Research Methods and Communication in the Social Sciences
This page intentionally left blank
Research Methods and C...
196 downloads
1651 Views
7MB Size
Report
This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!
Report copyright / DMCA form
Research Methods and Communication in the Social Sciences
This page intentionally left blank
Research Methods and Communication in the Social Sciences Tesfa G. Gebremedhin and Luther G. Tweeten
PRAEGER
Westport, Connecticut London
Library of Congress Cataloging-in-Publication Data Gebremedhin, Tesfa G. Research methods and communication in the social sciences / Tesfa G. Gebremedhin and Luther G. Tweeten. p. cm. Includes bibliographical references and index. ISBN 0-275-94930-3 (alk. paper) 1. Social sciences—Research. 2. Social sciences—Methodology. 3. Written communication. I. Tweeten, Luther G. II. Title. H62.G39 1994 300'.72-^dc20 94-13731 British Library Cataloguing in Publication Data is available. Copyright ©1994 by Tesfa G. Gebremedhin and Luther G. Tweeten All rights reserved. No portion of this book may be reproduced, by any process or technique, without the express written consent of the publisher. Library of Congress Catalog Card Number: 94-13731 ISBN: 0-275-94930-3 First published in 1994 Praeger Publishers, 88 Post Road West, Westport, CT 06881 An imprint of Greenwood Publishing Group, Inc. Printed in the United States of America The paper used in this book complies with the Permanent Paper Standard issued by the National Information Standards Organization (Z39.48-1984). P
To our families, Meheret Ghebremeskel and children and Eloyce Tweeten and children
This page intentionally left blank
Contents PREFACE
ix
ACKNOWLEDGMENTS
xi
1. THE FOUNDATION OF THE SCIENTIFIC RESEARCH METHOD
1
2. GUIDELINES FOR CONDUCTING SCIENTIFIC RESEARCH
35
3. GUIDELINES FOR PREPARING RESEARCH GRANT PROPOSALS
73
4. GUIDELINES FOR COMMUNICATING SCIENTIFIC RESEARCH
91
5. GUIDELINES FOR REVIEWING MANUSCRIPTS
109
6. GUIDELINES FOR LIBRARY LITERATURE SEARCH
117
APPENDIXES APPENDIX 1 -
125 Proposal Format Guidelines of the Northeast Regional Center for Rural Development (NERCRD)
127
Contents
Vlll
APPENDIX 2 -
APPENDIX 3 -
APPENDIX 4 -
INDEX
Sample Research Proposal 1: An Economic Analysis of Alternative Sources of Household Income for Farm Families
131
Essentials of a Project Outline for Hatch, Mclntire-Stennis, or State Matching Funds
145
Sample Research Proposal 2: An Analysis of Rural Health Services in Louisiana
149 163
Preface This publication evolved from graduate courses in research methodology taught by the authors. The material arose out of an apparent need by students for a more comprehensive understanding of the principles and techniques of research methods and writing skills. We attempt to provide a readily usable reference for conducting scientific research and developing scientific writing skills as well as resource material for teaching. The value of this publication for teaching and research has improved from the experience and responses of many students and other users of the material over the years. The structure and content of the text are intentionally simplified to facilitate a better understanding of each component of scientific research methods and technical writing skills. The text is designed to promote scientific thinking and communication. It applies scientific thinking to scientific research methods. It helps to develop skills required to pursue and conduct scientific research. It is designed to assist graduate students who have had limited experience in using scientific research methods and technical writing skills. However, the ideas and principles in this text are equally applicable to all students. Emphasis is on how to organize and communicate scientific research in papers, articles, theses, dissertations, and research proposals. The text consists of six chapters. The first chapter discusses the rudiments of research and the scientific method. The second relates the scientific method to scientific research. Chapter 3 outlines procedures for preparing a research grant or contract proposal. The fourth chapter presents guidelines for communicating scientific research. The fifth shows how to review manuscripts for publication. The final chapter outlines the guidelines for library literature search. Exercises for classroom teaching in graduate research methods are also included at the end of Chapters 1, 2, 3, 4, and 5.
X
Preface
Appendixes 1 and 3 contain two samples of specific proposal guidelines or Request for Proposals (RFPs) from funding agencies. Applications for a grant are prepared according to the guidelines of the RFPs listed in Appendixes 1 and 3. These applications, found in Appendixes 2 and 4, are sample research proposals.
Acknowledgments Every publication is a collaboration. Many of the ideas in this book are original, but more are borrowed from other authors-properly cited, of course. Special appreciation is expressed to the silent authors who indirectly served as joint authors of this publication. We acknowledge the valuable review comments and helpful suggestions of our friends and colleagues at West Virginia University, The Ohio State University, and Oklahoma State University. Special thanks is extended, particularly, to Dale Colyer, Virgil Norton, and Dennis Smith. Special gratitude is extended to Cloye Brower, Cynthia Dishon, and Lisa Lewis for typing the original and revised manuscripts. A special word of thanks is due to Renee Drury for typing and preparing the final copy of the manuscript. We are grateful to Cynthia Dishon and Shiva Makki for their assistance in writing Chapter 6. However, we are solely accountable for errors and views presented in this work. We are deeply indebted to West Virginia University and The Ohio State University for the staff and facilities we used in preparing this manuscript. We dedicate this work to the teachers, students, and administrators of West Virginia University and The Ohio State University. They encouraged us to complete the work when our motivation lagged. However, the views and opinions expressed herein are the authors' responsibility and do not necessarily reflect those of West Virginia University nor The Ohio State University.
This page intentionally left blank
Research Methods and Communication in the Social Sciences
This page intentionally left blank
1 The Foundation of the Scientific Research Method Creative thinking is our most fundamental human resource. The quality of our communication depends on the quality of our thinking. Creative thinking depends on skills (acquired through training) and not simply on intelligence. Making judgments, evaluating procedures, or contemplating alternative actions requires critical thinking. Effective scientific research and technical writing reflect scientific thinking. The first step in scientific thinking is organizing one's thoughts into a logical pattern, so the first step in scientific writing is organization. The maturation of one's thinking process is closely related to the development of a professional background, the adoption of an organized program of reading, and a concentrated effort to understand the methods of observation, experimentation, and interpretation. Reliable knowledge comes through scientific methods employing observable facts and logic. The scientific research method begins with curiosity about a given situation, which is articulated as a problem and frequently phrased in the form of a question. The problem may come from personal observation or from a consideration of work done by others. A tentative answer to the question or solution to the problem is formulated. This hypothesis is derived from previous experience and theory. The investigation proceeds toward a goal or solution. Problems are addressed by the method of investigation in an attempt to obtain evidence related to a hypothesis. The hypothesis directs the thinking of the researcher. If the problem is stated as a question, then each hypothesis is a possible answer to the question or a possible explanation. Data are gathered, and the hypothesis is tested as the analysis proceeds under a prescribed procedure.
2
Research Methods and Communication in the Social Sciences
The scientific method provides the framework for collecting and organizing the data so that they can be analyzed and interpreted. The collection of data highlights the empirical component of scientific research. Scientific research is not merely gathering data for the purpose of accumulating facts. Interpretation of the data is a key component of the research process. Interpretation employs a conceptual framework. A hypothesis that has gained widespread professional approval because it makes sense in a causal system and has not been rejected in repeated empirical sampling by numerous analysts working independently is sometimes called a principle, theory, or law. A law has value not only because it accounts for ("explains")past observations but also because it predicts future observations and experiments. Scientists need to retain a healthy skepticism. Hence laws or theories ordinarily are viewed as unrejected hypotheses that may be discarded for more plausible explanations and predictors as evidence accumulates (Barrass, 1978;Tuckman, 1978). The scientific research method draws conclusions based on the weight of evidence. Sometimes the evidence is quantified and statistical inference is used to measure the likelihood of making errors. Statistical inference enables the analyst to reject or fail to reject hypotheses based on probability. Due to the possibility of error in sampling and in measurement, analysts ordinarily do not accept a hypothesis. However, repeated inability to reject a hypothesis tested by numerous analysts working independently may elevate a hypothesis to the status of a law. An example is the law of diminishing returns, also called the law of variable proportions. The research process frequently provides feedback of information as the process is repeated by analysts. This characteristic of scientific research sets it apart from other branches of knowledge.
THE SEARCH FOR RELIABLE KNOWLEDGE As scientists, we never fully attain truth, but we strive for reliable knowledge. Personal experience, reasoning, and research have been prominent means for solving problems and understanding reality. To solve problems, most of us draw on personal experience—our own individual accumulated body of knowledge and skills from encounters and long acquaintance with facts and events in the environment. Much of the wisdom passed from generation to generation is the result of experience. Where solutions to problems clearly lie beyond personal knowledge, the wider or different experience of others with authority or expertise is sought.
The Foundation of the Scientific Research Method
3
Other means by which knowledge is discovered include reasoning and scientific research (Cohen and Manion, 1985). The following concepts and illustrations are useful in understanding the importance of critical thinking and the foundations of scientific research in the search for reliable knowledge. The concepts of Ary and colleagues (1979), Andrew and Hildebrand (1982),Emberger, Riley, and Hall (1955), Leedy (1981), Peterson (1961), Popper (1965),Tuckman (1978), Goldstein and Goldstein (1979), and Tweeten (1983) are incorporated in the following discussions. Science Science covers the broad field of human knowledge concerned with observed facts held together by principles and the relationships that exist among those facts. It is an organized body of knowledge, a method of ordered thinking, and a set of predictive propositions. Science is an intellectual, practical, and social activity undertaken for systematic growth of knowledge. Science seeks not only to understand and find reliable knowledge about the real world, but also to predict its natural phenomena through scientific investigation. Economists use principles of economics to predict the price of oil next year. Physicists use principles of science to predict the discovery of small particles such as quarks. In short, science is characterized by three features: (1) searching for understanding or knowledge, (2) expressing that understanding or knowledge achieved from science in the form of general laws, principles, or theories applicable to the widest possible variety of phenomena, and (3) testing experimentally the laws or principles achieved. Science differs sharply from other ways of understanding and explaining the world in its ultimate reliance on the authority and concept of an experimental test (Goldstein and Goldstein, 1979; Best and Kahn, 1986). In the physical sciences, the experiment often can be controlled. In the social sciences, where controlled experiments are rarely possible, we statistically control a dynamic environment to identify impacts of variables of interest. However, in both physical and social sciences, the ultimate test is the ability to predict and explain emerging outcomes. Scientific knowledge is always advancing as facts are accumulated and understanding of fundamental principles and laws is refined. For any knowledge to be reliable, it must be repeatedly tested and verified. Science is a continuing endeavor in which the end of one investigation is often the
4
Research Methods and Communication in the Social Sciences
starting point for another. This characteristic of science sets it apart from other branches of knowledge. Science also differs from other types of knowledge in that the scientific process depends on new ideas expanding or replacing old ones. For example, humanities such as theology, philosophy, and the arts deal with ideas about human nature and the meaning of life. Such ideas cannot be scientifically tested and accepted or rejected on empirical grounds.
Research Research may be defined as the systematic and objective recording and analysis of controlled empirical observations that may lead to the development of principles, laws, or theories resulting in prediction and possible ultimate control of events (Best and Kahn, 1986). Research applies the scientific method to study hypothetical propositions of presumed relations among phenomena. In physical science experiments, all variables but one may be held constant. Outcomes are observed as the one experimental factor is varied over some range of applications. In the social sciences, variables can rarely be held constant to isolate the impact of just one variable on outcomes. Multiple regression and other multivariate statistical techniques often are utilized in such circumstances to "control" the environment so that the impact of one variable on outcomes can be inferred. Characteristics of research can be summarized as follows (Best and Kahn, 1986): 1.
2. 3.
4. 5.
Research is directed toward the solution of a problem or the search for answers to unsolved questions. Analysts look for cause-and-effect relationships between the measures used in the study. Research emphasizes the development of principles, laws, or theories helpful in predicting future occurrences. Research is based upon observable experience or empirical evidence. It involves gathering new (e.g., survey) data from primary sources or using existing (secondary) data from published sources. Research demands useful, accurate, and factual observation and description by using quantitative measures where possible. Research strives to be objective and logical, applying appropriate tests to validate the procedures employed. References are
The Foundation of the Scientific Research Method
5
carefully documented, the data are collected, results are analyzed, and the conclusions are reached. Research in one branch of science often helps research in another. For example, advances in communication and computers have played important roles in facilitating research in other branches of science. The importance of research may be judged from the new knowledge, products, and services it has produced. These in turn have created higher standards of living (Tuckman, 1978; Peterson, 1961). The ultimate test of research is whether it provides reliable knowledge to serve objectives of society. These objectives include improving decisions to enhance the well-being of people and the quality of the environment. Basic research may have no immediate application nor provide usable information for solving immediate problems. It may seek to develop a model or theory that identifies fundamental factors in a particular environment or situation and hypothesizes their relationship. It may seek entirely new knowledge or extend the range and boundaries of existing scientific knowledge to better understand a particular subject. The investigation may be abstract and general without obvious practical use. Most basic research, however, is undertaken with the worthwhile, if distant, goal of accomplishing objectives that benefit society. A surprising number of scientific discoveries (penicillin is the classic example) have been made by luck. In this age of complex economies, technology, and social organization, the untrained tinkerer sometimes makes a lucky breakthrough. But the best "luck" comes to well-trained scientists working with modern techniques and apparatus in a facilitative, intellectual atmosphere. Alexander Fleming probably could not have discovered penicillin without such conditions. Some people justify basic research, including space exploration, by the serendipitous spin-off of products useful to society. The highest payoffs ordinarily come from purposeful research, however, whether basic or applied. Breakthroughs in superconductors, highly efficient batteries, or nuclear fusion energy are likely to come at least expense from research founded on developing those technologies. Albert Einstein's theory of relativity had no immediate application but subsequently influenced the broad sweep of applied research in physics. In economics, basic research often adds to or modifies theory. Nothing is as practical as good theory. Good theory may be no more than sophisticated common sense accumulated from controlled experiments or careful observations of the way the real world functions. In other cases, such as
6
Research Methods and Communication in the Social Sciences
neoclassical economic theory, mathematics has facilitated building a large, sophisticated conceptual edifice on simple assumptions of profit and utility maximization. Basic research extending neoclassical economic theory to the allocation of time and information has won Nobel prizes because it opened new fields of science with high payoffs to society. Thus basic research played an essential role in solving practical problems (Stock, 1985; Tuckman, 1978). Applied research is a specific, systematic, purposeful effort to solve real problems with more or less immediate applications (Stock, 1985; Tuckman, 1978). The primary objective of applied research is either (1) to suggest or recommend practical means of problem resolution, or (2) to provide information through in-depth analysis for decision makers (Andrew and Hildebrand, 1982). Applied research often has the characteristics of basic research, including the use of sampling techniques and subsequent inferences about the target population, formulating hypotheses, developing procedures, analyzing empirical results including tests of hypotheses, and reaching conclusions. However, its purpose is improving a product, decision, or process by applying theoretical concepts in actual problem situations (Best and Kahn, 1986). Hence problem-solving, applied research can extend to two additional stages: applying results (conclusions) and evaluating the consequences in the real world. If considerable resources of engineering and science are required to convert the outcome of applied research into a practical, functioning product for widespread use, we call that development. An example of applied research is to find the least-cost combination of farming practices to reduce soil erosion and pesticide contamination of surface and ground water to prescribed levels. Basic research, in contrast, might focus on the cellular biology linking pesticide intake to cancer in humans. A teacher seeking to resolve problems of the classroom will utilize findings of the latest educational research. Determining the chemical processes in the brain that retain memory is basic research; devising teaching techniques that reinforce long-term memory may be applied research. Of course, biological research is not necessarily basic, and social science research is not necessarily applied. Both social and biological sciences conduct basic and applied research. Applied research often draws on new findings from basic research. Effective applied research efficiently utilizes available research resources to raise the probability of achieving meaningful results that help solve problems. Because of its immediate usefulness, applied research is more widespread than basic research.
The Foundation of the Scientific Research Method
1
Private firms more often can appropriate profits for applied research than for basic research. Consequently, much basic research must be financed by the public sector, whereas much applied research is financed by the private sector. Studies find that, on average, basic research has as high a social payoff as applied research. In today's technocratic society, research is essential to address complex problems demanding objective analysis. Research is a means of verifying fact and ascertaining the state of affairs (diagnosis), predicting outcomes (prognosis), and prescribing solutions (remediation). Research is a methodical quest for truth and the solution of problems. The more wealthy and advanced a society, the greater the supply of resources for research and the greater the demand for output of research. All that is called "research" is not research. For instance, a student is sent to the library to find the composition of human blood. His instructor tells him to do considerable "research" in fulfilling the assignment. The student will be seeking information to dispel his own ignorance of the subject, but will not be doing research. The student has a problem indeed, but it is a personal rather than a research problem: he is uninformed. The answer to his problem has previously been found through research and recorded for his (and others') use in books in the library. A researcher encounters unresolved roadblocks to human knowledge having no answer or for which no answer has been found. Although complete truth may never be attained, we follow a process of inquiry to come as close as possible to understanding causes and predicting cures. Through research, we attempt to answer questions for which no answers exist, to solve problems that have not as yet been solved, or to reaffirm those answers and solutions that we assume to be valid but need confirmation (Leedy, 1981). We must use some common sense in deciding which problems research has a chance to solve at all or at least to solve with an acceptable commitment of resources. Some view research as essentially an attitude of mind, a way of thinking. The mental approach we bring to the quest for the answer to a problem differentiates research from other forms of truth-seeking attitudes. Historically, various pathways have been followed in the search for knowledge. Prehistoric man attempted to explain a problem through his own resources and imagination, personal thoughts, or religious beliefs. He invented myths, and myths grew into belief systems. Beliefs grew to become accepted answers for what otherwise would have remained a mystery. Such myths provided answers to deep philosophic questions but do not provide detail to make decisions in a modern economy.
8
Research Methods and Communication in the Social Sciences
Following explanation by myth came explanation by reason, in the age of science and technology. For this tradition of rational search for truth by reason, we owe much to the influence of ancient Greek philosophy and Christian thought in the Renaissance and Enlightenment periods. The attitude was to reason out the problem, to find the way out of a dilemma by thinking clearly and logically toward a conclusion drawn from basic premises. The pursuit of truth through rational thinking reached its summit in logical reasoning. Problem solving through logical reasoning was one way to provide answers to questions. But too often it was subjective and whimsical because it relied too much on untested premises and too little on the empirical processes gleaned by observing objective realities (Leedy, 1981).
TYPES OF REASONING To understand the nature of reality through an organized body of knowledge and a method of ordered thinking, two approaches are used: inductive and deductive reasoning. Deductive Reasoning Deductive reasoning can be described as a thinking process going from general principles to specific statements using prescribed rules of logic. It is a system for organizing what is known to reach research conclusions. It is finding the right rule for the circumstances and then applying it correctly. The mathematics discipline, for example, reaches deductive conclusions on the basis of general principles or axioms generally assumed to be true. The use of formal logic in a series of statements called syllogisms disciplines our thinking in deductive reasoning. Study of formal logic makes us more careful in reaching conclusions based on the weight of facts and logic. Syllogisms are merely systematic, widely accepted forms of reasoning. They consist of three propositions, two of which at least initially are assumed to be axiomatic and hence beyond the need for justification, substantiation, or proof. The basic proposition is known as the major premise, which is based on a "self-evident" proposition. A second statement, called the minor premise, typically is an application of the major premise to a specific situation or a particular instance. Applying the rule of logic to these two statements, the conclusion is an inevitable consequence. The basic assumption underlying the
The Foundation of the Scientific Research Method
9
syllogism is that through a sequence of formal steps of logic, from the general to the particular, a valid conclusion can be deducted from valid premises. An incorrect conclusion may be reached if the premises are incorrect or if the syllogism is incorrectly structured (see Cohen and Manion, 1985; Leedy, 1981). Example 1: All men are mortal. (Major premise.) Daniel is a man. (Minor premise.) Therefore, Daniel is mortal. (Conclusion.) Example 2: All birds can fly. A bat can fly. Therefore, a bat is a bird. In Example 1, the premises are correct and the syllogism is correctly structured. Logic leads to a correct conclusion. However, in Example 2, the major premise is incorrect (all birds cannot fly), and therefore the conclusion is suspect at best. From the fact that a bat can fly we cannot conclude that a bat is a bird. A bat can fly but is not a bird. We could have corrected the first premise to say that "some birds can fly,' but the shift from the universal "All" to the particularistic "Some" does not enable use to draw a valid conclusion. We could change the major premise and conclusion thus: All flying creatures have wings. A bat can fly; therefore, a bat has wings. The logic and premises stand scrutiny. In deductive reasoning, if the premises are valid and the logic is correct, the conclusion is also valid. Although deductive reasoning enables one to organize premises into patterns that provide a valid conclusion, we have noted that deductive reasoning has pitfalls. It is necessary to begin with correct premises. The conclusion of a syllogism cannot exceed the content of the premises. A conclusion reached from premises that are merely "true by definition" is called a tautology or truism. Syllogisms most useful in social science research have at least one premise with empirical content. Our confidence in the conclusion will depend not only on logic but also on the validity of the premise(s) tested by statistical inference or other means.
10
Research Methods and Communication in the Social Sciences
Firm conclusions can be difficult to draw in the social sciences because empirical premises often must be particularistic rather than universal. Who can afford to survey an entire population to be able to say that "All buyers respond to incentive X" or "No buyers prefer Y to Z?" Analysts attempting to draw firm conclusions from syllogisms under such circumstances are tempted to substitute definitions for empirical premises, or to arbitrarily elevate particularistic premises into universal premises. Analysts need to be alert to such pitfalls and the wrong and costly conclusions flowing from them. The first exercise at the end of this chapter provides examples of faulty syllogisms that have caused much mischief. Scientific inquiry cannot be conducted through deductive reasoning alone. Science requires empirical evidence as a basis of proof. Deductive reasoning is often hampered by preconceived notions, personal biases, ambiguities, incomplete or false premises, and the absence of supporting evidence (Best and Kahn, 1986). A discipline based solely on deduction, such as mathematics, is not a science. Despite its limitations, deductive reasoning is useful in the research process. Deductive reasoning can organize what is already known and can identify new relationships as one proceeds from the general to the specific. Premises or principles are a useful repository of previously acquired and validated knowledge. Deductive premises make research more efficient, eliminating the need to reestablish what is known when a new research project is conducted. Deductive reasoning provides a means for linking theory and observation. It guides research to deduce from existing theory what phenomena should be observed. Deductions from theory provide hypotheses central to scientific inquiry (Ary, Jacobs, and Razavieh, 1979; Tweeten, 1983). Inductive Reasoning Inductive reasoning is an empirical process of reaching a conclusion or arriving at new principles (general rules) from known data and experience by observing objective realities (facts). This type of reasoning is the reverse of the processes employed in deductive reasoning. Induction seeks to learn reality from observation. Inferences, therefore, do not depend upon previous knowledge. Induction alone is no more fruitful than deduction alone. Insofar as history is merely a collection of facts from the past, it is solely deduction and not a science. However, one motivated by George Santayana's dictum "Those who cannot remember the past are condemned to repeat it"
The Foundation of the Scientific Research Method
11
may apply science to history as a means to predict the future (Santayana, 1905,p. 218). As the Renaissance emerged out of the medieval Age of Faith, it was apparent that great logical edifices were "full of sound and fury, signifying nothing." The result was the Age of Reason and the Enlightenment, which emphasized rational approaches to solving problems: science flowered as inductive was joined to deductive thinking. The inductive process of moving from specific observations to generalization made deductive logic more powerful by giving premises empirical content (Best and Kahn, 1986). The new emphasis on inductive reasoning directed the attention of scientists to the real world for problem solution and demanded empirical evidence for verification. It is not possible to separate induction from deduction in useful science. An example of reaching a conclusion in deductive and inductive reasoning is illustrated below: Deductive Every fish has gills All sharks are fish Therefore, every shark has gills. Inductive Every shark that has been observed has gills Therefore, every shark has gills. It is apparent that the first premise of the deductive example must have empirical (inductive) content for the conclusion to be more than a tautology. Statistical Inference Statistics is a discipline that achieves an inductive conclusion on the basis of factual information or empirical results. Judgments whether to accept or reject conclusions from assembled data can be made based on the probability of making errors. The student t-test, widely used in the social sciences, can be used to illustrate statistical inference. The procedure begins by choosing a null
12
Research Methods and Communication in the Social Sciences
hypothesis, say that a true population parameter b = 0. For reliable science, the null should not necessarily be naught (zero) but should be the best estimate of the parameter based on prior knowledge (Tweeten, 1983). The value of t calculated from the sample is compared with a tabulated t. The t-test shows the probability of obtaining a sample value of / as large or larger when sampling from a population in which the null hypothesis is true. If the probability of obtaining the sample value of / is very small, we reject the null hypothesis. If the probability of drawing a t-value of the size actually obtained is very large, we fail to reject the null hypothesis. Note that we have only two choices: (1) reject or (2) fail to reject a hypothesis. Note that we never accept a hypothesis because some probability, however small, exists for error. Good science is skeptical of received wisdom. The conclusion that all science is a set of unrejected hypotheses is not comforting to lay persons, but it is valuable because it leaves science open to revision. As noted earlier, in deductive reasoning premises must be valid for consistently valid conclusions to be reached. Because in inductive reasoning a conclusion is reached by generalizing from samples to the whole class, it follows that inductive conclusions can be absolute only when the population in question is fully and accurately enumerated. Even complete enumeration of the population can bring incorrect conclusions due to measurement (enumeration) error, as noted in Chapter 2. Statistical inference more often focuses on errors in generalizing from a sample to the entire population than on within-sample measurement error, however. Combining Induction and Deduction in the Scientific Method Many problems cannot be solved by induction. Even if they could, induction alone would be inefficient because it fails to systematically utilize past information. Random collection of individual observations without a unifying concept or focus often obscures investigations and rarely leads to a generalization or theory (Best and Kahn, 1986). Inevitably, scholars learned to integrate the most important aspect of the inductive and deductive methods into a technique called the scientific method, which is shown in Figure 1.1 and discussed in more detail in Chapter 2. As shown in Figure 1.1, the inductive and deductive approaches are integrated to search for unresolved issues by operating inductively from observations (collection of factual data) to hypotheses, and then deductively
The Foundation of the Scientific Research Method
13
from these hypotheses to analyze their implications, and to test their validity from the standpoint of compatibility with accepted knowledge. In subsequent research, these hypotheses are revised based on previous findings and submitted to further testing and analyzing with additional data. This dual approach, in essence, is the modern scientific research method of empirical science (Cohen and Manion, 1985). Methodological models employing econometrics, simulation, or other frameworks often are used to facilitate and structure the analysis. The study of factors explaining academic achievement of elementary and secondary students illustrates how social science grows by inductive-deductive interactions and feedback. Early analysts explaining student achievement as a function of school variables (class size, teacher qualifications, spending per student, etc.) included characteristics of the parents as a control variable almost as an afterthought. To analysts' surprise, characteristics of parents such as educational attainment, income, occupation, and marital status dwarfed the impact of the schools on students' achievement. Repeated statistical significance of the parental background hypothesis when explaining student achievement made omission of the variable in subsequent studies unthinkable. Armed with modified theory recognizing the importance of parents, analysis introduced a new set of hypotheses in their research. Among these were the influence on schooling achievement of early reading to children and of single parenthood. Repeated statistical support for the influence of these factors led to new theories and attendant hypotheses. Some examples include the role of public assistance (welfare) in family solidarity and the role of preschool programs in compensating for inadequate parenting. In short, knowledge grows as theory interacts with empirical evidence using the scientific method as shown in Figure 1.1. In problem-solving research, additional steps can be added to Figure 1.1. These include decision making to implement a solution, execution of that solution, and bearing responsibility for the consequences of solutions (Johnson, 1986,p. 15). Sometimes the consequences may be credit or blame to an analyst. In other cases, the consequences can be financial success or failure to a firm, or socioeconomic progress or setback for a nation. A scientific approach enhances the empirical credibility of itsfindingsby appropriate standards and procedures for observation, experimentation, and confirmation of results by replication (Leedy, 1981). If the most distinctive feature of scientific research is its empirical nature, the next most important
14
Research Methods and Communication in the Social Sciences
Figure 1.1 The Scientific Method
characteristic is its set of procedures, which not only show how findings have been arrived at but are sufficiently clear for replication and reproduction (Cohen and Manion, 1985). Even a scientific inquiry, for instance, in economics, psychology,chemistry, or physics, can run into pitfalls causing faulty conclusions. Some examples are presented to illustrate the pitfalls in scientific study (see Miller, 1991). The confusion between correlation and causation is one pitfall. Many facts in the real world seem to be correlated-characterized by mutual relationships between two or more things. One variable appears to move with or opposite the movement of another. Statistical correlation in scientific
The Foundation of the Scientific Research Method
15
methodology is not equivalent to causation. You can probably think of lots of correlations that do not necessarily imply causation. Another pitfall in scientific thinking is the fallacy of composition, or believing that what is true for a part is therefore true for the whole. The whole is not always the sum of the parts. Otherwise stated, what is true for the individual is not necessarily true for a group. One farmer can increase profit by adopting a new technology. But profit falls due to increased industry output and consequent falling market price when all farmers adopt the new technology. Electronic facilitators of information technology-computers of all varieties, scanning devices, and telephone and satellite communication systems-are available to capture, store, manipulate, analyze, and communicate data. New, innovative uses of information technology provide tools to enhance scientific research. However, low-cost data handling systems can cause undue output of undigested information. The result is information overload of decision makers. Statistical tests mean little when we let the computer pick, say, five statistically significant variables (at the 0.05 probability level) from 100 prospects to predict the price of food. Resulting statistical bias can lull analysts to explain everything (in the past) and predict nothing in the future with reliability. Also, as modern data and analysis systems overwhelm us, we depend so much on computers that we may fail to develop common sense and communication skills. The result is paralysis of analysis, as a flood of information reduces rather than enhances our ability to make sound decisions
ATTRIBUTES OF WORTHWHILE SCIENCE A worthwhile science has internal validity and external worth (Tweeten, 1989,ch. 3). A scientific investigation can be said to be internally valid to the extent that, within its own confines, its results are credible. A scientific discovery is internally valid if it is clear, logical, and reliable (Ary, Jacobs, and Razavieh, 1979; Travers, 1978; Tuckman, 1978). For such results to be useful, they must be generalizable beyond the confines of the particular investigation. Without internal validity, a scientific investigation cannot be externally useful.
16
Research Methods and Communication in the Social Sciences
Internal Validity A science does not have to be simple, but it needs to have clarity in the sense of being no more complex in analysis nor demanding of assumptions than necessary. This principle of clarity is sometimes called Occam's Razor after the medieval English cleric, William of Occam. A related principle of parsimony calls for no more parameters and assumptions than necessary in research models. Internal validity also requires a logical or coherent structure. We expect internally valid science to make sense. The analyst should be able to make a case that A is caused by systematic structural forces B, C, and D. In addition, A may be influenced by random elements called unexplained shocks or disturbances. Internal validity of science depends on reliability as measured by the ability of its propositions to predict. That predictive ability is apparent in how closely the propositions correspond to past experience and, more important, in how accurately future or previously unknown phenomena are predicted from forces B, C, and D. A science is objective if researchers are willing to have their work reviewed for clarity, logic, and predictiveness and are willing to abide by the results. Objectivity is apparent if two or more researchers independently studying the same phenomenon consistently reach the same conclusions. Intellectual honesty or integrity in research represents a higher calling to rise above conventional wisdom and above political and social pressures to report findings well grounded in fact and logic (see also Travers, 1978). When scientists fabricate or selectively report data or in other ways "cook" results by managing them to fit biases, science is discredited and society is poorly served. Our best effort at objectivity, although never fully achieved, is what we owe to society for the privilege of being public researchers and educators. Stated another way, objectivity is the price we pay for maintaining the integrity of science, without which education is impossible. To promote objectivity and intellectual integrity, ten suggestions are offered: 1.
Beware of spurious correlations. Carl Wilken noted that gross farm income was one-seventh of national income in the 1930s, then erroneously went on to conclude that raising farm income by $1 would raise national income by $7. Adolph Hitler noted that no German field marshall had ever been captured by the enemy. He raised the rank of General Paulus, who was commanding
The Foundation of the Scientific Research Method
2.
3.
4.
5.
6.
7.
German troops surrounded at Stalingrad, to field marshal, but Paulus and his army were captured by the Soviets. Beware of averages. The fact that on the average only 14 percent of consumers' income is spent on food is little consolation, when food prices rise, for the many low-income consumers who spend over 40 percent of their income on food. Having feet in water averaging 80 degrees is not comfortable if one foot is in 40degree water and the other in 120-degree water. Beware of case studies. Case studies are anecdotal evidence with unknown generality. An intellectual scoundrel can find a case study to support almost any position. A random sample reduces opportunities for bias. Beware of injecting personal bias. Some fields of pseudoscience behave much like cults. They have their dogma, prophets, gods, devils, apocalypse, and millennium. In the physical sciences, examples of quack medicine and research are perhaps more obvious than in the social sciences. Beware of experts. Many researchers too willingly accept the views and findings of recognized authorities in their field. A scientist must always be a skeptic but must avoid becoming a cynic. Beware of data. Data can be of low quality even if printed in a slick publication with a hard cover. To use an actual unfortunate case, the state level demand function for farm tractors was estimated using state data on tractor sales and farm income constructed by a national agency. The demand analyst was unaware that the national agency had constructed the tractor demand quantity data from a single national number allocated to each state based on farm income, farm numbers, and other variables. (This procedure was used because the single national number was reliable for the nation but was not reliable for states.) The analyst's well-fitting demand functions were merely reproducing the way the data were constructed. Beware of persuasive people who can bias views of researchers. Many agencies, firms, or people with self-interests to protect will have every reason to bring the researcher to their point of view. Proprietary funding of social science research may bias outcomes. Liberal or conservative friends or colleagues intent on pushing their agenda can also introduce bias, perhaps subconsciously.
17
18
Research Methods and Communication in the Social Sciences 8.
Beware of failing to serve your client. A publicly employed university agricultural economist serves society through agriculture. An economist who becomes an advocate of special interests at the expense of society has lost intellectual integrity. 9. Beware of intervening or mediating variables. A study showing that rates of economic growth are higher without foreign trade than with may be meaningless if differences in growth rates among the countries being studied are determined by culture, technology, stage of development, and form of government. 10. Beware of the Hawthorn effect or Heisenberg principle. The very act of experimentation changes results. Control groups used in social sciences are likely to change their behavior when repeatedly interviewed. Of course, if the experimental and control groups are both interviewed and react similarly to the interview, Hawthorn differences between the control and experimental groups will average out and the experimental effect may be measured accurately.
External Worth A science with only internal validity may have value as a curiosity or arcane diversion to researchers, but the main value of science to society derives from external worth. Such worth is in turn a function of (1) the ability to turn human and material resources devoted to the science into output, an attribute closely related to internal validity, and (2) the value of that output. Trade-offs are apparent. Meteorology is unable to turn its resources into a product (forecasts) with much reliability. But weather information is so highly valued that imprecision is tolerated. Economics presents similar situations. On the other hand, using a minimum of resources, some fields of astronomy are able to predict with precision the course of objects in space. Even if predicting the course of a distant star is of little value, the science of astronomy is considered valuable. Thus, two very different sciences can be equally worthwhile: in one, predictions are highly valued even though resources devoted to it seldom predict correctly, whereas in the other, science consistently predicts well, but each precise prediction is not of much value.
The Foundation of the Scientific Research Method
19
POSITIVE AND NORMATIVE ECONOMICS In a search for intellectual integrity and respectability in conducting and reporting research, economists have classified the discipline into various approaches as a way of ruling out certain forms judged to be unscientific. Two broad classifications are normative economics, which is shunned, and positive economics, which is embraced. Normative economics is the economics of "what ought to be." In its extreme form, normative economics entails advocacy by economists of ends and means that they themselves subjectively have judged to be best for society rather than ends and means that society wants and that would improve the well-being of people. A publicly employed economist who advocates Marxism or laissez-faire policies for making society better off is practicing normative economics and is operating as a politician or preacher and not a scientist. The objective is to avoid advocacy of subjective issues but not to avoid issues dealing with human values. Measuring well-being or utility is a valueladen task long shunned as unobjective by economists. As science devises more objective means to measure well-being, it will be more feasible to analyze and provide objective information and policy options to policymakers on value-laden issues of what is good or bad for the well-being of people. There are times for an economist or sociologist to step out of the role of social scientist and into a role of private citizen or politician free to practice advocacy, but that changing of hats from social scientist to private citizen needs to be recognized by all concerned. In scientific writing, avoid words or phrases such as "should," "must," "ought to," "good," and "bad," which imply moral imperatives and value judgments. Issues concerning values are central to almost everything in social science. Social scientists are frequently asked to weigh the evidence and give their best judgment (opinion) on controversial issues. It is proper for scientists to state, "After comprehensive analysis of the options, here in my judgment is the best course of action," and give reasons why. Persons addressing issues concerning the environment, for example, may deem the field so important that strong advocacy is warranted in the absence of data and analysis. Several ways to avoid a messianic approach that is more appropriate for an evangelist than a scholar include: (1) state value judgments up front so the public is forewarned; (2) use data and analysis honestly and objectively, even when the goal is, say, the value judgment that an endangered species is to be preserved; and (3) present all sides of issues as fully as possible.
20
Research Methods and Communication in the Social Sciences
It is fashionable among some social scientists to argue (correctly) that subjectivity attends every step of the research process, from the choice of project, to the processing of raw data into statistics, to the interpretation of statistics in drawing conclusions. They go on to reason from this premise that there is no objective reality. They contend that because all research is subjective and biased, activists should get their own (biased?) research to counter the biased research of the establishment. If such thinking becomes widespread, it will undermine science. The answer to subjective science is not additional subjective science but rather is objective science. The several variants of positive economics are legitimate activities for professional economists. Descriptive economics is the purest form of the positive economics of "what is." It merely describes the economy without making judgments of "what ought to be" or predicting "what will be." It may be analytical, but because it does not predict, it is a primitive "form of economics. It has much the same status as the history discipline-it is primitive science at best. It is useful for some purposes but falls short of living up to the capability of economic science to be worthwhile. Predictive economics—the economics of "what is likely to be"—is another form of positive economics. This approach to positive economics is mostly concerned with forecasting and need not confront issues of what goals are being met for whom. Instead, it focuses on issues such as the likely state of the economy or the price of hogs next month or next year. Prescriptive economics, another form of positive economics, is the economics of "what could be" and is the principal basis for economic policy analysis. It utilizes if-then statements. It does not advocate objectives invented by the economist but holds that // the economic decision maker wishes to pursue objective X, then the most efficient use of resources is allocation Y. Prescriptive economics may assume a norm or objective such as profit maximization, a more equal distribution of income, risk minimization for the firm, or utility maximization for consumers and society. Linear programming or simulation used to determine allocations to increase income, employment, well-being, or other recognized objectives or norms of society is positive prescriptive economics. Such analysis is not normative because it does not advocate what ought to be. It specifies the prescriptive economics of "what could be" if the objective is profit or other norms, hence is conditionally positive. In keeping with the tradition of positive economics under such prescriptive economics, it is wise to specify alternative means or options that would achieve various objectives, letting the decision maker observe trade-offs before choosing policy options.
The Foundation of the Scientific Research Method
21
SCHOOLS OF THOUGHT Institutions researching agricultural economics are recognized for their emphasis on various components of the scientific method. Schools of thought are listed in Figure 1.2. Induction or empiricism characterizes those institutions that emphasize use of primary or secondary data to provide research results. The Economic Research Service in the U.S. Department of Agriculture relies heavily on that approach. The agency does so for a number of reasons, such as (1) having sizable resources to collect or access large primary and secondary data bases,and (2) working with Congress and other decision makers who are impatient if not downright distrustful of conclusions based on theory as opposed to facts. Other institutions emphasize deduction in a school of thought called rationalism. Rationalists emphasize theory, nonempirical principles, and logic. Rationalism is particularly prominent in academic disciplines with substantial scholarship but inadequate research funds to collect and manage data. A chief pitfall of rationalism is reification, the error of elevating economic principles to the level of truth. Rationalists sometimes make extensive use of data but "massage" the data and analyses until they conform to perceived principles. Marxist economists practice rationalism. The socalled Chicago school of economics has sometimes been accused, often unfairly, of practicing reification in defense of the market. Pragmatism emphasizes that research begins with a problem and ends with a solution to that problem. Most social scientists probably are pragmatists in that they believe science should solve problems. The philosopher John Dewey is regarded by many as the father of pragmatism. Pure pragmatists are eclectic about choice of methods. They properly use induction, deduction, or whatever legitimate means are useful to solve the problem. A shortcoming of pragmatism is that practitioners frequently become populists, unwilling to invest the time and resources necessary to analyze the ultimate implications of their recommendations. A simple, popular, and incorrect solution can be found for almost any complex problem. Social sciences are especially tempted to fall into the populist trap because populist audiences often are their most vocal and aggressive supporters. Analysts "captured" by special interests are prone to shallow or unobjective research. Other social scientists deride economists for emphasis on neoclassical theory and instead propound a multitude of problem-specific theories. Such theories explain everything and predict nothing. Three great failures of lay
22
Research Methods and Communication in the Social Sciences
Figure 1.2 Schools of Thought
analysis-failure to account for indirect, long-term, and unintended effects-too often result in counterproductive public policies or private-firm strategies. Pragmatists say they are concerned about people rather than about theory. Such excuses often attend special pleading in favor of policies for a special interest group at the expense of society. In-depth analysis might show that the policy advocated hurts others more than it benefits the special interest group. Institutionalism is a derivative of pragmatism and combines the problem-solving approach with economics, law, political science, history, and other disciplines. John R. Commons of the University of Wisconsin is one of the fathers of institutionalism. Some of its early applications were to resolve economic problems of the cutover area of the northern Great Lakes,
The Foundation of the Scientific Research Method
23
which could not be solved by conventional economic tools or the market alone but required government intervention. Institutionalists properly place much emphasis on a close working relationship with institutions such as public agencies and private firms relating to their research. Institutionalists work closely with involved agencies to know their subject, plan and conduct research, and implement findings. As with pragmatism, a weakness is that too close a relationship to a clientele tends to bias the researcher so that he or she becomes a normative advocate of a special interest group to the detriment of the public interest. Empiricism alone or rationalism alone is inadequate. The two are joined in the school of thought called experimentalism (Figure 1.2). Experimentalism became prominent in economics at Iowa State University, where a strong statistics program (borrowed to no small extent from the British) was combined with economic theory. Econometrics (combining statistics, mathematics, and economics) was used to estimate the magnitude and statistical significance of parameters such as elasticities of supply, demand, and production. Through use of statistical inference, economics was made more scientific by reducing the subjective judgments of economics to objective judgments based on statistical probabilities. A serious shortcoming of experimentalism is that many of the most important problems in economics cannot be quantified and subjected to the methods of controlled experiments or even of statistical inference, hence are not suitable subjects of research for experimentalists. Disenchantment with experimentalism and institutionalism led to combining advantages of the various approaches into a school of thought called relativism. Relativism holds that research begins with real problems, that the scientific method employing induction and deduction are appropriate, and that values are inseparable from research. The challenge is to make these values explicit and treat them objectively in research. For example, research at Oklahoma State University used tools of sociology, psychology, and econometrics to estimate the marginal utility of income (Tweeten and Mlay, 1986). This enables economists to work with seemingly subjective issues of equity (optimal income distribution) in a more objective manner than before. The approach has advantages over traditional positivisticeconomics, which divides the economics field into dimensions of equity and efficiency, then throws away the equity issues in favor of studying "objective" efficiency issues. Social scientists outside of economics also frequently divide problems into efficiency and equity dimensions, then throw away the efficiency issues
24
Research Methods and Communication in the Social Sciences
to concentrate on equity. Confining inquiry to equity alone or efficiency alone is inadequate. Critics have accused the economics profession of becoming increasingly irrelevant to the real world and instead caught up in arcane and esoteric theory and mathematical formulations. That criticism is not new but has become more intense, as economists must spend more of their professional career learning a narrow specialty to be able to read the appropriate literature, write for journals, and obtain university tenure. The profession is now large enough to support pathbreaking basic, perhaps arcane, research as well as pragmatic, broad-based applications of economics to pressing current issues. The problem is incentives. Applied journals need to be encouraged, and articles in them need to be given as much weight for tenure as that given to articles in journals reporting basic disciplinary research. Earning a Ph.D. in social science provides no exemption from using common sense. A recent rural sociology book advocated the discredited conspiracy theories and central planning despite the collapse of Marxism in much of the world. The book gave the impression of being scholarly, with numerous references, footnotes, and difficult terminology. The author seemed to have read everything but the newspapers! Collaboration with someone possessing practical experience in economic development might have avoided a propaganda piece. In conclusion, each of the schools of thought listed in Figure 1.2 makes a contribution to science. The challenge to the scientist is to draw on the best from each school of thought. It is recognized that tension and conflict will arise among social scientists over the bounds of objective science; i.e.,the question will arise, is relativism objective? Nonetheless, the thrust of future social science is toward recognizing and quantifying the social welfare function with arguments (variables) measuring the level, distribution, and variability of income (see Tweeten, 1988). Many of the variables once considered to be too subjective for science will one day be measured with sufficient reliability to be judged as being objective.
CROSS-DISCIPLINARY RESEARCH In appraising opportunities for cross-disciplinary research, it is useful to begin with Glenn Johnson's (1986, pp. 11-27) classification of research. Disciplinary research extends the frontiers of a single discipline such as economics, sociology, or mathematics. Subject matter research, such as agricultural economics, combines disciplines of agriculture and economics.
The Foundation of the Scientific Research Method
25
Problem-solving research seeks to alleviate a problem such as crime, drug use, poverty, or environmental degradation. Individual interdisciplinary research occurs when one individual combines more than one discipline in research. Multidisciplinary research combines individuals representing various disciplines. Each specialty is called upon as needed to address all dimensions of the project and accomplish the research mission. Analysts in each discipline prepare reports for their profession and write for their journals. Collaborative interdisciplinary research occurs in a multidisciplinary setting. Analysts bring their unique specialties, but interact to cross-fertilize the overall effort from their unique disciplinary perspective. They dialogue among disciplines and coauthor publications. These seek a coherent product that joins the contributions of various disciplines. They do not stop with one report from political scientists, one from sociologists, one from economists, and one from demographers. In nature, members of a species separated too long become separate species unable to inter-breed. Without collaborative interdisciplinary efforts, the same outcome can befall social scientists. Disciplinary research extending pure theory is likely to require little interdisciplinary content and no multidisciplinary content. Someone writing for a theoretical journal must have great depth in his discipline, but the work is never "pure" because some knowledge of grammar and perhaps mathematics is required. Subject matter research is often interdisciplinary for the individual. Any agricultural economist doing farm management research is expected to have knowledge of economics, agriculture, and statistics required to do the job and to conduct research without bringing in a sociologist or computer scientist. Problem-solving research, say, to address an environmental hazard or inner city blight, is no respecter of disciplinary boundaries. Such issues are likely to require expertise of several disciplines from the social and perhaps physical sciences. We have spent much of our careers in multidisciplinary settings. Our experience is that collaborative interdisciplinary research is unusual. That is, in most settings the economists, sociologists, political scientists, or professionals from other disciplines working on the same problem will operate independently. Rarely do scientists from economics and sociology cross-fertilize or coauthor publications. One reason is a shortage of journals publishing collaborative interdisciplinary efforts. Journal editors and reviewers usually require a rich sprinkling of jargon and author citations from the discipline that journal represents.
26
Research Methods and Communication in the Social Sciences
The following suggestions are offered for success in collaborative interdisciplinary research: 1.
2.
3.
The administrator(s) in charge of the various disciplines or multidisciplinary units must be committed to foster truly collaborative interdisciplinary work and reward it. Researchers must be compatible and committed to collaborative interdisciplinary work. Neoclassical economists may not work well with Marxist sociologists, for example. Researchers need a disciplinary home to keep abreast of subject matter, earn and receive professional recognition including tenure, and interact socially and professionally with colleagues. Because issues change but disciplines remain, the best approach is to bring together persons from necessary disciplines only as long as necessary to complete a multidisciplinary study. Thus, specific multidisciplinary teams can form to meet specific needs. Special funding, careful selection of participants, and leadership can help turn such efforts into truly collaborative interdisciplinary efforts.
COPING WITH PARADIGMS As noted above, multidisciplinary research has been thwarted by the difficulties of communication between scientists in different disciplines. A reason is that the deepening of knowledge within specialties requires scientists to become narrower, while many of the problems faced by society, such as the environment, become broader. It is useful to digress to explore some explanations for the narrowness of disciplines and how disciplines evolve over time. Some penetrating insights are provided by Thomas Kuhn (1970). He writes about scientific paradigms. On the one hand, a paradigm stands for the entire constellation of beliefs and values shared by members of a given scientific community. On the other hand, it denotes concrete methodology employed as models or theories that can replace explicit rules as a basis for solution of the remaining puzzles of normal science. A scientific community consists of persons who share a paradigm. A superior paradigm is recognized by accuracy of predictions, particularly of quantitative predictions, and the number of different problems solved. Less useful,
The Foundation of the Scientific Research Method
27
though also important, are values such as simplicity, scope, and compatibility with other specialties. In the early stages of science, different men confronting the same range of phenomena describe and interpret them indifferent ways. But according to Kuhn (1970), a unique characteristic of a scientific discipline is that such initial divergences disappear once and for all as a science matures. Their disappearance is caused by the triumph of one of the preparadigm schools. The confidence in a paradigm causes the end of interschool debate and, because it assures scientists that they are on the right track, encourages them to undertake more precise, esoteric, and consuming sorts of work. No part of the aim of normal (conventional) science is to call forth new sets of phenomena; indeed, those that will not fit the box are often not seen at all. Nor do scientists usually aim to invent new theories, and they are often intolerant of those invented by others. Instead, normal scientific research is directed to the articulation of those phenomena and theories that the paradigm already supplies. A discipline steeped in its extant paradigm restricts the scientist's vision. The science becomes increasingly rigid. But normal science leads to a detail of information and to a precision of the observation-theory match that could be achieved in no other way. Novelty ordinarily emerges only for the scientist who, knowing with precision what he should expect, is able to recognize that something has gone wrong. Anomaly appears only against the background provided by the paradigm. By ensuring that the paradigm will not too easily be surrendered, resistance guarantees that scientists will not be lightly distracted and that the anomalies that lead to paradigm change will penetrate existing knowledge to the core. Journal editors and reviewers guard the integrity of the paradigm from the barbarians who think differently. The fact that significant scientific novelty so often emerges simultaneously from several laboratories is an index both to the strongly traditional nature of normal science and to the completeness with which that traditional pursuit prepares the way for its own change. As in manufacturing, retooling is an extravagance to be reserved for the occasion that demands it. The significance of crises is the indication that the time for retooling has arrived. A scientific theory is declared invalid only if an alternative candidate is available to replace it. Very often, those who achieve fundamental inventions of a new paradigm have been either very young or very new to the field whose paradigm they change. Ordinarily, new paradigms do not contain the old, but rather reject and displace the old-according to Kuhn (1970). When paradigms change, there are usually significant shifts in the criteria determining the legitimacy both of problems and of proposed solutions. The
28
Research Methods and Communication in the Social Sciences
crisis that brings about a change in paradigm is evoked when an alternative candidate exists, and the testing occurs as part of the competition between two rival paradigms for the allegiance of the scientific community. The contribution is based on the ability of different theories to explain the evidence at hand. Theories are not true or false, they only differ in their ability to explain and predict phenomena. The issue is which paradigm should in the future guide research on problems, many of which none of the competitors can claim to be able to resolve completely. Scientific revolutions close with a total victory for one of the two opposing camps. In summary, Kuhn's (1970) central proposition is that science evolves through a metamorphic process. Normal science (the conventional wisdom including existing theory, assumptions, and perceptions) is gradually undermined by incongruence with perceived reality. Eventually, the accumulated inconsistencies, in the presence of an available new paradigm that predicts and explains better, give rise to a scientific revolution, and the old paradigm is replaced by the new. The new paradigm envelopes the scientific community. Kuhn's (1970) paradigm has had mixed success measured by its own standards. Evolution, not revolution, characterizes the neoclassical paradigm dominating economics for decades. Some challenges (e.g., socialism or Marxism) have proven unworkable; others (e.g., Keynesian economics) have been absorbed by becoming mere appendages to neoclassical economics. While economics has been criticized for having too narrow a paradigm, other social sciences have been criticized for having too many paradigms. Neoclassical economics imperfectly predicts the future. Other social sciences with numerous theories have even more difficulty predicting the future. Some theory can be found to explain almost any outcome ex post facto! But ex ante predictions are difficult because there is no basis for choosing which theory to use. What are the lessons for students of research methods? One is the importance of understanding and communicating with other disciplines. Another lesson is to beware of the status quo in any discipline.
SOME COMMON GROUND-SYSTEMS ANALYSIS Every discipline has its unique theories. However, some research strategies cross disciplinary lines. An example is systems analysis. Systems analysis is not a technique, but is a systematic way of solving problems. To
The Foundation of the Scientific Research Method
29
the extent that it represents the common sense that any good researcher uses in solving problems, systems analysis is as old as problem solving itself. Although no single, concise definition describes systems analysis, the approach to solving problems does have certain more or less accepted characteristics. 1.
2.
3. 4.
5.
Recognition of the total problem; all parts of the phenomenon in question that bear significantly on the solution are accounted for within the system. Of course, abstraction from reality to model essential elements is unavoidable. That is, a model attempting to represent the complexity of reality collapses of its own weight. Each component or subsystem must be understood in its relationship to other subsystems and the total system. The problem solution must recognize time sequences and meshing the components properly in reaching the chosen target(s). The system is tied together by communication networks and other linkages. The process is monitored for efficiency. Cost effectiveness, benefit-cost analysis, program planning and budgeting, and other terms describe evaluation techniques. The performance of the system is evaluated in relation to the targets or objectives, with feedback to adjust the process from information gained from experience. Quantitative approaches such as simulation and programming are often used to gain "experience" through small-scale operation of the system. The study of the control mechanism for the system is sometimes called cybernetics.
A POSTSCRIPT ON RESEARCH PLANNING This brief section on research planning serves as a transition from systems planning to the components of the research process discussed in detail in the next chapter. A systematic approach to planning research can avoid needless delays and bring a quality effort while meeting deadlines.
30
Research Methods and Communication in the Social Sciences Program Evaluation and Review Technique (PERT)
Figure 1.3 Sample PERT Chart
The suggested tool is a Program Evaluation and Review Technique (PERT) chart-a planning and scheduling approach. The PERT chart may look like Figure 1.3. It outlines tasks by timing. Failure to meet critical deadlines in the plan signals your need to redouble efforts by working smarter or more nights and weekends to catch up. A PERT chart highlights hard-to-get materials, such as unpublished literature or obscure data, for which a search needs to begin early in order to have it when needed. A final note is appropriate. An excellent background of training, careful planning, and a sound strategy are of little help to a procrastinator. Lack of drive to move ahead with research is probably the greatest problem among today's students. Research is often a solitary task requiring self-motivation. Inner drive is strengthened when a student sees the importance of the issue being addressed, the excitement of finding a solution to a problem and adding to the knowledge base, and the joy of interacting with stimulating people in a research environment. Don't necessarily count on motivation provided by outsiders who are badgering, cajoling, or giving pep talks.
The Foundation of the Scientific Research Method
31
REFERENCES Andrew, Chris, and Peter Hildebrand. (1982). Planning and Constructing Applied Research. New York: MSS Information Corporation. Ary, Donald, Lucy Cheser Jacobs, and Asghar Razavieh. (1979). Introduction to Research Education. Second edition. New York: Holt, Rhinehart, and Winston. Barrass, Robert. (1978). Scientists Must Write (A Guide to Better Writing for Scientists, Engineers, and Students). New York: Wiley. Best, John W., and James V. Kahn. (1986). Research in Education. Fifth edition. Englewood Cliffs, NJ: Prentice Hall. Blaug, Mark. (1980). The Methodology of Economics. Cambridge: Cambridge University Press. Bonnen, James. (1988). "Improving the Socioeconomic Data Base." Pp. 452-489 in R. J. Hildreth, K. L. Lipton, K. C. Clayton, and C. C. O'Connor, eds., Agriculture and Rural Areas Approaching the Twenty-First Century. Ames: Iowa State University Press. Cohen, Louis, and Lawrence Manion. (1985). Research Methods in Education. Second edition. Dover, NH: Croom Helm. Emberger, Meta Riley, and Marian Ross Hall. (1955). Scientific Writing. New York: Harcourt, Brace. Goldstein, Martin, and Inge Goldstein. (1979). How We Know (An Exploration of the Scientific Process). Second Edition. New York: Plenum Press. Johnson, Glenn. (1986). Research Methodology for Economists. New York: Macmillan. Kuhn, Thomas. (1970). Structure of Scientific Revolutions. Second edition. Chicago: University of Chicago Press. Leedy, Paul D. (1981). How to Read Research and Understand It. New York: Macmillan. Miller, Roger LeRoy. (1991). Economics Today. New York: HarperCollins. Muller, Jerry. (1993). Adam Smith in His Time and Ours: Designing a Decent Society. New York: Free Press. Nelson, Robert. (1991). Reaching for Heaven on Earth: The Theological Meaning of Economics. Savage, MD: Rowman and Littlefield. Peterson, Martin S. (1961). Scientific Thinking and Scientific Writing. New York: Reinhold. Popper, Karl. (1965). The Logic of Scientific Discovery. New York: Harper Torchbooks. Santayana, George. (1905). "Reason in Common Sense" in The Life of Reason, vol. Ill, ch. XII. New York: Charles Scribner's Sons. Stock, Molly. (1985). A Practical Guide to Graduate Research. New York: McGraw-Hill. Travers, Robert M. W. (1978). An Introduction to Educational Research. Fourth edition. New York: Macmillan.
32
Research Methods and Communication in the Social Sciences
Tuckman, Bruce W. (1978). Conducting Education Research. Second edition. New York: Harcourt Brace Jovanovich. Tweeten, Luther. (August 1983). "Hypothesis Testing in Economic Science." American Journal of Agricultural Economics 65:548-552. . (1988). "Domestic Food and Agricultural Policy Research Directions." Pp. 121-144 in RJ. Hildreth, K. L. Lipton, K. C. Clayton, and C. C. O'Connor, eds., Agriculture and Rural Areas Approaching the Twenty-First Century. Ames: Iowa State University Press. . (1989). Farm Policy Analysis. Boulder, CO: Westview Press. Tweeten, Luther, and Gilead Mlay. (1986). Marginal Utility of Income Estimated and Applied to Economic Problems in Agriculture. Agricultural Policy Analysis Project Report B-21. Stillwater: Department of Agricultural Economics, Oklahoma State University.
EXERCISES 1. Examine the following syllogisms one by one. Do you agree with the conclusion? If you feel the conclusion is wrong, is the reason a faulty premise(s), faulty logic, or both? State how you will correct each faulty syllogism. a. All willful taking of the life of a human being is murder. Abortion is willful taking of the life of a human being. .". All abortion is murder. (.'. means therefore.) b. Imports of rice into Thailand have increased. The staple food of Thailand is rice. .'. Thailand should produce more rice. c. All Communists are liberals. A is a liberal. .*. A is a Communist. d. All farmers benefit from improved price forecasts. This study develops models to forecast prices. .". Farmers will benefit from this study. e. The U.S. economy is performing unsatisfactorily. Economists provide remedies for unsatisfactory performance of the economy. Economists are responsible for unsatisfactory performance of the economy. f. Perfect competition is the only efficient economic system. The free market is perfect competition. .'. The free market is an efficient economic system.
33
The Foundation of the Scientific Research Method
2. Apply the scientific method step-by-step to solving the problem of why a car will not start. Do professional car mechanics use the scientific method? 3. A t-test is a widely used form of statistical inference and is explained in any introductory statistics book. In a single linear regression of the form Y = a + bX + e, assume that Y is the dependent variable, X is the independent variable, e is the error term, and a and b are estimated coefficients. Explain the meaning of a t-test of the statistical significance of b. 4. An obscure tinkerer claims she has invented a car battery 10 times more efficient than the conventional battery and cheaper to produce. The "establishment" engineers and scientists from the big corporations say that her invention is bogus and that she is a fraud. She says they are just trying to protect the battery industry from competition. Who would you believe? Why? How would you go about finding the facts or the truth? 5. Into what schools of thought in Figure 1.2 reasons?
do you fall?
For what
6. Until 1992, every time the Minnesota Vikings won a football game just before the presidential election, the Republican candidate won. Would you bet on the outcome of the election using the Viking's game outcome to predict? Explain. 7. Are history, philosophy, mathematics, and English literature sciences? Give the basis for your answer. 8. Construct a PERT chart for your research. Discuss the chart with students or your supervisor. Are your timetables realistic? 9. Some students seek a richer diet of readings in the philosophy of science than possible within the scope of this book. For background in the epistemology (study of origins, nature, and limits) of social science data, see Bonnen. Blaug and Popper provide seminal work on the current philosophy of economic science. Muller and Nelson provide excellent contemporary insights into how economics relates to the major value systems of moral philosophy and theology. These works are given in the reference list for the chapter.
This page intentionally left blank
2
Guidelines for Conducting Scientific Research The scientific method is a way of thinking about problems and solving them. It helps to explain the problem and its alternative solutions in an orderly way. It focuses research. It provides common ground for inquiry by those attempting to understand reality. No rigid formula can be used for all scientific discovery or scientific writing. However, two basic assumptions underlie scientific investigations: (1) events taking place in nature can be understood if one asks the right questions and does the right analysis, and (2) a systematic relationship holds between a cause and its effects, although, as noted earlier, random elements that give rise to error in predictions also may be present. The scientific method has at least six major checkpoints: (1) stating the problem; (2) formulating the hypothesis or, alternatively, stating the objectives; (3) developing the method of analysis; (4) data collection and analysis; (5) interpreting results; and (6) drawing conclusions. The research process is not complete without a systematic process of reporting and disseminating research results. Although research methods and scientific writing techniques vary from one person, discipline, or problem to another, some basic principles and mental techniques are commonly used in scientific investigation and communication of results in theses, journal articles, or other forms (Beveridge, 1957). The basic guidelines for conducting scientific investigation are discussed in detail in the following sections. A sample problem statement with objectives, hypotheses, literature review, and procedures for a scientific research study is illustrated in the text.
36
Research Methods and Communication in the Social Sciences
PROBLEM STATEMENT Scientific inquiry originates from felt needs as apparent in a problem in need of a solution or a question requiring an answer. A problem may be defined as a situation of confusion, doubt, or uncertainty. It may be defined as a puzzle or gap between the existing situation and a norm or desired situation. It must be possible to formulate a problem in such a way that observation or experimentation in the natural world can provide an answer. For a problem to be a useful basis for inquiry, the means for finding answers, empirically if possible, should be available. As noted earlier, value judgments must be handled carefully in science. When selecting a worthwhile problem for analysis, particular concern centers on whether economists are substituting their own values for the ends or goals of society. If economists are to address worthwhile issues, they must gather them from somewhere. Newspapers and magazines can be as useful as professional journal articles in deciding what puzzles to address. In originating and writing the problem statement, several points are useful. Criteria helpful in defining and evaluating the significance of a research problem are listed below (see Ary, Jacobs, and Razavieh, 1979; Tuckman, 1978; Andrew and Hildebrand, 1982; Travers, 1978). 1.
2.
A problem potentially is worthwhile. The problem should be based on and appeal to felt needs of individual or group clientele and societies. The expected benefits from resolving the problem should be greater than the expected research costs, although the expected benefit-cost ratio is likely to be a crude judgment. The problem may be one whose solution will make a significant contribution to the body of organized knowledge or extend the boundaries of scientific research. Problems include gaps in existing knowledge, inconsistencies in previous pertinent research work, or perceived needs by people having to make decisions. William Proxmire, when he was a United States Senator from Wisconsin, used to give a Golden Fleece award for what he perceived to be trivial, worthless research. Ask yourself the "so what" question-so what is the potential value of your research to society? Who will benefit? How will decisions be improved? Could you defend your work before Senator Proxmire or a less populist critic? Ideally, a research problem should be economically and technically feasible, politically acceptable, and socially desirable.
Guidelines for Conducting Scientific Research
3.
4.
5.
6.
That statement applies to all phases of research: the process itself, findings, and recommendations. Research that submits human subjects to extreme danger or that calls for euthanasia to raise per capita income might make narrow economic sense but is not socially desirable or politically feasible. The world is replete with scientific analysis, diagnoses, and policy recommendations possessing narrow social desirability but not economic or technical feasibility. A suitable problem is researchable. The problem and the research to resolve the problem must be relevant and manageable within the resource restrictions. To be researchable, a problem must deal with variables that can be defined and measured. The most basic resource restriction is likely to be research funds. The topic needs to be narrowed to proportions manageable with the data and other research resources at hand. The researcher should not conduct a project without first precisely defining the problem and reducing the proposed project to manageable proportions within the time requirement and within the human, financial, and other resources available for the project. Many important questions in economics cannot be answered by empirical research alone but must be investigated through other means, such as moral or philosophic inquiry. For instance, determining the optimal health care system in a society is impossible using empirical research alone. At the least, multidisciplinary analysis is necessary-analysis that includes social and political scientists and practitioners. The causal relationships expressed in a problem statement are nonhypothetical and relevant to the problem. Because the statements of the problem serve to orient the entire research process, problem statements must suggest meaningful, testable hypothetical relationships that, when analyzed, yield relevant and nontrivial results. Research is undertaken specifically for the purpose of obtaining factual information to help resolve clearly stated problems. Research ordinarily should not duplicate what has been done. It should yield new knowledge. In some instances, however, it is useful to replicate or update previous scientific investigation. It is important to know the literature to avoid undesirable duplication. The problem must be suitable for the particular researcher.
37
38
Research Methods and Communication in the Social Sciences a.
b.
Ideally, the researcher should have enthusiasm for addressing the problem. The research should make the best use of the investigator's talents and interests. It may be a problem whose solution is important personally as well as useful to the profession and society. An uninterested and indifferent researcher is unlikely to expend the time and energy required for thorough research. The problem ideally should be suited to the researcher's knowledge and experience. One needs to be familiar with the existing theories, concepts, and established facts to identify a worthwhile problem. One needs to have or be able to acquire the necessary skills and competencies to carry the study through to completion. Collaboration with colleagues can be rewarding.
The written statement defines the problem and tells why it is worthy of research. The statement needs to contain some brief "so what" sentences of a kind you would use to convince a lay listener that your research is worthwhile. The statement also may indicate who will benefit from the research, define the scope (time period, region, etc.) and limiting conditions of the study, and note general successes and failures in previous attempts to address the problem. A detailed definition may result in a better understanding of the problem. Generally, the problem statement does not include the review of literature, objectives, or other parts of the inquiry.
A Sample Problem Statement A sample of a problem statement from a project entitled "Modeling the Interaction between Farm Structure and Community Vitality" (Henderson, pp. 6,7) is presented below: Most of the towns, small cities, and farms in the 500 United States' counties having agriculture as their economic base share a common feature: depopulation. The loss of people from farms, towns, and small cities raises per capita costs of providing basic services such as schools and infrastructure such as roads. Some communities have responded by raising taxes to maintain services and infrastructure; other communities have allowed them to deteriorate. Neither strategy has been successful in stopping the loss of population. Continuation of the current downward spiral of population loss and rising
Guidelines for Conducting Scientific Research
39
service costs could cause near abandonment of large areas of the nation, especially in the Great Plains. An important part of the nation's agrarian heritage-family farms and rural communities-is being lost. [Note: The above paragraph identifies a problem.] The failure of past public policies to stem rural decline may trace to lack of knowledge of basic relationships. To make wise decisions, those who make rural development policy need to know what factors give vitality to communities. [Note: The above paragraph begins to narrow or delimit the problem to something manageable for research. The paragraph also begins to identify beneficiaries.] Of interest is the impact of farm size and numbers on income and employment of local communities. Do communities surrounded by small farms have higher income and employment, ceteris paribus, than communities surrounded by large industrial-type farms? Do increases in farm income favor large trade centers over nearby small towns as farm families travel farther to seek more variety and better shopping? Also of interest is the impact of communities on farm structure. For example, does a rapidly growing rural community increase the number of small part-time farms and slow exit of operators from farming? [Note: Asking questions helps create interest and focuses the issues.] Rural people have struggled to revitalize their communities and farms. They have sought industry to create jobs and payrolls but have been unsuccessful for lack of agglomeration economies found in larger population centers. They have invested in schooling only to see the brightest and most talented youth leave for better employment in larger cities. They have received billions of dollars of farm commodity program benefits from government only to see continued loss of family farms. These public policies have failed to stem the decline in farms and rural communities. [Note: The above paragraph identifies failures in past efforts.] As local, state, and federal governments reexamine their failed policies for rural development, it is essential also to review the causal forces underlying the socio-economic strength of communities. One such basic relationship is between farm structure and community viability. [Note: This is a transition paragraph between the statement of the problem and the review of literature or objectives.]
REVIEW OF LITERATURE The literature of science is a permanent record of research facilitating communication among scientists. The literature contains historical records
40
Research Methods and Communication in the Social Sciences
of scientific work: new approaches, special emphases, new departures, experiments, knowledge, and hypotheses that have been ignored or have withstood the test of further observation (Barrass, 1978). The review of research informs us of the growth and sophistication of scientific studies within a specific research area (Leedy, 1981). The search for related literature should be nearly completed in the early stages of a research project, but will continue as research progresses. In the ultimate sense, reading the articles in your field becomes insurance against doing less comprehensive research and writing a naive manuscript. The manuscript that lacks insights about previous work stamps the researcher as being uninformed, or worse still, as being either ignorant or indifferent to the work of others. A major reason for reviewing pertinent literature is to build on what others have done. As such, a review brings us to the frontiers of science. But just as important, a review of literature reveals what analysis has not been done. The written review of literature should clearly state how the proposed analysis will extend the frontier of knowledge by doing what has not been done. In summary, the review serves several important functions (Ary, Jacobs, and Razavieh, 1979; Peterson, 1961):
1. It provides the knowledge to define a specific problem along with the scope and frontier of the study. 2. It improves understanding of previous theories, laws, and interpretations that were skillfully or perhaps improperly used in previous research. Through studying related research, investigators learn which procedures or methods of analysis have proven useful and which seem less promising. The successes and failures of past work provide insight for designing one's own study. 3. The search provides a theoretical background to establish a perspective on the problem, to formulate appropriate hypotheses of the study, and to acquire broad knowledge of the research area. 4. A thorough search avoids unintentional replication of previous research studies and indicates how various approaches to similar problems in the past differ from one's own methods and results. 5. The study of related and pertinent literature enables researchers to better interpret the significance of their own results. The review prepares researchers for fitting the findings of their research into the body of organized knowledge.
Guidelines for Conducting Scientific Research
41
Searching and reading are, of course, time consuming. In most cases, the major difficulty will be too much rather than too little literature to survey. The technical literature grows rapidly each year and readers can become swamped with the publications in their specialty. Priorities must be set and reviews confined to the most pertinent contributions. After you have read several recent review articles in your field and have developed a sense of the relative importance of the various published papers, you can read selectively. For these reasons, researchers need to selectively cite sources that illuminate or support the content of the research work. A citation should indicate briefly who did what - the specific contribution. If you are to find the most literature references with the least effort, you must do two things. First, decide on the exact subject of your research and keep the scope within reasonable limits. Second, consult with your librarian about the sources and also about available information retrieval services. A computer terminal that gives you access to an information retrieval system is a powerful tool for searching the necessary literature. The following suggestions can help you to organize the search for literature review: 1. Begin with the most recent studies and work backward through earlier volumes. One advantage of this approach is that recent studies have incorporated the thoughts and findings of previous research. This improves the chances that earlier misunderstandings have been corrected and unprofitable approaches have been avoided. Another advantage is that these studies include references to seminal earlier works. (Your written review may follow an order opposite that of your search.) 2. Read the abstract or summary section of a research report first to determine whether it is relevant to your research. Doing so can save much time. 3. Before taking notes, skim the research report quickly to find those sections that relate to your research-another way to save reading time. 4. Make notes directly on file cards. They are easier to sort and organize than sheets of paper, backs of envelopes, and unorganized notebooks. Typing notes on a computer facilitates organization of the literature review. 5. Write out a complete bibliographic reference for each work. If you know which style manual will be used in the finished report, you can save time by using that reference form while note taking. Add the library call number to facilitate finding the work again, should it be necessary.
42
Research Methods and Communication in the Social Sciences 6. To facilitate sorting and organizing, put no more than one reference on each card. Including only one contribution per card facilitates later sorting into related themes. 7. Indicate which parts of the notes are direct quotations from the author and which are your own paraphrases. Failure to make this distinction can lead to inadvertent plagiarism. It is also wise to clearly separate an author's evaluation of his research from your own ideas and research results. 8. Do not include the entire literature review but only the relevant citations in the main body of the study. Use appendixes or annexes for technical procedures or additional materials. 9. Avoid perfunctory citations. Cite a work only if it is important to your research and is the source of the material you use. Some journal article authors seem to cite numerous other authors for the purpose of winning favor with possible reviewers. 10. Avoid "grandfather" citations not taken from the original source but from someone else's interpretation of that source. Go back to the original source to make sure your quote or interpretation is correct.
The completed, written review of literature should be tightly relevant, focused, specific about contribution of others, and precise in noting the remaining gaps that the current research will address. It is important to note the shortcomings and successes of previous research. Don't cite references you haven't read. A sample of the preliminary literature review of the study "Modeling the Interaction between Farm Structure and Community Vitality" (Henderson, p. 4; also see earlier problem statement) is illustrated as follows: Goldschmidt's (1946) "Small Business and Community" report to Congress implied that rural communities surrounded by larger, higher income farms would have a less viable nonfarm economy than similar communities surrounded by smaller, lower income farms. Goldschmidt estimated that a rural town surrounded by relatively smaller farms received $68 more retail trade per $100 of farm output than did a similar rural town surrounded by relatively larger farms. Numerous other national and regional studies have supported the Goldschmidt finding of a decrease in retail trade in communities surrounded by relatively larger farms (Heady and Sonka, 1974; Fujimoto, 1977; Marousek, 1979; Johansen and Fuguitt, 1984; Tweeten, 1984; Skees and Swanson, 1986; Lobao, 1990). (Note: The review begins with a seminal study and follow-ups.]
Guidelines for Conducting Scientific Research
43
Academic studies supporting the Goldschmidt finding (of a specific decrease in retail trade or a general decline in the nonfarm economy in communities surrounded by larger size farms) suffer several limitations. Most fail to adequately control for regional or local differences in the structure of the nonfarm economy and population. Most use small samples or county rather than community data. Serious methodological problems have attended the use of cross-sectional procedures and spatially aggregated units of observation (Hayes and Olmstead, 1984; Henderson, Tweeten, and Woods, 1992; Lobao, 1990; Wimberly, 1987). Spatial aggregation is a common limitation underlying the previous attempts to directly test the Goldschmidt hypothesis. Theoretically, county level estimates represent an average for all communities within a county and are less accurate than are community estimates because impacts vary by community size (Shaffer, 1989). Empirically, models using county data have been shown to underestimate the impact of a change in farm income for larger communities and overestimate the impact of a change in a farm income on smaller communities (Henderson, Tweeten, and Woods, 1992). [Note: The review has identified shortcomings of previous studies.] Regression models will address the methodological problems of the previous studies. The proposed methodology will reconcile the spatial aggregation problem by employing community level observations (Henderson, Tweeten, and Schreiner, 1989). The proposed application will resolve the cross-sectional limitation by combining time series with cross-sectional data in multivariate statistical analysis (Henderson, Tweeten, and Schreiner, 1989). [Note: The review identifies methodology by Henderson to correct shortcomings of previous studies.]
OBJECTIVES The objectives state what the research intends to accomplish. Specific objectives are generated from the problem statement and may be implicit in the hypotheses. The objectives link the theoretical relationships to the analytical and methodological orientation necessary for conducting the research. Each specific objective should be directly related to one hypothesis and help to clarify the direction of the research. Objectives of a research project suggest what information will be obtained to help resolve the problem that initiated the research. This information, in turn, can be used to establish an action program or plan to alleviate the problem. Objectives limit the scope of the research work and may clarify the means of performing the research but avoid detailed listing of procedures.
44
Research Methods and Communication in the Social Sciences
Objectives may identify the clientele(s) and describe the nature of the potential or/and expected product of the research for the client(s). It is necessary that the objectives be sufficiently broad to satisfy the needs of the clientele(s) but also sufficiently specific to conform to budgetary restrictions. Some researchers prefer to list one general objective and several specific objectives. The summary written upon completion of the research responds to each objective. If an initial objective was not addressed, it should not be removed from the list of objectives. Rather, an explanation of why it was not addressed should be given (Andrew and Hildebrand, 1982). Reporting objectives with unsuccessful results helps other researchers to avoid dead ends in their research. Selected objectives of the study "Modeling the Interaction between Farm Structure and Community Vitality" (Henderson, pp. 7,8; also see earlier problem statement and literature review) are illustrated as follows: 1. To develop and apply a cross-sectional time series regression model to test hypotheses regarding effects of changes in farm structure on the vitality of rural communities. 2. To develop and apply a cross-sectional time series regression model to test hypotheses regarding effects of community vitality on the local structure of farming. 3. To develop and apply a general model accounting for the interaction between changes in local farm and nonfarm structure.
HYPOTHESIS FORMULATION Formulating and testing hypotheses is the heart of the scientific method (Tweeten, 1983). The hypothesis is a powerful tool in scientific inquiry. Existing information is used to form hypotheses that explain, unify, or organize known facts. The hypothesis is a tentative statement or proposition regarding the diagnosis or solution of the problem. The hypothesis directs inquiry by drawing on principles or theories learned from accumulated past experiences. A mechanic faced with an auto engine that refuses to start does not need to waste time checking the paint or the tires. He forms hypotheses that the gas tank is empty or the battery is dead. Similarly, a scientist uses hypotheses to make inquiry more efficient and effective. Hypotheses also open new areas of research that can be tested by experiments or other means.
Guidelines for Conducting Scientific Resea
45
The following purposes are served by the hypothesis (Ary, Jacobs, and Razavieh, 1979; Andrew and Hildebrand, 1982; Borg, 1981): 1. Hypotheses should be stated as briefly and clearly as possible in terms of both theoretical complexities and implications. 2. The hypothesis provides a tentative explanation of phenomena and facilitates the extension of knowledge. It should be consistent with the existing body of knowledge. 3. The hypothesis provides direction to the research effort. The hypothesis determines the nature and type of data needed to test the proposition relevant to the solution of the problem. The hypothesis, when well formulated, will suggest the appropriate data and analytical techniques for testing that should be employed in the research process. Thus, a set of hypotheses can be thought of as a plan of action that guides the researcher in selecting the type of data and techniques necessary for analysis. This implies that hypotheses are formulated before the data collection activity of the research project has started. The hypotheses, then, link the problem to the data collection and analytical stages of research. 4. The hypothesis provides the researcher with a rational statement that is directly testable in a research study within the limits of the research resources. Information is collected and analyzed. A testable hypothesis is verifiable in such a way that empirical observation may support or not support the hypothesis. Where possible, it is desirable to choose hypotheses that are quantifiable and testable by statistical inference. However, many hypotheses concerning important problems are not quantifiable or testable by statistical inference. They must be tested by analysts' judgments based on the weight of evidence. 5. The hypothesis provides a framework for interpreting the empirical results, verifying the expected relationships between the measures used in the research, and reporting the conclusions of the research. The researcher will find it convenient to test each hypothesis separately and state conclusions relevant to it. Three types of hypotheses are widely employed in research. These are reported by Tweeten (1983) from class notes of Emeritus Professor John Timmons of Iowa State University.
46
Research Methods and Communication in the Social Sciences 1. Conceptual Hypothesis-a. premise or working hypothesis defining, delimiting, and analyzing the problem. Examples are (a) the expected value of the research is greater than the expected cost, (b) the research problem and analysis can be confined to the manageable subsystem under study of the total interdependent market universe, or (c) the problem is capable of being researched and hypotheses tested with available data, techniques, and theory. Such subsumed working hypotheses are usually not listed or tested; indeed, most are not testable in a formal way. But the study rests on such premises. 2. Diagnostic Hypothesis-2L tentative proposition about the cause of the problem. For example, (a) poverty is caused by inadequate inherent human resources, (b) poverty is caused by failure to develop human resources, or (c) poverty is caused by lack of jobs. Diagnostic hypotheses may or may not be quantifiable. The analyst needs to carefully formulate them so they can be tested. Such testable hypotheses are the core of most research. 3. Remedial Hypothesis-potential solution to the problem being analyzed. Such hypotheses may follow logically from diagnostic hypotheses and can be stated as conditional if-then propositions. Such propositions state that if a. given situation prevails, then certain consequences follow. Remedial hypotheses suggest one or more solutions to the problem. An example is: If poverty results from failure to develop human resources, then the most cost-effective means to alleviate it is with general schooling, apprenticeship, and vocational-technical training programs. Remedial hypotheses are often not quantifiable and not testable by statistical inference. However, they may be tested less formally by weight of evidence. In recent decades, large social experiments such as the negative income tax (income maintenance) scheme have tested remedial hypotheses previously considered too costly to test.
When using judgment to test hypotheses not amenable to statistical inference, the analyst needs to heed some advice. Before reaching a conclusion based on the weight of evidence, as much evidence as feasible needs to be assembled to make a compelling case for and against a hypothesis or theory. Assembling complete evidence avoids bias. Scientists often express hypotheses to be refuted because they find it easier to refute than to verify a proposition.
Guidelines for Conducting Scientific Research
47
Hypotheses may be primitive or advanced. Primitive hypotheses are used when little is known about a parameter. For example, in early research we may hypothesize that the elasticity of demand for beef is zero with respect to price. We reject or fail to reject this primitive null hypothesis based on a statistical test using regression analysis. After studies have rejected the null hypothesis that the elasticity is zero, we are ready for advanced hypotheses testing. The null hypothesis then is the best previous estimate or average of estimates of the elasticity (Tweeten, 1983). In advanced hypothesis testing, we wish to determine the value of the parameter as precisely as possible to use in forecasting or other analysis. Analysts using zero as the null hypothesis will not reject it when samples are small (as is often the case) even when the hypothesis is false. Maintaining a working hypothesis of zero effect then means biased science and predictions. This result is avoided by using the best previous parameter estimates as the null. Three common mistakes made by students are failing to formulate hypotheses that are (1) nontrivial, (2) testable, and (3) capable of refutation. For example, the hypothesis that "the linear programming model constructed herein will show how to raise farm income" will be true by construction and hence will be trivial. The hypothesis that "the optimal farm plan from the linear programming solution will raise net income of the representative farm" probably will not be tested and can be omitted. Just because a computer printout shows that the set of enterprises in our mathematical model raises farm income does not ensure that our findings will work or be used by actual farmers. Ordinarily we can't afford to properly follow up to see if research recommendations are implemented. However, truly comprehensive problemsolving research evaluates whether research results are implemented and problems solved. Some analysts treat a hypothesis as a proposition to be proven at all costs. Will Rogers quipped, "The trouble is not so much what folks don't know as it is what they know that ain't so." Heavy commitment to preconceived notions (called priors) biases research. We should be prepared to give at least as much effort to refuting hypotheses as to supporting them. A sample of hypotheses of the study "Modeling the Interaction between Farm Structure and Community Vitality" (Henderson, p. 9; also see earlier problem statement, literature review, and objectives) is illustrated below: 1. Income per farm family from all sources has a greater effect than size of farm on local community economic vitality. (The statistical test is between standardized regression coefficients on farm income and size variables.)
48
Research Methods and Communication in the Social Sciences 2. 3.
Entry into farming increases as off-farm employment of farmers increases. Communities surrounded by small farms have higher income per capita than do communities surrounded by large farms. 4. Communities surrounded by small farms have more social organizations per capita than do communities surrounded by large farms. [Note: These hypotheses will be tested by statistical inference from equations presented later in the procedure section.]
METHOD OF ANALYSIS The scientific approach rests on two basic assumptions: (1) that useful determinations of the nature of reality can be derived from observation, and (2) that phenomena conform to laws or systematic relationships. Laws are principles that make sense and are valid as measured by repeated successful tests of prediction. Logical positivism holds that we need not know the true causal relationship but we can learn enough from observed behavior to provide clear, logical, and reliable science.1 The materials and procedures used in conducting research are selected according to the objectives and hypotheses and must conform to scientific practice. Models are abstractions from the detail of reality designed to provide explanation and prediction. The real world is far too complicated to model in all its complexity. To make the model workable, only essential relationships are retained. A model is a set of interrelated constructs containing variables useful for explaining or predicting phenomena (see note 1). The framework may be an informal mental, nonquantitative set of verbal statements and relationships. On the other hand, the model may be formal: data, theory, and assumptions in a systematic framework or algorithm to explain or predict the phenomenon being analyzed. Linear programming is an example of a quantitative algorithm used for optimization. Quantitative econometrics, computer simulation, and computable general equilibrium models are often used to analyze and predict economic outcomes. No prediction is perfectly accurate. Measurement, specification, and sampling errors limit the accuracy of prediction. Specification error arises when we do not choose the correct model, including the form and content of the variables. Sampling error arises when we make population inferences from only a subset of the entire population. Measurement error arises when we do not measure precisely even that subset of the population included in the sample (Travers, 1978).
Guidelines for Conducting Scientific Research
49
Predictions based on numerate observations are called quantitative predictions. Predictions based on observations that do not contain numbers are called qualitative predictions. A farmer uses a mental model of his land, his assets, market prospects, and past weather conditions to decide which crops to plant each year. A surveyor constructs a physical model-a map-to help in planning a road. An economist uses mathematical models to understand and predict the flow of international trade. A natural science researcher may use an experiment or laboratory work to test a hypothesis. Scientific models are formulated and experiments are designed for educational purposes, to simulate natural or social phenomena, or to improve technology for further experimentation or scientific research. Decision makers at every level use informal mental models to choose among policies. These mental models are, of necessity, very simple when compared with the reality from which they are abstracted. The human brain, remarkable as it is, can only keep track of a few complicated, simultaneous interactions that depict the real world. The construction of a model is a creative act that admits to no standard procedure, but the validation of the model follows a regular process similar to the scientific method. Models are tested and subject to revision. Each new model uses feedback from the old model, retaining the successful parts of older models. A model may involve a highly complex analysis using large and sophisticated computers. Large models are costly to build and maintain. Models need to be kept as simple as possible for clarity and for more mundane reasons, such as the ability to keep them operating on a tight budget. Experience and knowledge of data, concepts, and institutions are important factors in the formulation and specification of valid models. No one procedure is relevant to all research. The procedure is as varied as the imagination of the researcher, and depends partly on resource limitations (Travers, 1978; Tuckman, 1978). A sample of procedures for the project "Modeling the Interaction between Farm Structure and Community Vitality" (problem statement, objectives, literature review, and hypotheses stated previously) is presented next. The reader need not understand the mathematics to catch the basic approach being illustrated. The procedure to meet Objectives 1 and 2 will employ partial equilibrium regression models. The equations will include dummy interaction terms to test for differences in the coefficients by community size across regions. The following regression equations will test the Goldschmidt hypothesis: where NFYj and NFE; are real income and employment in the nonfarm economy by community i, a (intercept term) and p are the coefficients to be
50
Research Methods and Communication in the Social Sciences
where NFYj and NFEj are real income and employment in the nonfarm economy by community i, a (intercept term) and P are the coefficients to be estimated, FYy is the real income of farm j in community i, ASy is acres or sales of farm j in community i, POPs is the nonfarm population of community i, D (0 or 1) is a dummy variable used to stratify the communities (m = different community sizes), t is year, and e is an error term. The procedure for Objective 2 will employ regression analysis to test the hypothesis that changes in community structure affect farm structure. Increases in community employment and income (nonfarm economic development) are hypothesized to affect farm structure by increasing the availability of off-farm employment of farm people. Increases in off-farm employment are hypothesized to increase part-time farming, increase entry into farming, and expand the number of smaller farms around communities. Conversely, decreases in off-farm employment increase exits from farming and numbers of larger farms around communities. The equations in the model are:
where OFE is number of days worked off the farm, PO is principal occupation of the operator, FN is the total number of farm operators, EIF is the total number of new farm operators who entered farming, EOF is the total number of farm operators who have exited farming, a (intercept term) and P are the coefficients to be estimated, NFY is the real nonfarm income,
Guidelines for Conducting Scientific Research
51
NFE is nonfarm employment, POP is the population of the community, D (0 or 1) is a dummy variable used to stratify the communities (m = different community sizes), i is community, j is farm, t is year, and e is an error term. The procedure for Objective 3 will be to combine equations (1) to (7) in a simultaneous system. Thus the system of equations will include a set for farm structure as well as a set for nonfarm sector structure. The simultaneous equation system estimated by three-stage least squares recognizes that the community and farming economies interact, so that causality is joint and two-way rather than running in one direction.
COLLECTING AND ASSEMBLING DATA Information is the foundation upon which research is based. Information in general and data specifically are as critical to the problem identification phase of the study as they are to analysis. Their availability profoundly affects both the quantity and quality of research which can be produced within a given period of time (Andrew and Hildebrand, 1982). Based on the problem statement, hypotheses, objectives, and method of analysis discussed above, the research work is built around empirical information within a conceptual framework. Information may be quantitative or qualitative. The less factual information available, the more difficult the research is likely to be. Scientific data are meaningless unless organized into patterns that permit the researcher to discern relationships or trends that will lead to prediction and/or tests of hypotheses regarding the cause of or solution to the problem. The researcher needs to compile data into suitable graphic composites or develop table(s) grouping the data into essential elements. Some analysts call the raw data "statistics" only after they have been collected in accordance with an appropriate conceptual model and standards, and then processed in accordance with principles specified in the discipline called statistics. The following sections describe principal types and sources of data. Primary or field data often are gathered directly from original units of observation by the researcher and are processed for analysis. The analyst ordinarily will subject the data to statistical tests of reliability such as sampling variation. Primary data give the researcher a close-up view of what has happened in a specific population perhaps not enumerated by agencies such as the Bureau of the Census. Secondary data are from published or other sources not generated or accumulated specifically for the study but borrowed for the particular project by the researcher.
52
Research Methods and Communication in the Social Sciences
Researchers like primary data gathered specifically to meet the needs of their study. Such data usually are too expensive, however. Hence analysts utilize secondary data where feasible. Sometimes primary published information provides the basis for problem formulation, while primary data generated in research is the source of information for analysis. Another classification is time series data and cross-sectional data. Time series data or observations are obtained at specific intervals or in a series over a period of time. Cross-section data are from one point in time (or from a relatively short period of time) and usually include observations of several different strata of a population or geographic area (Andrew and Hildebrand, 1982). Experimental data are another source of information for use in research. These are data generated from laboratories or field tests under controlled conditions. Experimentation has been the principal basis for obtaining information in the physical sciences and will continue to remain of paramount importance. An advantage of experimental data over survey or nonexperimental data is the ability to isolate the effect of one variable on another by controlling for all other variables. Proper experimental design (the form in which the experiment is set up) allows the researcher to select the factors that will vary, the levels at which they will be included in the experiment, the measurement of phenomena resulting from the experiment, and the pattern in which they will be used. By using appropriate equipment, the researcher can also usually obtain quite accurate measures of the input variables as well as of the results of the experiment (Andrew and Hildebrand, 1982). Survey data are widely used in the social sciences and are nonexperimental data. The researcher relies upon personal interview, mail, or telephone to obtain primary data and to communicate measurement. The researcher cannot determine the design and level of the factors nor directly measure the phenomena, but controls only the technique used in measurement. With survey or nonexperimental data, the levels and combinations of factors or variables are predetermined by nature or society, so the researcher must measure and use them as they exist. Because these variables are very difficult to identify and measure, survey data are usually more subject to inaccuracies than are experimental data (Andrew and Hildebrand, 1982). Survey data are frequently the only feasible alternative in the social sciences. Factors that cannot be controlled in data generation are "controlled" in analysis by multivariate statistical techniques such as multiple regression analysis.
Guidelines for Conducting Scientific Research
53
The discipline of statistics is useful in making inferences from the part to the whole. A questionnaire is administered usually to a subgroup of the larger group of interest. The small group that is observed is called a sample, and the larger group about which the generalization is made is called a population. A population is defined as an entire group or all members of any well-defined class of people, events, things, or objects having at least one trait in common, a totality consisting of all items that would be surveyed in a complete enumeration. A finite population contains an exact upper limit. Sampling is selecting a small number of units from among the population in such a manner that they can be used to make estimates (inferences) about the whole group. Defining the population upon which the survey is to focus, the researcher must make sampling decisions early in the overall planning of a survey. It is usually not practical to obtain measures from an entire population due to the factors of expense, time, and accessibility. Information is collected from a smaller group or subset of the population in such a way that the knowledge gained is representative of the total population under study (Cohen and Manion, 1985). Sampling is necessary when the researcher lacks the time and money to study all possible members of a population. The soundness of the sampling plan will determine how reliably the sample can be generalized. Appropriate procedures must be followed if researchers are to make valid inferences for a population based on data collected from a portion of the population. Sampling entails several steps (Ary, Jacobs, and Razavieh, 1979; Hayes, 1965). 1. Having determined the kind of information needed for the study based on the problem statement, hypotheses, and objectives, the first step in sampling is the identification of the population to be represented in the study. Make a map of the area or location of the study and understand the socioeconomic characteristics of the population based on census and other data. 2. One must identify the accessible population defined as that portion of the population to which one has access. It is from this group that the researcher will take the sample for the study. 3. From the accessible population, one selects a sample of proper size in such a way that it is representative of the characteristics of the accessible population considered in the study. Past experiences are helpful in selecting the sample and devising the sample size.
54
Research Methods and Communication in the Social Sciences
Random sampling is often the best approach. Random sampling means selecting a sample in such a way that each item or member of the population under study has an equal chance of being selected for the sample. It is based on the assumption that, while individual events cannot be predicted with accuracy, aggregate events can. For example, it is difficult to predict an individual academic achievement, but it is possible to predict the average academic performance of a group (Borg, 1981; Best and Kahn, 1986). Thus, in random sampling, because of chance, the sample is expected to contain subjects with characteristics similar to those of the population as a whole. Working with a list of all members of the accessible population (sampling frame), the required number of subjects is chosen at random. Assuming no bias in measurement, the larger the sample, the greater the probability that the sample chosen at random will display the characteristics of the population and in the same proportions (Cohen and Manion, 1985). Random samples yield research data generalizable to a large population within margins of error that can be determined statistically (Borg, 1981). Stratified sampling is often used when information is desired on each of a number of subgroups or strata that may differ in the characteristics being studied. Stratified sampling guarantees representation of defined groups in the population. Usually each stratum is sampled proportional to the population represented by it. This procedure is known as proportional stratified sampling. A big advantage is that results are self-weighted. However, characteristics of the entire population, together with the purposes of the study, must be carefully considered before a stratified sample is chosen (Borg, 1981; Best and Kahn, 1986). Analysts especially interested in, say, a minority subgroup may wish to oversample by taking a disproportionately larger share of the sample from that subgroup. This improves the reliability of inferences made for that subgroup. Inferences are made for the population by weighting the oversampled stratum so as to restore proportional sampling. Systematic sampling, like simple random sampling, is used to obtain a sample from the defined population. If the sampling rate is 10 percent, every tenth name on the list is sampled. Systematic sampling is a little more convenient than simple random sampling. It differs from simple random sampling in that each member of the population is not chosen independently. Once the first member has been selected, all the other members of the random sample are automatically determined. Thus only one number needs to be selected at random. Systematic sampling can be used instead of simple random sampling if one is certain that the population list is in random order (Borg, 1981; Best and Kahn, 1986).
Guidelines for Conducting Scientific Research
55
In cluster sampling the unit of sampling is not the individual but rather a naturally occurring group of individuals. An example is to circle clusters, say, of six contiguous houses in the population from a map, number each cluster, and select clusters for sampling from a table of random numbers. Cluster sampling is convenient when the population is widely scattered geographically, because personal interviews are cheaper and driving is less when households are taken in clusters. Cluster sampling also is used when it is more feasible or convenient to select groups of individuals than it is to select individuals from a defined population. This situation occurs when it is impractical or impossible to obtain a list of all members of the accessible population and when it is difficult to break up established groups in order to collect data from a few selected group members. This method of sampling may introduce sample bias because some clusters overrepresented in the sample do not represent population characteristics. This method is recommended only when a simple random sample is impracticable (Borg, 1981; Best and Kahn, 1986). Sample size influences the reliability of statistics. There is, of course, no clear-cut fixed number or percentage of subjects that determines the correct or adequate sample size. The size of the sample depends upon the purpose of the study, the nature of the population under scrutiny, the precision the researcher desires in estimating the population parameter, and the available budget. There is usually a trade-off between the statistical desirability of a large sample and the economic necessity of a small one. The ideal sample is small enough to afford but large enough to statistically detect meaningful relationships in the population about which the researcher wishes to generalize (Best and Kahn, 1986). The best approach is to work with a statistician who can estimate the size of the sample required to detect meaningful traits or differences in the population. The important factor for deciding on size of the sample is the variance of the population. If variance is zero, a sample of one will do. Thus prescribing sample size as a certain percentage or proportion of the population is not a realistic procedure. Some use the crude "rule of thumb" that 10 percent of the accessible population be included in the sample. In principle, we select the sample giving the smaller variance for a given cost, or the smaller cost for a prescribed variance. In practice, the sample is usually set as large as possible subject to budget constraints. A sample is too large when relationships or differences found from the sample are statistically significant but otherwise of no consequence. A sample is too small if, when sampling from a
56
Research Methods and Communication in the Social Sciences
population where differences are of consequence, large and meaningful sample differences are repeatedly insignificant (Tuckman, 1978). A common error in selecting samples is to take refuge in sheer size of the sample, as though sample size by itself assures representativeness. If the sample is biased (not representative to begin with), size will not ensure a reliable estimate of the population parameter. A sample of 1,000, improperly selected, can be less representative and less reliable than one of 100,properly selected. For example, an attempt to infer political preferences of the nation from a survey of cellular telephone users will be biased toward the wealthy. In short, proper sample size for any study depends on (1) desired level of statistical accuracy; (2) costs in terms of time, labor, material, and other resources; (3) variance of the population; and (4) contemplated intensity of analysis (Travers, 1978). If age, education, farming status, location, income, price, and other selected factors are presumed to be important to our analysis, then the sample must be adequate to assure a strong likelihood that most of these characteristics will be well enough represented to permit a reasonable amount of subsorting on each of them. One way of collecting primary data is simply to ask questions. Personal or group interviews, mailed questionnaires, case studies, and telephone surveys utilize this approach. Questions may be carefully structured and quantified or they may be open-ended. Structured as opposed to open-ended questions are generally most precise, objective, and suited for analysis with computers. The personal interview is a well-established method of data collection which, despite cost in time and other resources, is still widely used. The personal interview enables many complex questions to be asked and is flexible. The response rate is high. Because of high costs of personal face-to-face interviews, many analysts are gathering information from questionnaires presented to subjects by mail or telephone. The lower cost per unit of a mail survey may permit a larger sample, but the response rate may be low, leaving the chance of sample bias. Those not enumerated may be from a different population than those who responded. Modern sampling procedures including aggressive follow-up have resulted in high response rates for mailed surveys (see Dillman [1978] for elaboration). Mail or telephone surveys ordinarily must be shorter than personal interview surveys to elicit high response rates. The following are practical guidelines for constructing a survey questionnaire:
Guidelines for Conducting Scientific Research 1. Carefully construct the instrument. Numerous revisions may be necessary to include all essential items and to eliminate ambiguous or unnecessary items. 2. Keep the questionnaire as brief as possible so that it requires a minimum of the respondents' time. The questionnaire should solicit only those items needed for the study. 3. Questionnaires need careful wording. Phrase questionnaire items so that they can be understood by every respondent. The questions should be clear without elaboration by the interviewer, whose extemporaneous clarifications could bias the response from the respondent. 4. Do not use "loaded" questions or emotional terms that elicit ambiguous or biased answers. Phrase questionnaire items to avoid prejudice that might predetermine a respondent's answer. 5. Avoid leading or abstract "if" questions whenever possible. The question "Would you attend the symphony if it were located nearby?" invites a "yes" response. 6. Avoid complex questions. Long questions with lots of phrases confuse respondents. Know something about the capabilities of your respondents: most are unlikely to be brain surgeons. 7. Avoid open-ended questions. Multiple-choice answers should be exhaustive; that is, all the most likely alternatives on the issues should be expressed. A place to respond "other" will complete the set. 8. Avoid irritating questions that might elicit reactions of embarrassment, suspicion, or hostility in the respondent. Also, avoid questions that use negatives. 9. Ranges may be used instead of direct questions for some sensitive answers. For instance, it is better to ask in what net worth range the respondent falls (list the various ranges for the respondent to check in a multiple-choice answer) than to ask for exact net worth. Respondents may be less offended if asked to specify their age range or their year of birth rather than their actual age. 10. Arrange questions incorrect psychological order. If both general and specific questions are asked on a topic, ask the general questions first. Put the interesting questions that invite participation first. Put the most sensitive questions last. 11. Arrange questions in such a way that responses can be tabulated and interpreted readily. The best questionnaires provide data easily entered in the computer, where it can be edited and
57
58
Research Methods and Communication in the Social Sciences
12. 13.
14. 15.
16.
analyzed. In a telephone survey, the interviewer can directly insert responses in the computer. The computer can be programmed to branch to the proper next question based on the response to the last question. Look at an old questionnaire when preparing a new one. Learn from what has been successful in the past. Let your colleagues review your questionnaire and incorporate their comments and suggestions. Experienced scientists and statisticians who have worked with surveys can help you avoid mistakes. After designing the questions, conduct a pilot test with a few people to detect deficiencies in the questionnaire. In collecting the data, it is necessary to win the cooperation and confidence of respondents. A close identification with the population in question is very helpful. A letter of introduction and endorsement by a respected official or role model and news releases explaining the survey build confidence and trust. Give the respondent a good reason to complete the questionnaire. Send several follow-up letters and cards and make telephone calls to encourage mail response. Mailing another questionnaire also helps. Be courteous and offer to share findings with respondents. Don't send a follow-up to those who have responded.
The U.S. Bureau of the Census gets a high response rate by noting that a response is required by law. You are unlikely to enjoy that advantage, so you have to use other approaches. One way is to pay respondents. Sometimes a prize is offered to the lucky respondent. The best approach, however, ordinarily is to appeal to the self-interests and felt needs of the respondent. Once the raw data are collected and carefully edited for errors, the statistics can be compiled for tables, charts, and other uses. Careful labeling, spacing, arrangement of headings, and placing of the tables with respect to the text is desired. Put appropriate definitions, limitations, and sources of the data with the tables.
RESULTS AND DISCUSSIONS The earlier parts of your report (introduction, literature review, objectives, hypotheses, and method of analysis) are designed to tell how you got the
Guidelines for Conducting Scientific Research
59
results; the results (including discussion or interpretations, inferences, and conclusions) are designed to tell what you found and what it means. The real skill of the researcher and the importance of the study come to play after the necessary empirical results have been obtained. The research work rises or falls on the strength of the results. Professions applaud an analyst for "mining" data and criticize "undigested data" and analysis that is "shallow." Results need to be precisely, clearly, and simply expressed to avoid ambiguity, obscurity, and error (Tuckman, 1978;Day, 1988). Results address objectives and report on hypothesis tests-one by one. Drawing inferences means analyzing or interpreting the empirical results, not simply repeating them. It requires analyzing the causes and effects, stating implications for future conditions, recognizing trends and changes, drawing attention to comparisons and contrasts, and summing up differences. General limitations and constraints of the study and suggestions for future research are included. Although the results section is the core part of the research report, it may be the shortest. Many papers are rejected by journal reviewers and editors because of a faulty presentation, although the results of the paper might be both valid and interesting. Too often, the significance of the results is not recognized or is exaggerated. The true meaning of the results may be completely obscured by the interpretation and may be lost in a mass of words and data presented in the discussion. It is important to present only essential information (Day, 1988). Research reports rely upon two main forms of communication to convey results. One is verbal communication in which description and narration explain to the reader what the researcher has done and what results have been obtained. The communication may contain a brief description of the experiment or statistical results without repeating the experimental details or procedures previously provided in methods of analysis. A second form is symbolic communication in which graphics, tabular array, or statistical values convey to the reader the factual data essential to communicate the researcher's message. Results of the research depicted in tables, charts, or graphs, if properly assembled, can suggest by mere visual inspection the interpretation to be placed on them (Tuckman, 1978; Day, 1988; Leedy, 1981). The researcher presents the data on which the conclusions are based. The summary tables, graphs, or charts should be self-explanatory to the extent possible. The discussion highlights only the main facts and figures from the tables and graphs. The discussion interprets the data and notes conclusions. Representative data rather than endlessly repetitive results are presented and
60
Research Methods and Communication in the Social Sciences
analyzed. Behind every table, graph, or chart may reside great masses of raw data in the form of responses that were gathered directly from the population of the study. The raw data need not be published but should be held as necessary for review and scrutiny by other analysts. In short, a useful discussion of results from the main facts and figures does the following (Day, 1988, p. 44): 1. Presents the principles, relationships, and generalizations shown by the results without recapitulating all of the results. 2. Thoroughly explains graphs and tables that need interpretation. If many graphs and tables are of a similar type, only the first needs a detailed explanation. 3. Points out any exceptions or any lack of correlation, and notes unsettled points. 4. Shows how results and interpretations agree or contrast with previously published work. 5. Discusses the theoretical implications of the research as well as any possible practical applications. 6. States conclusions as clearly as possible and summarizes evidence for each conclusion. Briefly mention the unsuccessful experiments, wrong turns, and dead ends. That helps subsequent investigations avoid minefields. Negative results are not very exciting and are difficult to publish. The less relevant findings can be presented in appendixes to the main report (see Tuckman, 1978; Barrass, 1978). In the main text, the writer emphasizes only the pieces of information that are necessary to make the argument logically acceptable and technically complete.
PREPARING THE CONCLUSIONS AND SUMMARY The discussion of the research work should end with conclusions and/or a summary. For several reasons, writing the concluding section requires your full effort. One reason is that a well-written set of conclusions or summary brings your main contributions into focus. Another reason is that your concluding section has a more "captive audience" than the other parts of your manuscript. After the title, this is what most readers will examine before they read the manuscript. Many readers will scan only portions of the main body, but will carefully read the last section. One more reason is that most
Guidelines for Conducting Scientific Research
61
readers will expect you to explain the key results of your research in brief, readable form. If there is nothing in the summary to inspire the interest of the reader, the study may go unread (Peterson, 1961; Day, 1988). The summary is a synopsis that can be written only when the research is otherwise complete. A good summary briefly recapitulates the nature of the problem, objectives, hypotheses, the scope of the research work, methods of analysis, significant empirical findings, and the advances that have been made in understanding, solving,or alleviating the problem studied (Tuckman, 1978). It concentrates heavily on the accomplishments of the research but also may list shortcomings of the research and possible future directions. Ideally, the summary is complete, interesting, and informative. The summary should be as concise as possible, however. It is sometimes called an Executive Summary and placed at the front of the report where it will catch the eye of the potential reader. No table numbers, figure numbers, references, or citations should be included. There should be no information, ideas, or claims other than those in the research report. Statements of conclusions and inferences should be accompanied by an indication of their range of validity. The summary also lists constraints and limitations of the study. The conclusions do not summarize the study but only review the results. Point out what is important and why the results are valid. The conclusions present major findings regarding each objective. You may show the gaps in your work, the limitations of your findings or your data and procedures, and point out directions for future research. At times it may be necessary to include a recommendations section to indicate the policy implications or the application of the research findings to real situations. Writing recommendations is an opportunity to demonstrate your analytical skills, the soundness of your reasoning, and your expertise in deciding on a practical course of action. The character of the final section will depend on the purpose of your report: (1) to analyze a problem, derive a solution from assembled facts, and propose a solution; (2) to select a new procedure or technique and indicate why it is preferable; (3) to identify a need and suggest a way to fulfill it; (4) to propose a new project and show why and how it should be carried out; and/or (5) to explore a new concept and recommend how it should be applied to existing problems.
62
Research Methods and Communication in the Social Sciences
FOOTNOTES Footnotes are not intended to be read by all readers. Footnotes contain information not central to the report but useful to a thorough reader. Footnotes provide additional facts or further explanation of material that appears in the body of the paper. Relevant material not of sufficient importance to be brought into the text of the paper may be put in footnotes. The intended reader may be unfamiliar with some terminology used in the text or research proposal. Terms that are new or that are used in a different sense from the normally accepted one should be explained in the footnotes. Footnotes are used to give a translation of a passage quoted in a foreign language. Footnotes may be used to indicate the source of a specific passage or information from another's work. References in footnotes follow the same general pattern of an entry in a list of references except that the author's name is usually given in normal order, that is, Luther Tweeten, not Tweeten, Luther. Example: Luther Tweeten and George Brinkman. Micropolitan Development. Ames: Iowa State University Press, 1976, pp. 200-215. Footnotes usually are indicated by superscript Arabic numerals. The numerals may begin with "l"on each page, or they may run consecutively throughout the entire paper. The numbers usually refer to footnotes at the bottom of the page. An alternative preferred by some is endnotes listed at the end of the manuscript rather than at the bottom of the page in the text (Samuels, 1978). Failure to reference an idea or a way of expressing an idea that is not your own is plagiarism. Proper credit must be given to the source when you quote or paraphrase another's work. Quoting or otherwise drawing from others adds the weight of their evidence to your own. The originality of a research paper rests in the way you use the ideas and opinions of others to support and explain your own thesis or point of view (Samuels, 1978). Students may be tempted to plagiarize for several reasons: (1) lack of understanding of the ethics governing the publication of research, (2) lack of knowledge of the subtleties of referencing, (3) poor language abilities, or (4) sloth. Each must be resisted, and a first step is to gain confidence in one's ability to do original research. Proper guidance and direction helps students
Guidelines for Conducting Scientific Research
63
to verbalize and communicate effectively, to organize properly, to write logically and in an orderly manner, and to give proper credit when borrowing from others.
BIBLIOGRAPHY OR REFERENCES Numerous formats have been used to reference other studies. One is footnotes, another is numbering references in the text, and a third is to input the last name of the author parenthetically in the text. The latter two methods require all references to be included in a bibliographic summary at the end of the chapter or report. Some researchers always include the publication date with the author's last name (in parenthesis in the text) to convey helpful information on chronology to the reader. References in the text may be noted by a number referring to the footnote/endnote or bibliography where the full citation is listed. The bibliography, literature cited, or references section at the end of the report may duplicate references listed in footnotes but list them in alphabetical order by last name of the author. References are frequently listed only in the bibliography. A growing and useful practice (the one employed in this report) is to insert a partial citation in the text sentence. When the author's name is part of a sentence, a parenthetical reference is unnecessary unless page numbers or a date are needed, for example, "As Smith (1983, p. 10) indicated in his report, farm size and off-farm employment were closely related." If two or more publications are listed for Smith, the year is also included, so the citation includes the author's surname, date of the publication, and, when necessary, page numbers, as in "(Smith, 1980)" or "(Smith, 1980, p. 110)." Advantages of this system are that citations are easily included as the report is written and are easily removed, added, or rearranged. The use of cumbersome bibliographic footnotes is avoided. In scientific papers the general practice is to collect all the references alphabetized by author's last name on a separate page at the end of the paper under such headings as "References" or "Literature Cited." Only cited works are included in "Literature Cited,"but other relevant literature may be included in "References." The list of references is arranged alphabetically by surnames of authors, and also included are the date, publication title, and place of publication and name of publisher (if a book). Some insert the date immediately after the author's name for ease of identification with the text citation. Others include
64
Research Methods and Communication in the Social Sciences
the date at the end of the reference. In the case of articles, the list of references also includes the volume of the periodical containing the article and the page numbers. In the case of pamphlets, some prefer to include the serial number (if there is one). Enough information should be included for the reader to locate the publication and the specific material cited. The Appendix to this chapter contains a sample of references following the style of the American Journal of Agricultural Economics.
NOTE 1. Much debate has centered on structural versus predictive science. If the purpose of a study is to predict the price of wheat, it may be argued that explanatory variables such as quantity of wheat, income, population, and exports need not be structural (i.e., causal, with each predictor justifiable conceptually, and accurately accounting for its separate impact on the price of wheat). If sunspots or today's onion prices accurately predict the price of wheat next year, predictive "science" tells us to use these variables even if we can't explain the causality. We do not resolve this controversy but note that it ordinarily is advantageous to choose variables that are predictive and structural (see Friedman, 1953).
REFERENCES American Journal of Agricultural Economics. (1994). "Preparing a Manuscript for Publication in AJAE." Ames, IA: AAEA Business Office. Andrew, Chris, and Peter Hildebrand. (1982). Planning and Constructing Applied Research. New York: MSS Information Corporation. Ary, Donald, Lucy Cheser Jacobs, and Asghar Razavieh. (1979). Introduction to Research Education. Second edition. New York: Holt, Rhinehart, and Winston. Barrass, Robert. (1978). Scientists Must Write (A Guide to Better Writing for Scientists, Engineers, and Students). New York: Wiley. Best, John W., and James V. Kahn. (1986). Research in Education. Fifth edition. Englewood Cliffs, NJ: Prentice Hall. Beveridge, W. I. B. (1957). The Art of Scientific Investigation. Revised edition. New York: Random House. Borg, Walter R. (1981). Applying Educational Research: A Practical Guide for Teachers. New York: Longman. Cohen, Louis, and Lawrence Manion. (1985). Research Methods in Education. Second edition. Dover, NH: Croom Helm.
Guidelines for Conducting Scientific Research
65
Day, Robert A. (1988). How to Write and Publish a Scientific Paper. Third edition. New York: Oryx Press. Dillman, Don. (1978). Mail and Telephone Surveys. New York: Wiley. Friedman, Milton. (1953). Essays in Positive Economics. Chicago: University of Chicago Press. Fujimoto, Isao. (1977). "The Communities of the San Joaquin Valley: The Relationship between Scale of Farming, Water Use, and Quality of Life." Hearings before the Subcommittee on Family Farms, Rural Development, and Special Studies of the Committee on Agriculture. 95th Congress, 1st Sess.: 480-500. Washington, DC: Government Printing Office. Goldschmidt, Walter. (1946). "Small Business and Community." Report of the Special Committee to Study the Problems of American Small Business. 79th Congress, 2nd Sess., N. 13. Washington, DC: Government Printing Office. Hayes, Michael, and Alan Olmstead. (1984). "Farm Size and Community Quality." American Journal of Agricultural Economics 66(4):430-436. Hayes, Robert. (1965). Principles of Technical Writing. London: Addison-Wesley. Heady, Earl and S. Sonka. (1974). "Farm Size, Rural Community Income, and Consumer Welfare." American Journal of Agricultural Economics 56:534-542. Henderson, David. (1993). "Modeling the Interaction between Farm Structure and Community Vitality." Unpublished research proposal. Piketon, Ohio: Piketon Research and Extension Center. Henderson, David, Luther Tweeten, and Dean Schreiner. (1989). "Community Ties to the Farm." Rural Development Perspectives 5(3):31-35. Henderson, David, Luther Tweeten, and Mike Woods. (1992). "A Multicommunity Approach to Community Impacts: The Case of the Conservation Reserve Program." Journal of the Community Development Society 23(1):88-102. Johansen, H., and G. Fuguitt. (1984). The Changing Rural Village in America: Demographic and Economic Trends Since 1950. Cambridge, MA: Ballinger. Leedy, Paul D. (1981). How to Read Research and Understand It. New York: Macmillan. Lobao, Linda. (1990). Locality and Inequality: Farm and Industry Structure and Socioeconomic Conditions. Albany: State University of New York Press. Marousek, Gerald. (1979). "Farm Size and Rural Communities: Some Economic Relationships." Southern Journal of Agricultural Economics 11:57-61. Peterson, Martin S. (1961). Scientific Thinking and Scientific Writing. New York: Reinhold. Samuels, Marilyn Schauer. (1978). Writing the Research Paper (A Step-by-Step Guide). New York: Amsco College Publications. Shaffer, R. (1989). Community Economics: Structure and Change in Smaller Communities. Ames: Iowa State University Press. Skees, Jerry, and L. Swanson. (1986). "Examining Policy and Emerging Technologies Affecting Farm Structure in the South and the Interaction between Farm Structure and Well-Being of Rural Areas." Technology, Public Policy, and the Changing
66
Research Methods and Communication in the Social Sciences
Structure of American Agriculture 2:373-495. Washington, DC: Office of Technology Assessment. Travers, Robert M. W. (1978). An Introduction to Educational Research. Fourth edition. New York: Macmillan. Tuckman, Bruce W. (1978). Conducting Education Research. Second edition. New York: Harcourt Brace Jovanovich. Tweeten, Luther. (August 1983). "Hypotheses Testing in Economic Science." American Journal of Agricultural Economics 65:548-552. . (1984). Causes and Consequences of Structural Change. NPA Report No. 207. Washington, DC: National Planning Association. Wimberly, Ronald. (1987). "Dimensions of U.S. Agristructure: 1969-1982." Rural Sociology 52(4):445-461.
EXERCISES 1. In successive assignments over several weeks, prepare the first chapter of your proposed thesis, dissertation, or research project. Hand in each section assignment for your instructor's critique. Rewrite the section before proceeding to the next step. The sections, taken in sequence, are: a. Problem statement b. Review of literature c. Objectives and hypotheses d. Procedure • Data • Conceptual framework • Tools and techniques e. If you are far enough along with your research, go on to complete the analysis section and the conclusions section. Because of the importance of practice, it may be well to do exercise (1) first with a problem assigned by the instructor to all students and again with a problem of the student's choosing. 2. Invite professional authorities in various subfields of your discipline as guest lecturers to your class. Have each discuss the frontier puzzles and unresolved issues in his or her field. Why is each puzzle or issue a good candidate for your research? Why haven't these problems already been solved through research? 3. A researcher "wears many hats." He or she may be a research entrepreneur, manager, analyst, employer, employee, writer, public
Guidelines for Conducting Scientific Research
67
relations expert, and fund-raiser. Discuss each of the roles in your profession. Which are you best at? What roles can you add to the list? 4. Strict positivists contend that social scientists employed in public institutions to serve society should confine their analysis to determining alternative means to achieve objectives given by society. Strict positivists say scientists should analyze only issues brought to them by their private employers or by the political or other public choice process. Modern analysts argue that this stance unduly constrains them, and are activists in telling government and society what the issues are, which ones should be researched, how issues should be researched, and what the answers are. Discuss the advantages and disadvantages of each approach. 5. List several problems worthy of research in your field. To do so, survey the latest issues of journals in your profession. Also survey editorial pages of magazines and newspapers. What problems are identified? Compare these with your personal choice of scientific problems to address. Why do your choices differ from those of journal and newspaper and magazine writers? [Note: Finding a worthwhile problem manageable in terms of time, apparatus, ability, and funding is one of the greatest challenges facing any researcher.] 6. Researchers are part of the nation's service industry, which has a poor record of improving its productivity. Discuss ways you can improve your research productivity-getting more output out of available resources (see also the PERT chart in Chapter 1). How should your output be measured? 7. Society and idealists are ever more conscious of ethical issues in research. Discuss the following: a. Your telephone survey is an hour long and catches people at the dinner hour because that's when they're home. Is this an undue imposition on your respondents? b. Your research needs some cats for testing some treatments. A friend says he can pick up strays. Do you give him the go-ahead? Your treatment causes the cats to become ill and some to die. Do you continue the treatment? You lack funds for adequate cages. Do you crowd the cats in a few small cages, terminate the experiment, or try to find more funds to get bigger cages?
68
Research Methods and Communication in the Social Sciences c. Your sample survey contains some "outlier" observations with values more than three standard deviations above and below the mean of the sample. Do you throw these observations out? Under what conditions is it ethical to omit some of your sample?
APPENDIX TO CHAPTER 2 The following sample of a suggested style of entries is the list of references from the American Journal of Agricultural Economics (AJAE) from "Preparing a Manuscript for Publication in AJAE" (1994). It is similar to the style used in this manuscript. Books a) One author. Van Home, J. S. Financial Management and Policy, 4th ed. Englewood Cliffs,N.J.: Prentice Hall, 1977. b) Two or more authors (do not use "etal." unless there are more than four authors). Johnston, Bruce F., and Peter Kilby. Agriculture and Structural Transformation. London: Oxford University Press, 1975. c) Organization, association, or institution as "author." National Research Council. Nutrient Requirements of Beef Cattle, 5th ed. Washington, D.C.: National Academy of Sciences, 1976. d) Editor as author. Fishel, Walter L., ed. Resource Allocation in Agricultural Research. Minneapolis: University of Minnesota Press, 1977. e) Component part by one author in work edited by another. Duloy, J., and R. Norton. "CHAC: A Programming Model of Mexican Agriculture." Multilevel Planning: Case Studies in Mexico, ed. L. Goreaux and A. Marine, pp. 291-337. Amsterdam: North-Holland Publishing Co., 1973. f) Forthcoming book. Masters, Harold E. Land-Grant Colleges Today, vol. 1. New York: Macmillan Co., forthcoming. g) Two or more works by the same author. List works chronologically, in ascending order. Substitute 3-em dash for the author's name after first appearance only if the name or names appear exactly the same in the second instance.
Guidelines for Conducting Scientific Research
69
Fuller, Wayne A. Introduction to Statistical Time Series. New York: John Wiley & Sons, 1976. . "Some Properties of a Modification of the Limited Information Estimator." Econometrica 45(1977):939-53. h) Anonymous works. If name of author is unknown, list by title (without "Anon."or any such addition). Article in a Journal or Magazine a) Article in a journal. i) With volume number given and pages numbered consecutively throughout the year: Rausser, G. C. "ActiveLearning, Control Theory, and Agricultural Policy.' Amer. J. Agr. Econ. 60(1978):476-90. Two or more authors (do not use "et al." unless there are more than four authors). Gebremedhin, T. G., and W. M. Johnson. "Small Farm Research and Policy Implication." South. J. Agr. Econ. 17(1985):47-56. ii) With issue number rather than volume number, numbering beginning with page 1 in each issue: Mayer, Andre, and Joan Mayer. "Agriculture, the Island Empire." Daedalus, no. 103 (Winter 1974), pp. 83-96. b) Article in a popular magazine. Prufer, Olaf. "The Hopewell Cult." Scientific American, Dec. 1964, pp. 90-102. c) Forthcoming journal article. Lambert, Phineas. "Parameters of Social Disintegration." Polit.Sci. Quart., in press. Reports a) Author of report named. Bachman, Kenneth L.,and Leonardo A. Paulino. Rapid Food Production Growth in Selected Developing Countries. Washington, D.C.: International Food Policy Research Institute, Res. Rep. No. 11, Oct. 1979. Jones, Roger, Janet Hall, and Robert Masson. Milk Marketing. Washington, D.C.: A Report of the U.S. Department of Justice to the Task Group on Antitrust Immunities, Jan. 1977.
70
Research Methods and Communication in the Social Sciences
b) Committee, commission, or association the "author." Special Dairy Advisory Committee. Examining Dairy Policy Alternatives. Washington, D.C.: Report prepared for Milk Industry Foundation and International Association of Ice Cream Manufacturers, Sept. 1979. National Commission for the Review of Antitrust Laws and Procedures. Report to the President and the Attorney General. Washington, D.C., Jan. 1979. University Publications and Papers a) State agricultural experiment station bulletins. Small, L. E., V. Kasper, Jr., and D. A. Derr. Transfer of Development Rights: Marketability. Rutgers University Agr. Exp. Sta. Bull. No. 848, 1978. b) University departmental bulletins and monographs. Greig, W. S. "The Changing Structure of the Food Processing Industry." College of Agr. Res. Center Bull. 837, Washington State University, Sept. 1976. Bieri, J.,and A.de Janvry. "Empirical Analysis of Demand Under Consumer Budgeting." Giarmini Foundation Monograph No. 30, University of California, Berkeley, 1972. c) Departmental staff papers, mimeographed papers (unpublished papers). Wiggins,E. R. "SmallFarm Development Programs." Dept. Agr. Econ. Pap. 1972-8, University of Missouri, July 1972. McAfee, D. H. "Farm Opportunities Program." Mimeographed. North Carolina Agr. Extens. Serv., Sept. 1975. d) Unpublished theses. Polito, Joseph. "Distribution Systems Planning in a Price Responsive Environment." Ph.D. thesis, Purdue University, 1977. U.S. Government Publications a) Congressional publications (journals of House and Senate, debates, reports, hearings, committee documents, statutes). i) Congressional committees or task forces: U.S. Congress, House of Representatives, Task Force on Federal Flood Control. A Unified National Program for Managing Flood Losses. Washington, D.C.: House Document 465,89th Cong., 2nd sess., 10 Aug. 1966. ii) Congressional Record: U.S. Congress, Senate. Congressional Record, Washington, D.C.: 77th Cong., 1st sess., vol. x, party y, full date, page numbers. iii) House or Senate bill:
Guidelines for Conducting Scientific Research
71
U.S. Congress, Senate (House). Name of Bill, nth Cong., xth sess., full date of session. iv) Reports and hearings: U.S. Congress, Senate, Committee on Commerce. Conversion to Metric System: Hearing on S1278, Washington, D.C.: 88th Cong., 2nd sess., 7 Jan. 1964, p. 58. b) Executive departments and agencies. i) U.S. Department of Agriculture: With author names: Brown, Lester R. Increasing World Food Output: Problems and Prospects. Washington, D.C.: U.S. Department of Agriculture, ESCS For. Agr. Econ. Rep. 25, Apr. 1965. Without author named: U.S. Department of Agriculture. Changes in Agriculture in 25 Developing Nations, 1948-1963. ESCS For. Agr. Rep. 27, Washington, D.C.,Nov. 1965. ii) Other government departments or agencies: U.S. Bureau of the Budget. The Budget of the United States Government for the Fiscal Year Ending June 30, 1964. Washington, D.C., 1963. U.S. Department of Commerce, Bureau of Economic Analysis. Survey of Current Business. Washington, D.C.,selected monthly issues. Other Publications a) United Nations publications. United Nations. Yearbook of National Accounts Statistics, 1963. New York, 1964. United Nations, Food and Agriculture Organization. Agricultural Commodities, Projections for 1970. Rome, 1962. b) Publications by foreign governments or their agencies. Ghana, Republic of, Central Bureau of Statistics. Economic Survey, 1961. Accra: Government Printing Department, 1962. c) Newspaper articles. i) Author named: Russell, J. S. "Freedom from I.S.C. Curbed, Schultz Says." Des Moines Register, 19 Sept. 1943, p. 1. ii) No author named: Wall Street Journal. "Cattle on Feedlots Fall by 5%, Signaling Rise in Prices Later in Year." 20 Jan. 1981, sec. 2, p. 38.
72
Research Methods and Communication in the Social Sciences
d) Paper presented at a meeting. Author. "Title." Paper presented at the annual meeting of the AAEA, Blacksburg, VA, 6-9 Aug. 1978. e) Unpublished paper. Author. "Title." Unpublished paper for Social Science 59th Anniversary Celebration, University of Chicago, 5 Dec. 1979. f) Acts and laws. Acts and laws do not need to be referenced. The full name of the act must be mentioned in the text for federal laws; for state laws, name of state must also be mentioned. However, acts and laws should be referenced if they are quoted. Additional Information Regarding Citations Any work cited in the paper should be in the reference list. All citations must include the author(s) of the work. If you mention the author's name in the text of the paper and are not referring to a particular page of the work, no additional citation is necessary: Martin Evans discusses guaranteedprice adjustment in the United Kingdom. If you include the page number, the citation becomes: Martin Evans (p. 56) discusses guaranteed price adjustment in the United Kingdom. Otherwise, insert citations parenthetically at the end of sentence, or, if necessary, at the end of a main clause. Example: Between 1974 and 1979, the average total consumption of nitrogen per acre of corn in Ohio increased 51% (Smith, p. 69). Be sure page numbers are given for specific citations. If two or more items in your reference list are by the same author, use dates to distinguish between them: Total consumption of nitrogen per acre of corn in Ohio increased 51 % (Smith 1971, p. 69). If two or more items in your reference list are by the same author and were written in the same year, distinguish between them by using lower case letters after the date: Total consumption of nitrogen per acre of corn in Ohio increased 51 % (Smith 1971a, p. 69). These lowercase letters, of course, also should appear in the reference list: Roberts, Ralph E. Econometric Approaches to Farm Labor Problems. Boston: Little, Brown, and Co., 1976a. Finally, although styles will differ from publication to publication, the style must be consistent within a publication.
3
Guidelines for Preparing Research Grant Proposals Most research projects require some type of financial or material support. Obtaining such support often necessitates the development of a grant proposal for research. The proposal is constructed much like a research proposal, as discussed in Chapter 2. The proposal guidelines in this chapter are designed to assist graduate students in writing research proposals for theses and dissertations, and others preparing proposals to obtain grants and contracts. These efforts are based on scientific research methods and technical writing skills. No one set of guidelines or instructions on how to write proposals suffices for everybody. The guidelines are intentionally simplified in the following sections to facilitate an understanding of each component of a proposal. The guidelines are sufficiently flexible to adapt to a wide range of circumstances. The research proposal may be looked upon as a sales, planning, and management document. It describes the intended research in a manner that will stimulate the interest of the potential sponsor, pointing out its value, both economic and social, to the community or clientele. It shows the process by which the objectives of the research can be met and solutions can be achieved. Research proposals, like other forms of written communication, are best introduced by a short, carefully devised statement that establishes the overall area of concern. The preparation of a proposal does not follow a rigid formula. As the personality of a salesman is reflected in his sales approach, so the personality of the researcher is reflected in the presentation of the proposal. Proposals presented by different researchers will differ, though they might address the same subject and may offer the same alternative
74
Research Methods and Communication in the Social Sciences
solutions. However, a proposal must compete with other equally ambitious proposals. The task is to assemble a proposal that beats the competition. Identifying potential funding sources is the first step in preparing a successful grant proposal. The next step is to obtain the potential funding agency's published submission guidelines, which will vary among academic institutions, government agencies, and industries. When the funding organization invites proposals for a project, it issues a document called the Request for Proposal (RFP). The RFP normally outlines in great detail what must be addressed. Researchers should read the instructions in an RFP very carefully and use the RFP outline for the topics to be covered in a research proposal. To enhance the probability of funding, the submitter needs to learn the needs of the funding organization, to organize the proposal accordingly, to follow directions precisely, and to complete and submit the proposal on time to the right place. In this chapter we outline guidelines to help researchers and institutions prepare research proposals having favorable chances for funding (see also Bly and Blake, 1982; Meador, 1990). For further illustration, two samples of RFPs and two samples of research proposals are included in Appendixes at the end of this book.
GENERAL COMPOSITION AND ORGANIZATION Begin by preparing an outline of the planned proposal. The researcher should ensure that the proposal is complete but as concise as possible. Proposals are evaluated on the basis of several criteria, which may differ from one funding agency to another. However, criteria often include the clarity and relevance of the problem statement, the cogency of the conceptual approach, the availability of relevant data, the duration and budget, and the appropriateness of the research methods to the problem being addressed. A proposal is a challenge to the author's scientific thought and technical writing skills. It should be well organized to convey ideas logically and in an orderly manner. Typical proposals, solicited or unsolicited, willbe required to contain most if not all of the following elements in the final proposal package. The works of Locke and colleagues (1987), UNDESA (1973), UNESCO (1968), and Meador (1990) are incorporated into this discussion to illustrate the elements of a typical format.
Guidelines for Preparing Research Grant Proposals 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12.
75
Cover letter Title page Table of contents Proposal summary or abstract Introduction Text or main body of the proposal Significance or impact of the proposed research Personnel and capabilities of the institution Supportive services and cooperation Duration of the project Dissemination of the results Proposed budget
Some elements will carry more weight in the review process than others. Although all elements in the proposal are not equally important, each should receive careful attention and should be consistent with the proposal as a whole. The proposal instructions may dictate the length of particular sections. If the instructions only specify the total number of pages allowed, pages should be carefully allotted to give the elements that count most in proposal evaluation the greatest emphasis. If the proposal instructions do not provide any limitations on length, reasonable limitations should be established based on the requirements of the proposal. A bound and covered proposal ordinarily is used when the total number of pages is greater than five or when the proposal is complex enough to require formal subdivision. Cover Letter A cover letter will be the first page of a proposal unless it is specifically prohibited. The proposal instructions in a solicited proposal or RFP may not specify a cover letter. However, if the instructions do not prohibit such a letter, attaching a brief one with copies of the proposal will serve several purposes. In the case of an unsolicited proposal, the cover letter is the first thing the funding agency will read. Because it orients the recipient, the cover letter is a particularly valuable part of an unsolicited proposal. In the cover letter, the researcher can introduce himself or herself, start establishing his or her credentials for the project, highlight special features of the proposal, and add any useful details not included in the proposal (Meador, 1990).
76
Research Methods and Communication in the Social Sciences Title Page
The title page or proposal cover sheet often is a form provided by the proposal recipient (funding agency). Items commonly included on the title page are: a brief and professionally informative title of the research proposal; the complete name, mailing address, and telephone and facsimile (FAX) numbers of the principal investigator(s) and/or the institution(s) submitting the proposal; the name and location of the organization to which the proposal is being submitted; the date of submission; the proposal number (if necessary), to permit easy reference and filing;the total estimated budget; the proposed starting date and duration of the project; and signatures of the principal investigator(s) and an official authorized to commit the grantee institution in administrative and financial affairs. Proposals lacking these signatures and endorsement are considered incomplete and are usually returned to the applicant. The title of the proposal is the first contact a reader has with the proposed research. First impressions generate powerful expectations about what is to follow. The choice of an attention-getting and mind-grabbing title is critical. The first rule in composing a title is to achieve reasonable parity between the images evoked by the title and the opening pages of the proposal. The proposal title calls for careful consideration of all of the functions it must serve and the standards by which it will be judged. In general, the title should be brief and concise and describe as accurately as possible the main elements in the study.
Table of Contents A table of contents should be included if the text has more than one level of subdivision or if, having only one level, it contains more than about five headings. Page numbers should be provided for all entries and the entries should be listed in the order they appear in the text. Tables, figures, abbreviations, and literature cited should be listed with page numbers in the table of contents. Project Summary or Abstract The proposal summary ("abstract" and "summary "are used interchangeably herein) is likely to be one of the last written but most frequently read
Guidelines for Preparing Research Grant Proposals
11
elements of the proposal. The summary may be published by the recipient of the proposal as a description of the project. An abstract is typically prepared if the proposal is longer than about ten pages and has more than one level of subdivision. Purposes of the proposal summary are to: 1. 2. 3.
Focus the thinking of individuals developing the proposal by establishing a clear and brief summary of major elements. Obtain preliminary administrative approval or to solicit support and cooperation from other units. Serve as the basis for a letter of intent, which, in essence, is a one- or two-page abstract of the proposed study. Potential sponsors are likely to read the letter of intent first. Then, the letter is screened and ranked by a panel of reviewers on preestablished criteria. Whether or not a letter of intent is required, the abstract bears a disproportionate share of responsibility for success or failure of the proposal.
In the abstract, the contents of the proposal are summarized to no more than one or two pages, with the primary intention of conveying key information to those who need not read the full proposal. It should include clear statements of the need for the research work, the research objectives, the anticipated results and benefits, and the capabilities to achieve the objectives. Because the abstract is a one-shot communication, absolute clarity is essential. It must accomplish the dual tasks of providing a concise picture of the study while highlighting the unique characteristics that will sustain the interest and attention of the potential sponsor. Reviewers frequently first read the summaries (abstracts) in a stack of proposals to gain an overview of each proposal. Holding this background information in mind, a reviewer then deals in depth with each separate proposal. This review method makes it desirable for a proposal to be fronted by a forceful and persuasive proposal summary that wins attention and favorably influences the reviewers (Meador, 1990). Thus, in preparing the summary it is important that each word and sentence convey a precise message to the reviewer in plain language. Only information essential to a coherent, comprehensive statement of the proposed research work should be included.
78
Research Methods and Communication in the Social Sciences
Introduction The introduction is a particularly useful part of an unsolicited proposal. The introduction is used to provide background information on the project and the researcher. The introduction can help confirm the qualifications and experience of the researcher and the resources for performing quality research of the type proposed. Such information may appear at length elsewhere in the proposal, but the introduction provides advance, capsulized support of the information. An effective introduction gives a few lines of history about the capability and experience of the institution seeking a grant and briefly covers the goals, accomplishments, successes, and proven ability of the researcher in related studies. The introduction prepares the reviewer for the sections of the proposal dealing with the specific project (Meador, 1990).
Text or Main Body of the Proposal The main body is the heart of the proposal. The section is especially scrutinized by reviewers because it specifically identifies the nature and thrust of the project. The text of a proposal includes a thorough narrative statement of the nature and significance of the research problem, the hypothesis, objectives, literature review, and the procedure. The main body of the proposal should begin with a statement of a problem and of what will be accomplished. It may be presented in the format of a research question or testable hypothesis. In developing a problem statement, several cogent points are reviewed from the earlier discussion in Chapter 2. 1.
2.
To be worthy of research, a problem needs to identify a gap between current circumstances and preferred circumstances. The problem statement potentially should appeal to felt needs of individuals, groups, and/or societies. The statement needs to make a case that expected benefits exceed expected research costs to society, and that the research is a significant contribution to the current body of knowledge. The problem needs to have a solution that is economically and technically feasible, politically acceptable, and socially desirable. It must meet environmental requirements. A suitable problem should be researchable with the human, financial, and other resources made available.
Guidelines for Preparing Research Grant Proposals 3.
4. 5.
79
The cause and effect relationship expressed in a problem statement should be based upon conceptual foundations or previous research findings relevant to the problem under scrutiny. The statement must suggest meaningful, testable hypothetical relationships that, when analyzed, will yield relevant results or factual information to help resolve the problem. Research ordinarily should not duplicate what has already been done. It should yield new knowledge or update previous scientific research results. The problem must be suitable for the particular researcher with expertise in the field to exploit his or her knowledge, experience, and enthusiasm. The problem statement should link the project with the experience, activities, and history of the organization and the researcher.
A literature review of previous work is initiated in the early stages of preparing a research proposal and continues as the research progresses. The major functions of a literature review are to learn and understand previous research in terms of the theory developed, research procedures used, empirical results obtained, interpretations and inferences made, and conclusions reached; to identify remaining gaps that the current research will address; to identify the frontier or scope of the research study; to avoid replication or duplication of previous studies; and to provide a base for formulating hypotheses. It is not necessary to include an extensive literature review, but it is essential to include relevant citations in the body of the research proposal. A bibliography of literature pertinent to the proposed research is also included to show the conceptual and methodological thoroughness of the project. The objectives and hypotheses state what the research plans to accomplish and benefits to be attained, clarify the direction and limit the scope of the research study, and suggest what information will be obtained to help resolve the problem. The objectives and hypotheses of the proposed research must be sufficiently specific to conform to budgetary restrictions and sufficiently broad to satisfy the needs of clientele. The objectives should be concrete and clear-cut and the hypotheses testable and nontrivial. The clarity and precision with which the objectives, hypotheses, and problem statement are presented set the tone for the subsequent section on methods and procedures. To address the stated problem statement and accomplish the specific objectives, it is important to include a detailed step-by-step description of all
80
Research Methods and Communication in the Social Sciences
procedures necessary for study completion. The proposal should state precisely what methods will be used, what steps will be taken, and what schedule will be met in reaching the objectives. If the methods to be used in the project are unusual or original, these details should be positively stated with the reasons why the methods are appropriate. Particular abilities, qualities, and features of the project that put the proposal ahead of other proposals should be noted. The procedure must be sufficiently thorough to show that each task will be managed and performed in a methodologically sound fashion. The realities of the project are ultimately dictated by the feasibility of the methodology and work plan, reflecting the potential achievement of the objectives (Meador, 1990). The discussion of methodology used in the project must include the types (time series or crosssectional data) and sources (survey or secondary) of data and the proposed methodology to collect and analyze the required data. The research proposal should be built around factual information within a conceptual framework. A well-prepared project description relates the proposal to the present knowledge in the field, outlines a plan of work, specifies the technical approach and experimental methodologies and procedures to be followed, and describes the facilities and instrumentation to be used. How the research is to be performed determines its quality and potential, and is why some potential sponsoring agencies fund procedures more than the objectives of the research. The project description should be concise and not exceed the number of pages required by the sponsoring agent. The description generally will not exceed ten single-spaced typed pages. The main body of the proposal should be written on the premise that the cooperating or funding organization is unfamiliar with the problem(s). A grant or contract proposal is likely to contain more justification than a project statement or research report. The reason is the need to convince the potential funding agency of the merit of the project. By far the best sales pitch is a carefully designed and well-written project of interest to the funding agency. Careful formulation and presentation of the implications or possible applications of knowledge help to give the proposal an urgency, justifying its worth. In summary, a technical proposal succeeds or fails on the strength or weakness of the problem statement (including benefits to be achieved), objectives, and procedure.
Guidelines for Preparing Research Grant Proposals
81
Significance or Impact of the Proposal It is not unusual for funding agencies to require a section entitled "Significance of the Proposed Research." Funding agencies usually are accountable to the public, their benefactors, or another authority for the expenditure of their funds. The proposal and the analysis it will generate constitute the last opportunity to convince the prospective funding agency that the project is important and should be supported. Responsiveness to the criteria and the existing interests and commitments of the funding agency will improve chances to obtain support for the research. Because funding agencies frequently shift their emphasis, applicants need to work at knowing what areas or topics of research are timely.
Capabilities of the Personnel and the Institution The proposal must convince a potential sponsor or funding agency that the soliciting institution or principal investigator is capable of doing the work being proposed. A paragraph or two should describe the performing agency's human resources and support facilities, including the size and professional make-up of its staff, its commitment to performing the project, its professional strengths, and relevant collaborative relationships with government, educational, or professional organizations. Capabilities may be explained by describing disciplines in which exemplary or complementary research has been or is currently being undertaken. Special facilities or unique capabilities (those that competitors might not be able to offer) that would make the institution particularly well suited to perform the various tasks of the proposed project should be described. The success of completed projects and some important ongoing projects are mentioned to highlight skills and capabilities available for the proposed project. The identities, education, experience, and qualifications of the personnel involved in carrying out the project are essential ingredients of a proposal. Reviewers inevitably reach conclusions about the value, credibility, and prospects for success of a project on the basis of the people involved as much if not more than on the basis of the objectives, methodology, and plan of action (Meador, 1990). Thus the staff needed for the research project should be carefully selected and show their readiness and proven ability to perform the tasks. The proposal should note specific duties and responsibilities of those assigned to the research. Assignment of a well-qualified project leader and able supporting staff is often a key factor. Letters of commitment and
82
Research Methods and Communication in the Social Sciences
availability are important from personnel who are not with the applicant agency but critical to the project. Proposal instructions commonly require those submitting proposals to include resumes of the project leader or principal investigator and any other key professionals. Many formats exist for the preparation of resumes to accompany a proposal. "The simpler the format, the better" is a good rule to apply when compiling personal fact and achievement summaries. The length of the curriculum vitae may be restricted by the instructions. A twoto three-page resume is proper, though longer resumes sometimes are necessary. The following items are normally found in the curriculum vitae for a research proposal: 1.
2. 3. 4. 5.
6.
Name and title together with address and telephone number of the principal investigator and the department of the institution with which he or she is identified. Education-universities attended, degrees received, and field(s) of competence (not fields of interest). Professional experience -work record with detailed description of positions and responsibilities in reverse chronological order. Professional skills-a documented description of capabilities and credentials, unique skills, and evidence of relevant training. Professional roles and activities -professional membership, honors, scholastic awards, and other involvements pertinent to the project. A list of major publications completed during the past five years, selected for project pertinence; previous grants with involvement in a similar study; and related workshops and conference presentations.
The proposal instructions in the RFP usually will stipulate exactly what materials may be added to the proposal as appendixes. Appendixes that supply background information and documentation may accompany unsolicited proposals. Staff resumes, organization history, summaries of successful projects previously completed, letters of support and recommendation, and further documentation of materials in the main body of the proposal are examples of supplementary and support materials often best suited for an appendix.
Guidelines for Preparing Research Grant Prop
83
Supportive Services and Cooperation It is essential to describe the roles played by each institution and organization that will collaborate on the project. Written evidence should attend the proposal, showing that the appropriate professionals have reviewed the proposal and are prepared to participate if the grant is awarded. Funding chances are often enhanced if such individuals and groups have a high political or professional profile that will bring recognition to the project. Having an advisory board of persons with high professional and political (but noncontroversial) profiles can be decisive in a successful grant or contract application.
Duration A time frame for starting and completing specific tasks and parts of the research project should be specified (see PERT chart in Chapter 1). Useful functions of that frame include: 1. 2.
3.
4.
Helping to keep the principal investigator and supporting staff on schedule throughout the duration of the project. Allocating and scheduling time for hiring personnel, ordering equipment and supplies, putting equipment and facilities into operation, collecting and analyzing data, and writing progress and final reports. Enhancing the reviewer's understanding of the entire project. A well-conceived time frame will document the applicant's organizational skill. It will convince the reader that the applicant knows the area, the methodologies to be used, and the effort required. Identifying specific tasks to be performed, preparing program management and budgeting, and establishing progress benchmarks for the timely report of research findings.
Dissemination of Results Research is incomplete without dissemination of results. Funding agencies concerned about public relations usually want a major effort to distribute
84
Research Methods and Communication in the Social Sciences
findings. The proposal needs to explain which results will be reported, the appropriate audience, and the plan or form (popular press, research reports, bulletin, journal papers, etc.) to be used to disseminate the final results of the study. All parties involved should be aware beforehand if the funding agency holds proprietary rights to findings and/or editorial power over publication of results. Proposals that show promise of future life, further growth, and prospects of providing further information in the future are most favored.
Budget The funding organization rightfully will be concerned about the economic feasibility of a project. No funding organization wants to waste scarce resources on a project whose proposal budget shows excessive expense relative to anticipated results. The project should establish a clear connection between the proposed budget and the specific tasks to be performed. Most application formats for research project funding require that the budget be presented in brief form with no more than a page or two of appended explanation. An understanding of what portions of a budget go for specific purposes at specific times needs to be worked out between the parties involved. The principal investigator usually seeks as much flexibility as possible in budgets, while funding agencies often request specificity. Grant applications must include a summary budget covering the duration of the proposed project and separate budgets for each individual year of support requested. The budget lists personnel for whom funds are requested. Personnel time and responsibility for the tasks of each position are specified. The qualification or special certification desired for the project is listed. Contributions in cash or kind by sponsoring individuals or institutions are noted in the proposed budget. Expected expenses for supplies and equipment are included. Generally, funding agencies are not disposed to look with favor on requests for extensive equipment purchases. The position of many agencies is that the institution where the research is to be performed should supply all of the equipment, as well as most large items that represent expensive and permanent investments in research capability. If funds for the purchase of major equipment are requested, the budget justification section might explain why alternative plans such as rental or sharing will not suffice. If equipment is purchased, the proposal should clarify who will own, upon project completion, the equipment purchased with project funds.
Guidelines for Preparing Research Grant Proposals
85
Costs for data collection and dissemination of results can be built into the budget. Costs for everything from the preparation of the manuscript to publication printing and mailing reports are also appropriate. Only those travel funds that are necessary to complete the research and, perhaps, to disseminate findings should be requested in the budget. The proposed budget will include indirect or overhead costs that the funding agency will pay directly to the institution. Such costs vary widely and are negotiated by the agency and the university or other, recipient institution. Many large institutions have a standing policy on overhead costs for all grant contracts. Indirect charges presumably reimburse the university for the cost of administrative overhead, building maintenance, utilities, and all items that would have to be purchased by the grantor if the study were conducted at a facility owned by the agency. Other budget items dictated by the institution are fringe benefits such as insurance, retirement, and medical benefits for project personnel. Some funding agencies require a minimum match or sharing of costs of the research. It is always wise and generally necessary to take an early draft of the budget to the university officer in charge of grant negotiations and obtain assistance on budget policies and procedures.
PROCESSES FOR EVALUATING PROPOSALS The completed proposal should be reviewed both internally and externally. The researcher, team members, and other peer reviewers give the proposal a final evaluation before sending it to the funding organization. Project proposals also will be reviewed and cleared by the home institution before they are submitted to the funding agencies (Meador, 1990). The most important evaluation will be by the reviewers for the funding organization. Proposals usually undergo administrative and technical review at the sponsoring agency before approval for funding. The funding agency usually employs a rigorous review system that emphasizes scientific and technical merit. The proposal should be evaluated in light of the following checklist: A. Proposal Format 1. Does the format of the proposal adhere to suggested guidelines specified by the granting agency? 2. Is the proposal well prepared and concisely written? 3. Is the project succinctly summarized in the abstract?
86
Research Methods and Communication in the Social Sciences 4. Does the proposal properly review and cite pertinent previous and current research findings from the scientific literature? B. Relevance of Proposal 1. Are the objectives of the proposal consistent with the mission of the university (grantee organization) and the granting agency? 2. Do the objectives of the proposal address appropriate concerns and problems of the target clientele, considering the capabilities of the research unit and resources? 3. What is the overall scientific merit of the proposal in terms of importance and potential impact on progress and contribution to the scientific field? C. Appropriateness of the Methodology 1. Do the procedures address each objective? 2. Are the procedures and experimental design methods appropriate for achieving the objectives and testing hypotheses? 3. Does the methodology provide for appropriate statistical evaluation of the results? 4. Is a clearly defined time frame provided for completion of each objective? 5. Are the conceptual foundation and empirical approaches built upon the current level of knowledge documented in the review of literature? D. Personnel, Facilities, and Cooperation 1. Can the personnel involved (with regard to time commitment, expertise, and resource availability) adequately and reasonably be expected to carry out the research project? 2. Does the research proposal show evidence of appropriate multidisciplinary collaboration in planning and proposed conduct of research? 3. Is there appropriate collaboration with other organizations and institutions? Have suitable commitments been obtained? Have necessary liaison contacts been made with persons and agencies who will be part of or affected by this research?
Guidelines for Preparing Research Grant Proposals
87
4. Are adequate support personnel, facilities, equipment, and supplies available (or to be made available) for successful completion of the project? 5. Have guidelines for affirmative action, animal welfare, safety, and environment been met? E. Budget 1. Have the major budget categories (personnel, direct and indirect costs, etc.), as applicable, been included in the proposal? 2. Is the budget consistent with and realistic for accomplishing the specified objectives? 3. Does the budget reflect the contributions of participating institutions? F. Information Dissemination 1. Have suitable plans been included in the proposal for making the findings of the project readily available to the intended audience? 2. Is there a specific plan to publish the results in journals or other appropriate channels for utilization and application of results?
SUMMARY An institution or individual researcher may be able to plan and conduct a research project. But unless the ideas can be communicated in a convincing manner to a prospective funding organization, the project will not be funded. The research proposal is the medium through which such information is communicated. The proposal must stand alone in conveying to the potential sponsoring agency the ideas, concepts, and capabilities of the research personnel and institution. The proposal must tell the funding agency what the researcher plans to do, the research or service involved, the cost of conducting it, the facilities and resources available to do the proposed research or service, and the overall experience of the researcher and the associated institution. A proposal is the primary vehicle for communicating to a funding agency a plan for action and, potentially, a contract. The length of time and sequence required for approval of projects and funding by sponsoring agencies vary from one organization to another. A funding agency may review and process grant proposals once or twice a year. Proposals should be submitted at least six months prior to the start date.
88
Research Methods and Communication in the Social Sciences
Thus delay in preparation of a research proposal can be fatal for the grant or contract. Obtaining institutional review and approval, peer reviews, letters of support, budget information, commitments from personnel, vitae, and multiple signatures is a tedious but essential process.
REFERENCES Bly, Robert W., and Gary Blake. (1982). Technical Writing (Structure, Standard, and Style). New York: McGraw-Hill. Locke, Lawrence F., Waneen Wyrick Spirduso, and Stephen J. Silverman. (1987). Proposals that Work (A Guide for Planning Dissertations and Grant Proposals). Second edition. London: SAGE Publications. Meador, Roy. (1990). Guidelines for Preparing Proposals (A Manual on How to Organize Winning Proposals for Grants, Venture Capital, and R&D Projects). Seventh edition. Chelsea, MI: Lewis Publishers. U.N. Department of Economic and Social Affairs (UNDESA). (1973). Research Proposals: A Guide for Scientists, Technologists, and Research Institutes in Developing Countries. ST/ECA/187. New York: United Nations. UNESCO. (1968). Guide to the Preparation of Scientific Papers for Publication and Guide for the Preparation of Authors1 Abstracts for Publication. UNESCO/SC/MD/5.
EXERCISES 1. Prepare a grant proposal for class. Include the introduction or problem, objectives and/or hypotheses, and procedure. 2. List a few agencies that fund research in your field. Where can you get names and addresses of agencies? How do these agencies differ in the types of research they fund? 3. Give a brief oral presentation of your research proposal in class. Ask listeners to suggest improvements in your presentation. 4. Where can one go to get help in improving a research proposal? Pay a visit to the research foundation that handles research budgets at your agency or university-if such a foundation exists. Ask how the foundation can help you to prepare research proposals. (Some funding agencies
Guidelines for Preparing Research Grant Propo
89
provide seed funding support for analysts to prepare grant and contract proposals.) 5. Ask a successful grant or contract recipient to speak to your class on tips for successful grantsmanship.
This page intentionally left blank
4
Guidelines for Communicating Scientific Research Even the best scientific research is useless unless results are communicated. In a private firm where knowledge is proprietary, communication is likely to be internal. In a public agency, research results usually are communicated as widely as possible. The first written communication is likely to be a thesis, dissertation, or report to the sponsor of research. Results can be disseminated in published form through various outlets such as popular publications, books, journals, conference proceedings, research station reports, and bulletins. The style and order of presentation will differ according to the outlet and audience. This chapter emphasizes academic and professional publications rather than newspaper articles. Researchers and readers of scientific research often do not have a common understanding of the subject matter. Research will not be communicated if the scientific research is reported in tortuous prose, esoteric jargon, unfamiliar concepts, and meaningless statistics. Scientific writing follows the scientific process of performing research from its initial stage to its completion. Good organization is the key to sound research and writing. To write well is to think well. One secret of writing well is to know one's subject well. Writing skill, like most skills, is not inborn. The skill of useful and practical writing must be learned. The skill of research writing will come mostly from self-training. Clear, concise, well-organized, and precise writing is very hard work indeed. Of course, complex subjects may not be easy to communicate in simple and clear language to all readers. The ideal is to work for the greatest clarity possible given the difficulty of the material to be
92
Research Methods and Communication in the Social Sciences
communicated. This chapter specifically focuses on the basic requirements of written and oral communication.
REQUIREMENTS FOR SCIENTIFIC WRITING Scientific writing is the most objective type of writing. The main purpose of scientific writing is to convey precise, correct information in the most efficient and effective form to the intended audience. There are two basic types of intended audiences-the specialized or professional audience and the general or popular audience (Collins and Tuttle, 1979). Know your audience. Aim your message directly at your readers. To do this you need to know about readers' interests, desires, and level of knowledge. What works well for communicating with one group may not work for another. Do not assume that the reader has full knowledge of the subject matter. Always put yourself in the place of the reader and analyze what you would expect from the author. Scientific or technical writing for the popular audience is perhaps the most difficult task because the author has to have command of the content and vocabulary of the specialized field well enough to simplify it without distortion. Effective writing about scientific investigation conveys to the readers not only the scientific subject but also its significance in practical life. Scientific writing for the general public makes accessible concepts and theories often hidden by scientific jargon and techniques. Such writing translates scientific discovery into language comprehensible to the general audience. The unique situation of scientific writing for the general audience creates a special obligation for the author. The general audience usually does not have the knowledge and skills to verify the research findings or conclusion. The scientific writer has an ethical responsibility for stating limitations and biases that other writers may overlook. Of course, the specialized professional audience also needs accurate communication of the concepts or procedures from the author (Collins and Tuttle, 1979). Our prime concern in this chapter is not formal grammar. The writer needing further training in English composition and structure should consult an English instructor or a reliable, up-to-date reference on the subject. This text teaches the scientific writing process by applying first principles to sharpen communication skills. The following are some basic principles and requirements for effective scientific writing (see also Bly and Blake, 1982;
Guidelines for Communicating Scientific Research
93
Bates, 1985; Barrass, 1978; Day, 1988; Samuels, 1978; Hays, 1965;Turabian, 1969; Bolsky, 1988; Ladd, 1979). Completeness Scientific work is useless unless it is understood by the intended audience. Scientific research reporting begins with a clear and complete statement of a problem and proceeds toward a clear and comprehensive conclusion or solution. Too often, essential information is omitted because it is assumed to be obvious to readers. What is obvious to you (having experience in the subject matter) may not be known to others. Writing should be free from errors of omission, but also should show an awareness of the limitations of your knowledge in general and the general limitations and constraints of the study in particular (see Bly and Blake, 1982; Barrass, 1978).
Accuracy The scientific method depends upon care in observation, measurement, recording of observations and measurements, and analysis. Research procedures should be reproducible and conclusions verifiable. Because the intended goal of any scientific writing is the transmittal of technical information, even the most well-written scientific document is ineffective if the facts, theories, and observations are erroneous. Scientific writing that contains technically inaccurate statements reflects inadequate knowledge of the research and improper use of language (Bly and Blake, 1982; Barrass, 1978). Among other things, accuracy and clarity depend upon the care you take in the choice and use of words. Objectivity and Impartiality In science, statements should be based on evidence and not on unsupported opinions. Speculation cannot replace evidence. Preconceived biases too often emerge in the conclusions, undeterred by intervening facts and logic. Some analysts torture data until they confess to anything! An opinion is not a fact merely because it is believed by the majority. Authority is not a substitute for evidence. When no more information is available on any point, it is not time to speculate but to note the need for further work.
94
Research Methods and Communication in the Social Sciences
Do not hide unpleasant details with vague statements or generalities or by burying them in a mass of boring detail. If the final results are inconclusive, the writer should say so. Intellectual integrity and objectivity are more important than advocacy. Convince readers by the weight of evidence and not by emotion (Bolsky, 1988; Barrass, 1978). Incorrect assumptions lead to incorrect conclusions. State the assumptions underlying your analysis. Any conclusions, projections, or generalizations should be based on scientific evidence (Barrass, 1978). Indicate the source of the data and analysis and specify the limitations of your work, the source's probable errors in the data, and the range of validity of the conclusions. Beware of overestimating the importance of your research work. Neither omit evidence that is against your hypothesis nor undervalue the findings of other scientists when these seem to contradict your own work. Most disciplines contain schools of thought variously favoring conservative, liberal, radical, or revisionist points of view. The tendency is for scientists within a school to cite mostly each other. The appropriate approach is to utilize the most reliable knowledge from analysts making careful use of facts (probabilistic) and logic-whatever school of thought they come from.
Order Arrange your material logically and identify clearly the relationship between topics in the document. The reader will find your message easier to understand if information and ideas are presented in a logical order. Inconsistencies in the presentation will confuse readers and convince them that your scientific work is as disorderly as your prose (Bly and Blake, 1982). Begin with ideas the reader can readily understand. Difficult material can be introduced step-by-step. Evaluate the whole of your material to see if thoughts flow logically. The line of thought should remain consistent within a paragraph. It may change between paragraphs but appropriate transitional lead-in and lead-out sentences should prepare the reader for the change. Organize the document in short sections, each with a single idea. Put the most important information first. Start each major section on a new page. Combine related information in one place. Avoid issues in the text that are not central to the study. Especially avoid nonessential controversial issues. Let colleagues and others review your material. Revise the material based on their comments and suggestions (Barrass, 1978; Bolsky, 1988).
Guidelines for Communicating Scientific Research
95
Simplicity The primary goal of scientific communication is to transmit information, not necessarily to persuade. But the most persuasive scientific documents offer a clear and objective presentation of the relevant facts and logic that support the conclusions. Simplicity in writing, as in a mathematical proof, is the outward sign of clarity of thought. Scientific writing is direct, concise, logical, and precise. It is free from jargon, cliches, and other distracting elaborations. Use concrete terms instead of generalizations. Be direct to save words but polite to avoid misunderstanding. The author should not use scientific jargon just to impress the readers-it usually impedes the effectiveness of the communication. Some technical terminology is valid and necessary, but too much makes writing incomprehensible. If a term is essential or if it facilitates the flow of the writing, it should be used. The writer, however, needs to define and explain the difficult term or concept. Keep the writing short and simple. Break up the writing into short sections, and try to limit each section to one central theme or idea. In summary, the more direct and simple a composition, the more effectively it communicates (Bly and Blake, 1982; Barrass, 1978).
Clarity Simplicity promotes clarity. It is essential to include enough information in the document so that main points are conveyed without having to refer to other material. Ambiguity, redundancy, and wordiness in the writing detract from communication. Do not repeat words or ideas unnecessarily, but some repetition can increase clarity. In short, write not to impress but to be understood by applying the basic principles of scientific writing. A review of your research report by colleagues or other reviewers can bring out points you have missed. They can identify writing that is unclear and poorly organized. Training in research writing must be largely self-training, preferably with the guidance and critique of experienced scientists. As the old adage goes, "The wise man learns from the experience of others, the fool only from his own." Any training, of course, involves much more than merely being "told how." Much practice is required to learn the precepts and to develop a habit of using them (Beveridge, 1957). It is also well to keep in mind that many of the best writers developed that skill from being avid readers.
96
Research Methods and Communication in the Social Sciences Ethics
We have addressed issues of intellectual integrity in faithfully reporting research results, but we have not addressed a worldview of ethics in science. A comprehensive review of ethics is impossible within the limits of this book, but some observations are offered. This section cannot be a treatise on ethics but can alert scientific writers to be sensitive to ethical issues and the minefields they pose in communicating research results. The two major schools of moral philosophy are utilitarianism and moral imperatives. Utilitarians judge actions by whether they add to or detract from the well-being of people. Adherents to the moral imperative school judge actions by whether they agree or disagree with dictates such as "Thou shalt not kill,""Species diversity must be preserved," or "Allsensate animals have the same rights as people." Often, utilitarian and moral imperative schools agree, as in condemning racism or sexism. One school or another often berates scientists for avoiding ethical issues. Many of those critics want "scientific" preaching only if it favors their position, however.
Comment Clarity, completeness, impartiality, order, accuracy, objectivity, simplicity, and ethics are basic requirements in scientific writing. Other characteristics determining readability include appropriateness of the study to the reader; consistency in the use of numbers, names, abbreviations, and symbols; a sense of the dramatic by, say, posing questions to hold the reader's attention; persuasiveness in convincing the reader by evidence; precision supported by exact and appropriate definitions and accurate measurement; and sincerity demonstrated by the quality of frankness and honesty.
A CHECKLIST FOR PREPARING A MANUSCRIPT No set of guidelines can fully define the difference between popular and scientific writing. But we offer a checklist of some characteristics common to scientific and technical writing (see also Bly and Blake, 1982; Day, 1988). 1.
Decide who is your audience and write neither above nor below it.
Guidelines for Communicating Scientific Research 2.
Prepare an outline by going through the mental game of imagining that you will present your research without notes in an oral presentation. That will encourage you to construct a logical sequence that flows easily from one idea to the next. 3. Scientific writing is concise, precise, direct, and logical. 4. Write and rewrite! Writing is hard work. The first draft is for ideas, the second for organization, the third for precision, and the fourth for polish. Then, you're ready for someone else to review your manuscript. After that review, again rewrite your manuscript. Don't hand your major professor or a reviewer a manuscript so badly organized and written one doesn't know where to begin to improve it. Obtain in-house reviews from colleagues before revising and sending the manuscript for journal review. Let the manuscript age for a week between rewrites. 5. Place yourself in the reader's position. You are thoroughly familiar with the methodology, but your reader is not. If it is worth saying at all, it is worth writing so that the reader can comprehend. Include major assumptions and analytical steps in your write-up. Although these may seem so obvious to you that reporting is unnecessary, the steps may be essential in order for the reader to understand your work. But at times it is wise to sacrifice a thimble of clarity for a bucket of readability. 6. Go over each sentence and ask yourself, "Can I defend this thought?" If not, rewrite it. In writing any manuscript you become your own editor every time you change a word, add a phrase, or reorganize a paragraph to clarify problems and arrange the findings in logical order. 7. Keep sentences short and direct. Eliminate unnecessary words and phrases. Words or phrases such as "in order to," "with respect to," "that," "which," "there is," "it is," and "etc." are good candidates for exclusion. 8. Your logic should be relentless, especially within paragraphs. Don't try to pull the reader along by a "logical" chain full of missing links. Change your chain of thought only between paragraphs. 9. Let your manuscript read smoothly and simply. A long series of clauses and modifying phrases is a good way to discourage even the most ardent reader. 10. Say what you are going to do, do it, and then tell what you did. Summarize what you are going to do at the beginning, tell what
97
98
Research Methods and Communication in the Social Sciences
11.
12.
13.
14.
15. 16. 17.
18.
you did at the end. This procedure applies especially to chapters, but lead-in and lead-out sentences of paragraphs also tie a manuscript together. Avoid undue repetition. Words or phrases such as "problem area" (note the undesirable grammar-a noun modifying a noun), "resource use," "elasticity," and "ordinary least squares" become tiresome if repeated often. Ordinarily,firstperson pronouns are not used in scientific writing. Sometimes an "I" or "we" lightens a sentence and allows you to use the active voice. But don't overdo it. And be sure you know to whom "we" refers. "I" is sometimes preferred to the stilted term "the author." When possible, keep the voice active, not passive. When a sentence is written in the active voice, the subject performs the action and fewer words are used. Passive voice is flat because it has no actor. (Active) Example 1: Figure 1 illustrates growth in output. (Passive) Example 2: Growth in output is illustrated in Figure 1. However, be careful what a "chapter" can do: "This chapter analyzes" or "This chapter tests hypotheses" are dubious grammatical forms. Watch the tense! Don't change it in the same sentence or paragraph without good reason. If you have trouble, try using present or past tense throughout rather than becoming confused with future tense in the introduction, present tense in the text, and past tense in the conclusion. Do not underestimate your ability to make mistakes! Check and double-check everything. Avoid being mesmerized by the subconscious drive to overlook your mistakes. Avoid starting sentences with "this" or "these" and "that" or "those" if the reader doesn't know to what these words refer. Qualifiers can rob nouns of their vitality. You occasionally must use qualifiers, but avoid using words such as "may," "probably," "approximately," and "perhaps," "almost," and "at least" unless they are necessary. Reduce the use of rambling sentences, jargon, cliches, and platitudes because they bring ambiguity and distractions that interrupt the flow of thought. Avoid vagueness by using simple, straightforward words or phrases.
Guidelines for Communicating Scientific Research 19.
20.
21.
22.
23.
24.
25.
26.
Avoid unnecessary changes in verb form. "There has been an increase in funds for equipment, more monies to use to take surveys, and additional dollars for the purpose of paying salaries" can be shortened to "Funds increased for equipment, surveys, and salaries." Avoid words and verbose terminology more complex than necessary. Some writers never use a short word where a long word will do. For example, you may be able to replace the word "elasticity" by "response," the word "existential" by "existing," and the word "problematical" by "problem." Avoid digressing into diatribes and areas, especially controversial issues, not central to your research. You may never complete a scientific crusade if you stop to lop off the head of an innocent peasant on the way. Keep your words soft and sweet, because someday you may have to eat them! Remember that the expert selected by the editor to review your manuscript may be the researcher you attacked in your review of literature. Avoid normative terms such as "need to," "ought to," "should," "most," "good," or "bad." You have no special expertise to state that "the nation should pursue food self-sufficiency," or that "research must be done on beef cattle." Avoid writer's block. Perfectionists are especially reluctant to put something on paper. When paralysis sets in because nothing you compose seems acceptable, write anything. By the end of the first page, words will begin to make sense. There will be time later to go back and revise. The main thing is to get something on paper to revise. Bad writing leaves your reader full of questions you should have answered. Go through each sentence of your manuscript asking yourself, "What questions would a reader have about this sentence?" The questions may relate to omitted or included "facts" and "logic" or they may relate to meaning intended by the writer. Edit to answer or dispel those questions. Listen to your writing. Whether you hear the words spoken aloud or silently, listening helps you identify awkward phrases and sentences, lack of clarity, and gaps in logic. Keep the consecutive sections of your document interrelated by transitional devices. Connective phrases, sentences, and paragraphs are the cement that binds together your blocks of writing.
99
100 27.
Research Methods and Communication in the Social Sciences Arrange your material in logical sequence. Begin with ideas the reader can readily understand. If you must present difficult material, go one step at a time. Probably the single best way of arranging most material for easy understanding is what is called the general-to-specific order. This means that you start with a general statement, and then follow up with additional facts that elaborate or explain the statement. Popular articles follow the opposite order, however. Newspaper articles begin with a lead sentence giving the major findings of the analysis. Subsequent paragraphs elaborate on when, where, how, and from whom the findings originated.
A CHECKLIST FOR ENHANCING MANUSCRIPT ACCEPTANCE Few things emotionally devastate an author more than a harsh review and manuscript rejection. The first reaction is likely to be not only despair but also anger. Remember that even the most renowned professional gets plenty of rejection notices. In addition to the points for sound methodology and manuscript preparation already discussed, here are some guidelines for diminishing or coping with manuscript rejection. 1.
2.
3.
Keep manuscripts not only well written but short-10 to 20 double-spaced pages. It is better to prepare four 10-page selfcontained articles out of your research and have two rejected than prepare one catch-all 40-page article and almost certainly have it rejected. Remember that the review process has a big random element. Typically, only one in four articles are accepted. Be philosophical-remember that sometimes you will be unfairly rejected and at other times undeservedly accepted. Write enough articles so that it averages out! Also be persistent; if one journal doesn't accept, another might. But send to only one journal at a time. Select journals carefully. Pick one that favors your topic. General journals usually have the lowest acceptance rate and specialized journals the highest rate. Don't send a highly mathematical article to a journal attempting to reach a popular
Guidelines for Communicating Scientific Research
4.
5.
6.
7. 8.
9.
10.
101
audience. And don't send lots of verbiage to a journal appealing to the mathematical audience in a discipline. Especially if a journal prides itself in being timely, don't submit an outdated article. Know the current issues and respond to them. Get on the mailing list of leading scientists in your field. Professions move so fast today that ideas may be obsolete by the time they are published in professional books and journals. Carefully read the editor's rejection. It may in fact be a request for resubmission after major changes. The editor doesn't want to appear generous because that would encourage you to resubmit a poorly revised manuscript and he would feel obligated to accept it. A little humor helps: A politician says "yes" if he means maybe, "maybe" if he means no, and if he says "no" he's not a politician. An editor says "no" if he means maybe, "maybe" if he means yes, and if he says "yes" he's not an editor! Few reviews come back without requests for changes, often major changes. Address reviewers'comments carefully and forthrightly. Where changes are warranted, make them. Where changes are not warranted, state why. Reviews often take six months or more. Be preparing other articles while you wait for a review. Having several articles in the pipeline helps to maintain your productivity. Some scientists succeed by counting on timely topics and responding to every new scientific fad. Others succeed by researching perennial subjects in ever greater depth. Either approach can be successful. Know and emphasize your strength. Reviewers may be defensive about a manuscript that provides too large a deviation from received wisdom and perhaps even threatens the reigning paradigm. If you are brilliant enough to come up with such material, it may be wise to publish it in less shocking increments or in a book, rather than have to battle reviewers whose hostility is raised by the perceived threat to the discipline. Keep a low personal profile. Boldly identifying yourself in an article is usually detrimental to acceptance, especially if reviewers perceive cockiness. Sometimes hero worship may cause reviewers to accept work from a pillar in the profession. But rejection of well-known authors by reviewers jealous of too much success is an equally likely outcome.
102 11.
Research Methods and Communication in the Social Sciences Establish a time to complete your manuscript and keep it. Your results will have little impact if you miss the teachable moment for the public or the decision deadline of a firm or government. If you must spend 40 hours per week on tasks such as teaching and advising students, the few extra hours you put in at night and on weekends may be critical for promotion, pay, professional reputation, and job mobility.
A CHECKLIST FOR PREPARING AN ORAL PRESENTATION In addition to acquiring technical knowledge and proficiency in research, you must be able to communicate not only in writing but also orally what you are doing. Many opportunities exist for telling people about your research: seminars, committee meetings, oral examinations, thesis or dissertation defenses, and professional meetings at local, regional, national, and international levels (see also Stock, 1985). The following checklist of guidelines can help you prepare and make an oral presentation. 1.
2.
3.
4.
Appearance counts heavily in preliminary judgments of you by professionals who do not know you. It affects how people listen to you, since it affects how serious about your work you are perceived to be. Thus, consider what you are going to wear. It is important that you dress appropriately. You never get a second chance to make a first impression. On the day of the talk, get to the conference room a few minutes early to check the facilities (projection equipment, operation of room lights, microphone, podium and podium light, slide-change device, and pointer) for proper functioning. Then, select the best position of the lectern or spot where you will stand to make the presentation. Make sure that you have all the materials for your presentation. Check ahead of time to ensure that the projected prints, slides, notes, or other presentation materials are properly in place. As with a written presentation, analyze the characteristics of your audience in terms of their general background and comprehension of the subject matter of your presentation. For
Guidelines for Communicating Scientific Research
5.
6.
7.
8.
9.
10.
103
example, a presentation to economists will differ from a presentation to farmers. The audience will need a brief background statement from you to understand who you are and why you are speaking to them. If someone is going to introduce you, prepare appropriate information, neatly typed, and give it to the introducer, preferably sometime before the meeting. Know the purpose of your talk-why you are giving the talk and what you expect the audience to gain from it. Communicate this purpose to the audience in your opening remarks. Is the purpose to inform, to convince, or to entertain? Most people like to be entertained, but few of us are entertainers. Don't tell jokes unless you are adept at it and the humor will not detract from your message. Try to make your presentation of sufficient interest to hold the audience without resorting to gimmicks that distract. Most people can develop a unique and pleasing style over time to improve audience rapport. A good way to begin is to appear confident and to come immediately to the point. Make a case that what you are going to say is important to the audience. Then briefly and broadly outline for the audience the scope and major topics you will cover. You will keep the audience's attention if your presentation addresses their interests. Learn to deal with stage fright, a feeling that virtually all speakers have experienced. It is important to accept it as a natural reaction and to realize that the associated tension is not altogether undesirable. The best way to deal with stage fright is to be well prepared for your presentation. Bolster your selfconfidence by making prior contacts or casual conversation with some members of the audience. Rehearse your talk before a mirror or while mentally visualizing yourself before the audience. Practice the speech extemporaneously several times, speaking from an outline or three-by-five-inch cards. The advantage of extemporaneous delivery is naturalness, spontaneity, and flexibility. Dry run the speech with colleagues or friends matching the audience profile or with a VCR or cassette tape recorder. Ask for critical reactions to your speaking technique and technical content. Maintain eye contact with the audience. It will give your listeners the feeling that you are talking to them directly. Prepare an
104
11.
12.
13.
14.
15.
16.
Research Methods and Communication in the Social Sciences outline of your presentation. The best speakers rely little on notes so they can keep eye contact and appear to be spontaneous. Use notes for main themes, but do not keep looking at your notes. Ordinarily, oral presentations are not read. Presenting your paper extemporaneously gives your presentation some personality of interest to the audience. Some professional societies have strict rules against reading speeches at regular meetings. However, you may be one of the gifted few able to read while holding the attention of the audience. Rhetorical devices can help to explain, clarify, develop, and emphasize principal points. These include (a) repetition in a varied way (but don't overdo it), (b) summations that capsulize longer sections, and (c) examples and illustrations. Pause, raise your voice somewhat, or use other means to highlight critical oral points just as you would use punctuation and underlining in written communication. Speak up, but rhetoric should be more conversational than thundering. Project your voice at the audience. Speak slowly and clearly. Do not mumble or talk to the blackboard. Show enthusiasm about your topic, and talk to the audience rather than to your slides or to your notes. Also, being too far from or too close to a microphone may give irritating electronic feedback. Mannerisms should be appropriate. Keep your head up, smile when you can, and avoid leaning on the podium or against the wall. Don't be stiff, but stand comfortably erect. Try to be relaxed, but moving around excessively distracts. Other kinds of movement such as natural gestures are not at all distracting but instead provide emphasis and variety to your presentation. Occasionally moving away from the lectern serves to stimulate audience interest when it is lagging, but may take you away from the microphone. Unlike the written paper, the length and content of the oral presentation can be adjusted according to the feedback from the audience. Time your practice presentation. It is disconcerting to be half through the presentation to your audience and find your time is up. Have someone remind you when time is up if you (like many people) lose track of time when speaking in public. The primary aim of talking about research is to generate interest and awareness in the listener, not simply to convey information.
Guidelines for Communicating Scientific Research
17. 18.
19.
20.
105
One way to make an oral presentation interesting is to keep it simple. Keep the main points clear to the audience, concentrate on concepts, and eliminate confusing details. Do not present so much information that the audience gets lost in the forest of details and cannot see the important points. Any audience has a limited attention span. Use stretch breaks, a change in pace, jokes, visuals, and other devices to break the monotony. Plan and prepare suitable audiovisual aids (flip charts, blackboards, slides, handouts, overhead transparencies, films, models, or tapes) if they are helpful in presenting data and equations. Such aids can show relationships (tables and graphs), simplify and clarify an idea, involve the audience in the discussion, and help the audience follow an organized plan of presentation. People learn faster through the eyes than the ears, and they remember more if they both hear and see. In planning the use of audiovisual aids, remember that they should be (a) limited in number, (b) relevant to the topic and helpful to understanding (not a parade of unnecessary props competing for the attention of the audience), (c) easy to comprehend (stripped to essentials and adequately labeled), (d) limited to one idea per display, (e) pictorial rather than verbal whenever possible, (f) presented precisely at the moment needed (because earlier is distracting and later is ineffective), (g) removed immediately after discussion to prevent distraction, (h) visible to everyone and highlighted or in large letters as needed, (i) thoroughly discussed by the speaker, and (j) grouped in the body of the talk as a unit to prevent on-again-off-again lighting interruptions. You will need a portable microphone if you need an amplifier when you are moving around to change overhead transparencies, to point at the screen, write on the blackboard, or just to put some life in your presentation. The most common problem with visuals is to include too much in too small print on each display. Finally, summarize your main points at the conclusion of your presentation before you entertain questions and comments from the audience. A well-conceived ending to your presentation makes a strong and lasting impression. A weak ending gives the impression that you have nothing more to say. Present conclusions in a way that invites comments and discussion. Do
106
Research Methods and Communication in the Social Sciences not hesitate to summarize your best points and to show why they are significant in real applications. Your closing can acknowledge the limitations of your results and point out aspects that need further study.
REFERENCES Barrass, Robert. (1978). Scientists Must Write (A Guide to Better Writing for Scientists, Engineers, and Students). New York: Wiley. Bates, Jefferson D. (1985). Writing with Precision. Washington, DC: Acropolis Books. Beveridge, W.I.B. (1957). The Art of Scientific Investigation. Revised edition. New York: Random House. Bly, Robert W., and Gary Blake. (1982). Technical Writing (Structure, Standard, and Style). New York: McGraw-Hill. Bolsky, Morris I. (1988). Better Scientific and Technical Writing. Englewood Cliffs, NJ: Prentice Hall. Collins, Sarah H , and Frederick B. Tuttle, Jr. (1979). Technical and Scientific Writing. The Curriculum Series. Washington, DC: National Education Association. Day, Robert A. (1988). How to Write and Publish a Scientific Paper. Third edition. New York: Oryx Press. Hayes, Robert. (1965). Principles of Technical Writing. London: Addison-Wesley. Ladd, George W. (February 1979). "Artistic Research Tools for Scientific Minds." American Journal of Agricultural Economics 61(1): 1-11. McCloskey, Donald. (1987). The Writing of Economics. New York: Macmillan. Samuels, Marilyn Schauer. (1978). Writing the Research Paper (A Step-by-Step Guide). New York: Amsco College Publications. Stock, Molly. (1985). A Practical Guide to Graduate Research. New York: McGraw-Hill. Strunk, William, Jr., and E. B. White. (1979). The Elements of Style. Third edition. New York: Macmillan. Turabian, Kate L. (1987). A Manual for Writers of Term Papers, Theses, and Dissertations. Third edition. Chicago: University of Chicago Press.
EXERCISES 1.
Discuss places to get help in writing, for example, a technical or scientific writing course.
Guidelines for Communicating Scientific Research
107
2.
Who should be responsible for "cleaning up" the writing of a thesis: The student, the major professor, or someone hired by the student?
3.
Prepare an outline for an oral presentation. student critique it.
4.
Present your research orally to the class. Ask the class to suggest improvements.
5.
Initiate a discussion of memorable oral research or other academic presentations. What made the presentations outstanding?
6.
Suggest ways to brighten up your drawbacks, such as a monotone voice, stiff demeanor, or dull material.
7.
Discuss the advantages and disadvantages of 35-millimeter slides versus overhead transparencies versus paper handouts to assist your oral presentation.
8.
One way to create interest is to introduce a topic, then ask your audience for questions. Build your talk around those questions. Try it.
9.
Join a local Toastmasters or other such club if you need to improve your public speaking. Or visit a meeting of a club and report back to the class what you learned.
10.
You will feel well rewarded for spending a few hours with the brief but elegant books by Donald McCloskey and by William Strunk, Jr., and E. B. White on style in economics. These authors, who have a special ability to make the prose of economics come alive, suggest numerous ways for you to improve your writing. And what those authors do to make the "dismal science" of economics readable can do even more for other disciplines.
Let an instructor or
This page intentionally left blank
5
Guidelines for Reviewing Manuscripts Research follows a series of steps designed to develop and disseminate useful information. Once the analysis is completed, the information generated is prepared in manuscript form for publication. Manuscripts are critically reviewed prior to acceptance for publication for the following reasons (Soil Science Society of America, 1988): 1. To solicit opinions as to the appropriateness of the subject. Original findings suitable for publication are interpreted as the outcome of scholarly inquiry, investigation, or experimentation having as an objective the development of new concepts; the revision, refinement, extension, or verification of existing concepts; the application of existing concepts to new situations; or the development of new or improved techniques in the field of study. 2. To aid in maintaining a high standard of quality in manuscripts accepted for publication. Quality includes such factors as originality of subject or applications, appropriateness of methods, accuracy of mathematical equations and computations, validity of conclusions, organization of subject matter, clarity, and correctness of grammar.
A CHECKLIST FOR REVIEW COMMENTS The author should strive to prepare a manuscript in acceptable form so that only minor alterations will be required after submission for publication and review. Comments of reviewers will help to determine whether a
110
Research Methods and Communication in the Social Sciences
manuscript is acceptable for publication and what revisions are necessary to make it acceptable. The following questions are provided as guidelines for reviewing manuscripts: 1.
2.
3.
4.
5.
6.
7. 8.
9.
10.
What is the nature of the material covered in the manuscript? Is it a new application, a first attempt at empirical application of a method, a report of empirical work, a review or appraisal, or a proposal for action? Does the title describe adequately the subject of the manuscript? Is the title well chosen and intuitive? Can it be understood before the whole text is read? Can you suggest any improvement in wording? Is the abstract informative and concise? Does it tell briefly the reasons for the study, methods used, empirical results, and conclusions? Abstracts are the most widely read section of a paper. Is the manuscript of appropriate length, clearly written, well organized, and adequately documented? Is the outcome sufficiently projected? Does the author write the information in a relatively simple, straightforward manner that can be readily understood by a reasonably competent reader? Is the manuscript technically and logically sound? Is it factually correct? Is the research information presented in a logical order? Does it develop the subject effectively? Is sufficient introductory and summary material given to inform reviewers and readers of the content, importance, and field of applicability of the material covered? Is the problem precisely identified, clearly stated, adequately addressed, and resolved? Is the statement of objectives adequate and appropriate in view of the study? Are the specific objectives clearly stated in a concise and understandable form? Are the objectives sufficiently accomplished? Are the hypotheses properly formulated and adequately tested to meet the stated objectives? Have suitable measurements been performed to test the hypotheses? (Note: Not all research will state hypotheses; some rely on objectives or questions.) Is pertinent literature related to the study properly cited and adequately documented? Does the author give due credit to the
Guidelines for Reviewing Manuscripts
11.
12.
13.
14.
15. 16.
17.
18.
19.
111
relevant contributions of others? Is the number of literature citations inadequate or excessive? Are the methods appropriate for the purpose for which they are used? Do the methods logically follow the problem statement and use appropriate procedures to meet objectives and test hypotheses? Do the methods of analysis meet current scientific standards? Are the methods of analysis suitable for statistical definition of variables, specification of assumptions, formulation of the model, and development of basic mathematical equations? Are all of these items described in sufficient detail to permit a reasonably competent reader to understand and review them accordingly? (Note: Some high quality research is done without formal models or quantitative statistical tests of hypotheses.) Are data relevant and accurately incorporated into the method of analysis? Is data availability a crucial limitation on the procedures? Are the results and analysis consistent with the stated objectives? Are the results over- or understated? Are results built upon the current level of knowledge documented in the review of literature? Are results well interpreted? Are the conclusions adequate? Are they supported by reliable analysis? Are all tables and figures properly presented and referred to in the text? Do statements in the text correspond to the content of the tables and figures? Do the titles of the tables and figures convey the contents? Does the manuscript repeat unnecessarily the published work of other authors? Can the manuscript be shortened without loss of content by condensing the text or reducing the number of tables and figures? Are all of the figures needed if the same information is given also in tabular form? Is there unnecessary duplication in the text? What is the overall scientific merit of the study in terms of importance and potential impact on progress of the scientific field? Does the study advance the scientific frontier? Is the contribution original? Does the study recognize the importance of underlying issues to different audiences (academics, policymakers, society at large)? How well are policy implications drawn?
112 20.
21.
22.
Research Methods and Communication in the Social Sciences Is the general perspective adequate in terms of length of literature review, insightfulness, and depth of author's understanding of research priorities along the frontier? What are the basic limitations of the study? What issues are still open for research? What critical assumptions might usefully be relaxed for future research? Does the manuscript conform to mandated editorial style and format or can it with reasonable effort?
RESPONSIBILITY OF THE REVIEWER The well-known pressure to publish, expressed as "Publish or perish," is relentless. Academic scientists are well aware that their professional reputations, their promotions, and in some cases even their very jobs depend in no small part on their record of publication. Publication records are often scrutinized critically by peers and administrators. Ideally, the author should be the leading authority on his own work and its relation to the general body of scientific knowledge and theory. However, the reviewer who has not been involved in the process of solving the problem and who is not emotionally attached to it is often more able than the author to discern shortcomings of the research. The reviewer who has been engaged in research closely related to the author's field may have data at his disposal that are unknown to the author but that may be of material assistance in the interpretation of the author's own findings. Reviewers are usually selected for their acquaintance with, maturity, and competence in the technical subject matter so that they can offer appropriate constructive review comments and judge whether the manuscript merits publication (Meites, 1975). In fact, the reviewer may have so much experience, knowledge, and eminence in the field that he is actually better able to interpret the findings than the author, and often without the necessity of undertaking prolonged study of background material. The reviewer may render assistance to the author by pointing out errors that the latter may have overlooked or proposing crucial information that would more conclusively bolster the interpretation. Perhaps the reviewer can suggest a totally new interpretation that provides a better explanation not only of the author's findings, but also of other research findings of which the author may not be aware. Thus the reviewer can make a contribution to the
Guidelines for Reviewing Manuscripts
113
scientific literature that is no less real than that of preparing his own manuscript. Being a reviewer or editor often seems to be a thankless, anonymous job. However, for various reasons, an honest review of the manuscript should be given for the benefit of the author and the profession (see also Meites, 1975): 1. A profession issues credentials, awards, and other recognition. In general, it lends legitimacy to its members. Someone entering a profession receives an unwritten social contract. Part of that contract is to review manuscripts. The contract continues when others review your manuscripts. 2. The professional commitment of the reviewer should be sufficient to ensure that the manuscript finally published is of the highest possible quality. Although review of manuscripts yields no remuneration or personal recognition, it is essential in advancing the profession. The review is conducted on the basis of the following review qualities (Soil Science Society of America, 1988): 1. Objectivity. Strive to be objective in your evaluation. Make your evaluation on scientific grounds without bias based on personal reasons or professional jealousies. Objectivity in evaluations is difficult to attain because the manuscripts you receive are on subjects closely related to your own interests and professional work. Gratuitous subjective comments on the significance of contributions are inappropriate and unprofessional. It is always possible that the reviewer may misunderstand the author's assumptions and conclusions, may overlook an important link in the author's chain of evidence or reasoning, or he may be misled by a carelessly phrased portion of the manuscript. The reviewer should be courteous and considerate of an author's efforts no matter how poorly written the manuscript or how inadequate the content. 2. Accuracy. Manuscripts need to be as free of error as possible. Reviewers are not expected to verify the accuracy of all aspects of manuscripts. A good procedure to follow is to verify the points that take little time. An informal sampling of points, checking each for accuracy, is likely to reveal whether or not the author displays a pattern of carelessness. If you are not sure of your understanding,
114
Research Methods and Communication in the Social Sciences
do not hesitate to comment on your uncertainty regarding the point in question. 3. Relevance. Relevance of the problem addressed and subject matter is at issue. Remember that professionals can differ on what they consider to be important work. Chances are you were chosen by the editor to review the manuscript because your professional judgment of what is important is respected. Do not criticize the manuscript for lack of subject matter that you think should be present unless that subject matter is essential to the points covered by the author. 4. Thoroughness. To aid in preparation of your comments, study the entire manuscript intensively to obtain a good understanding of the subject matter. Reviewers are frequently criticized, with good reason, for submitting derogatory and even discourteous and sarcastic comments based on superficial reading and inadequate understanding. Such comments discredit the review process. Ask help of a colleague or other expert in reviewing a portion of the manuscript you don't understand. 5. Explicitness. Your comments should be explicit. For example, if you think the organization is poor or the methodology is not appropriate, your comments will be most valuable if you explicitly explain why. 6. Helpfulness. Your comments should help the author publish work of high quality. If, in your opinion, the manuscript is deficient in one or more aspects of quality, your review comments should help the author eliminate the deficiencies. If, in your opinion, the scientific content cannot be brought to acceptable quality, you will be of service to the author by advising that such is the case. Few things are more emotionally devastating, especially to a young author, than an unfavorable review. The first reaction is likely to be anger as well as anguish. It is helpful to remember that the hurt will diminish with time. It is also important to remember that reviewers, like other persons, are sometimes biased, careless, and even downright wrong. Although the reviewers' evaluations are usually helpful, as an author you should review all of their comments with great care. Negative comments can be highly constructive when they reveal the weaknesses in your manuscript and indicate how you can strengthen your final version. After revision to respond to weaknesses, resubmit the manuscript elsewhere if appropriate. Keep in mind that a manuscript should be submitted to only one journal at a time.
Guidelines for Reviewing Manuscripts
115
REFERENCES Meites, Louis. (1975). "Technical Editing: The Reviewer's Viewpoint." Pp. 103-114 in B. H. Weil, ed., Technical Editing. Westport, CT: Greenwood Press. Soil Science Society of America. (1988). "The Reviewers Guide." SSSA Journal 52:1-2.
EXERCISES 1. Critically review research proposals or professional manuscripts about to be submitted for publication. Suggest ways the manuscripts could be improved. 2. Discuss the ethics of someone (e.g., a major professor) editing the manuscript of another. At what point does it become unethical to substitute the words and thoughts of the editor or major professor for those of the author? 3. Ask a professor to let you help him or her review an article some editor has asked him or her to jury (referee) for possible publication in a journal. Compare your review with the professor's review. 4. The editor of a journal needs to be able to sort out hostile and incompetent reviews. What qualifications and background would you like to see in an editor? 5. Is it wise to become an editor or agree frequently to serve as a reviewer before you are tenured?
This page intentionally left blank
6 Guidelines for Library Literature Search Computer systems in the past decade have begun to profoundly affect and reshape the process of recording, transmitting, storing, and retrieving information. Computerized information retrieval makes it relatively easy to pull relevant citations from databases, but retrieving information in the most effective manner requires a thorough understanding of how library information is organized. College and university libraries continually revise the way they provide services. This condition results from the information explosion and technology revolution. The amount of published material is increasing rapidly, and libraries are losing ground in the proportion of new information they are able to acquire. In addition to being available in print form, information is found in a variety of other formats, including audio cassettes and video discs. The manner in which libraries catalog, store, and retrieve information has changed dramatically in recent years. These changes result from at least five important developments in library automation. First, development of the Machine-Readable Cataloging (MARC) system by the Library of Congress enables libraries to create machine-readable bibliographic files. Second, establishment of computer library networks such as the Research Libraries Information Networks (RLIN) permits libraries to use machine-readable cataloging done by other libraries, as well as to submit their own cataloging to the networks. Third, automation of certain key library functions (for example, cataloging and bibliographic control, interlibrary loans, acquisitions, and circulation) has changed the organizational structure at many leading academic and public libraries. Fourth, development of on-line public access catalogs (OPAC) enables library users to obtain bibliographic information without resorting to the time-consuming, manual card catalog. Finally, creation of local area computer networks (LACN) permits library users to
118
Research Methods and Communication in the Social Sciences
access library information without having to be physically present in the library. The user's personal computer is linked to the information network via a modem or direct line. These developments, combined with growing information resources, provide library users with access to more information and enable them to make more effective use of library resources (Neway, 1985). The focus in this chapter is on the subject matter and structure of information in broad fields rather than on the selection or operation of specific retrieval systems for which information abounds elsewhere. The unique differences between various search tools and databases in dozens of research fields cannot be covered comprehensively. This work is principally intended for the research worker or graduate student who is generally familiar with research in a given field but who may not know how to approach the development of search strategies that use computer literature searching systems. The section is not directed toward any particular system. The search facilities are described in a general fashion.
MANUAL LITERATURE SEARCH Before turning to the more specific techniques or tools for accessing information, one needs to consider and know the principal resources of the library, understand its literature classification system, and be able to find the shortest route to the information that it contains. One basic rule is to be self-sufficient and learn to use the library by actual experience. Nothing will take the place of actual hours of exploration in the library, getting acquainted firsthand with its particular characteristics and the nature of its holdings. Every library is different in terms of its emphases and unique characteristics. A basic requirement in using any library is to know its layout and explore it from one end to the other. Some libraries have manuals that give the necessary information about the general physical features of the library, the specific location of various holdings and special collections, the methods of classification, and the main tools for accessing information (Leedy, 1974). Manual literature-search techniques rely on the structure of printed retrieval tools. Printed indexes may take many forms. Most printed indexes are hierarchial, presenting citations in a logically arranged series of subject headings and related topical subdivisions. Users of such a work look under those headings that appear to describe most aptly the area of investigation. Having located the appropriate heading, they then read through the titles appearing under that heading to identify the ones they wish to use for their research. Although hierarchial indexes are particularly good in providing access to large volumes of literature and desired references under a carefully controlled
Guidelines for Library Literature Search
119
set of categories, such indexes are usually slow to reflect changes in terminology when new research areas emerge. Consequently, the user working on a new area of specialization not yet introduced into the list of subject headings has to seek information under headings not quite on target. The shortcomings of hierarchial indexes have been overcome by other approaches to indexing such as computerized rotations of keywords and/or title words that provide subject access. This approach enables the reader to search for articles that contain specific words or phrases in the title or in the strings of index terms. It displays terms in a way that enables the user to check the context in which the desired term appears (Gilreath, 1984). Coordinate indexing is another type of indexing scheme adaptable for both card files and printed retrieval aids. It is the indexing principle upon which computer literature searching systems are based. An index based on coordinate principles first identifies each bibliographic entry by means of some unique number or letter-number tag. Indexes are built for each significant word by listing together document numbers for every item containing that word. These document number lists that constitute the index's access points are then filed or printed in alphabetic or numeric sequence. When users desire to conduct a literature search, they consult those portions of the index containing the terms in which they are interested. By comparing specific item numbers, they are led back to those items in the sequential listing of bibliographic entries. The users look up the items identified and select specific items that are likely to be of potential interest (Gilreath, 1984).
COMPUTER LITERATURE SEARCH Similarities are many between computer or on-line searching and the manual searching of the conventional printed tools. Both processes involve the matching of words that describe the subject of the search against descriptors (entries) used to index the collection of references. Coordinate indexes in printed form can be extremely helpful in library searches, but they quickly lose their usefulness when the number of items to be indexed or terms to be coordinated becomes large. When these limits are reached, the manual task of comparing large lists of document tags or manipulating a dozen or more search terms simultaneously is burdensome. Computers, in contrast, are ideally suited for performing these tasks. It is precisely this set of operations that forms the basis of today's on-line retrieval systems. In a matter of seconds the computer transcribes hundreds of document tags into the active work space of a user and, when instructed, quickly compares any set of tags with any number of other sets of tags to identify references meeting the reader's needs. With a computer-based retrieval system, a library researcher can easily search indexes using a profile of a
120
Research Methods and Communication in the Social Sciences
dozen or more words and phrases that are to be grouped together in several ways. The output from such a search can quickly provide the citations the researcher could have found with a tedious manual search, and also a number of references that are on target but may have been missed or indexed under a heading the reader may not have thought to consult. The power of the computer to search large volumes of literature rapidly and thoroughly encourages researchers to query indexes and abstracts they would not normally search with manual search techniques (Gilreath, 1984). On-line searching occurs at a computer terminal. .The process starts with a searcher choosing sufficiently narrowed and focused terms describing his topic. The selected terms are then input into the on-line searching system. Searching is performed by looking for a match between search terms and terms stored in the description of the journal article or other literature. At each stage, the number of items referenced under a selected search description is printed. The user can modify the search by combining items or by trying new terms until a reasonable number of items has been selected from the database. The selected items can then be printed either on-line at the terminal, when there is a small volume of printed material, or off-line at the site of the computer, from where they are dispatched to the user. The selected items can usually be printed in a variety of ways: in a short format, such as an abstract number or title; as a full bibliographic reference with author, title, journal reference, classification codes, indexing terms; or as a full bibliographic reference plus an abstract (Henry et al., 1980). The computer, of course, has to be instructed with the specific system command language. The system command language is usually specific to a particular on-line service. The field of on-line literature searching continues to evolve with new system operations, databases, search facilities, and communication networks. One of the most valuable features of on-line information retrieval is the high degree of accessibility. With little more than a telephone and a terminal, it is possible to link into a range of computer systems spanning or reaching half the world and covering nearly all the major bibliographic databases (Henry et al., 1980). On-line searching strategy cannot be successful without the effective use of the relevant library manuals or search aids which are continuously revised. The best sources of current information and new developments, particularly with regard to library literature search techniques, are the user manuals and newsletters issued by the library or information center. Before turning to computer databases and search strategies for specific fields of research, it is helpful to review how the computer search service is administered in one's own institution. In most institutions, staff members in the library or information center act as the interface between computer search services and researchers. The search analyst is an information specialist or librarian with training in the construction and execution of strategies to retrieve information via computer. In some cases search
Guidelines for Library Literature Search
121
specialists may have extensive subject background in the areas to be searched. Because of their general familiarity with the content and structure of reference tools and their general expert knowledge of system protocols gained through frequent use, search analysts often can provide more thorough and cost-effective searches than end users can provide for themselves through direct searching of files on-line. Because searches frequently are conducted by intermediaries (search analysts), a special partnership should be developed between the requesters (researchers) and the providers of computer literature searches. The person with the original need for information, the requester, who may be a graduate student or research worker, realizes a need for information, and approaches the library or information center. The requester or the information staff may decide that on-line search is an appropriate method. Before the search strategy is initiated, the requester should do some initial research in the literature of his discipline and should provide full details of the substance of the inquiry (or question). Details start with a brief summary of the search topic; a list of the words or phrases normally used to describe the search topic; details of several recent relevant references and the number of references desired; current terminology and knowledge about the field to be searched; any limitations by date, language, and cost; and the urgency and confidentiality of the search. The words used in the search will determine the number of citations. Suppose one is reviewing literature on the income elasticity of demand for food at the retail level. Use of the search word food alone would produce tens of thousands of citations and would not be helpful. Adding the word demand would cut the citations to a few thousand, still too many. Adding the word elasticity would narrow the citation to (say) 1,000 and eliminate much irrelevant literature. Adding the word income would make a meaningful list of perhaps 100 publications. If greater specificity is required because of specific interest in theory, demand systems, time series, or meat demand, these words could be added to reduce citations to perhaps a dozen. A search analyst trained in library or information science can bring to the process a knowledge of the content and structure of information retrieval systems, familiarity with indexing vocabularies, and skill in constructing and executing on-line search strategies. Once the topic to be searched is explained and the databases to be searched are selected, the search analyst can run the specific program and provide the requester with the desired results of the search (see Gilreath, 1984; Henry et al., 1980). Hundreds of databases are available for searching. The library has access to these specialized databases, which offer the advantage of quick, electronic searching. Databases are available inmost scientific and technical disciplines and also, increasingly, in various social sciences and humanities areas. In addition, many machine-readable files of numerical information (databanks) and information systems contain the full text of documents for retrieval
122
Research Methods and Communication in the Social Sciences
(Henry et al., 1980). These electronic searches give access to an enormous range of research sources, and so can turn up material that one might overlook in more routine search. In a computer search, librarians employ special software to retrieve information from databases at remote locations. The databases are produced by commercial firms, government agencies, professional associations, and other institutions. Many of the databases are the electronic equivalents of familiar printed indexes. They contain citations or references to books, journals, and other published works. Some include abstracts or summaries of articles along with the citations. The result of an extensive search of this kind of database is a customized bibliography on any imaginable topic. In addition, source databases contain the full text of articles, lists of names and addresses, and statistical and other numerical data. A search of this type of database can retrieve facts and data to answer questions on many research topics. The improvement in library research and the availability of specialized databases made possible by computer systems are not without cost. Of course, manual search also entails the cost of the searcher's time, which must be balanced against direct costs of on-line computer searches. Costs for computer literature search services conducted at regular rates vary depending on the databases searched, the interaction and amount of time the terminal is physically connected to the databases, and the number of references printed. Libraries or information centers usually charge fees for the service based on hourly rates that vary from one institution to another and change from time to time. Cost decisions may weigh heavily in the choice of method of conducting library search. In general, the main consideration in deciding whether to use computer systems will be whether the value of the information exceeds the cost of obtaining it. However, the power of computer systems is evident in terms of the extensive number of references for a chosen search definition, the convenience of having a printed list of the references, the typical capability and speed of accessing directly different parts of a database, and the alleviation of much of the serial searching necessary when a printed index is used. The databases are usually designed for easy use. No special knowledge of computers is required. Generally, the necessary instructions are displayed on the computer screen. In addition, a variety of printed instructional materials are available for many of the databases.
ELECTRONIC LITERATURE SEARCH DATABASES Three database examples are briefly reviewed here: EconLit, ABI/INFORM, and AGRICOLA (Dowling and Leonard, 1992;SilverPlatter
Guidelines for Library Literature Search
123
Information, Inc., 1991-91; University Microfilms International, 1990). EconLit is a comprehensive, indexed bibliography with selected abstracts of economic literature. It covers over 300 major economic journals and collective volumes as well as books and dissertations on economics. This database is compiled and published by the American Economic Association and is updated quarterly. ABI/INFORM is an electronic database covering both theoretical and practical business topics. This source contains detailed abstracts and bibliographies for more than 800 North American, European, Asian, and Australian journals and is updated monthly. AGRICOLA (AGRICultural OnLine Access) is an electronic bibliographic database covering agricultural literature published in journals, monographs, technical reports, theses and dissertations, patents, and software. AGRICOLA is published by the U.S. Department of Agriculture. General procedures for using these electronic database systems are relatively simple. Most databases, including the three listed above, are operated by a menu-driven system. When the user begins to conduct the search, the database title screen (or the database selection screen, depending on how your computer terminal is set up) will appear on the monitor. The user can then move through the menu to the point where the search can actually be started. The systems require the user to provide a search statement or word (i.e.,market system, biotechnology, etc.). The system then retrieves the listings that contain the statement or word. The user can browse through the listings and print or download to a diskette any references that may be of use. Most of the menu systems provide excellent step-by-step instructions to the user on how to proceed and perform specific functions.
NOTE Cynthia Dishon and Shiva Makki are coauthors with Tesfa Gebremedhin and Luther Tweeten of this chapter.
REFERENCES Dowling, Carol L., and Thor Lehnert, compilers and editors. (January 1992). List of Journals Indexed in AGRICOLA 1992. Washington, DC: Technical Services Division, National Agriculture Library, U.S. Department of Agriculture. Gilreath, Charles I. (1984). Computerized Literature Searching (Research Strategies and Databases). Boulder, CO: Westview Press.
124
Research Methods and Communication in the Social Sciences
Henry, W. M., J. A. Leigh, L. A. Tedd, and P. W. Williams. (1980). Online Searching: An Introduction. London: Butterworths. Leedy, Paul D. (1974). Practical Research: Planning and Design. New York: Macmillan. Neway, Julie M. (1985). Information Specialists as Team Players in the Research Process. London: Greenwood Press. SilverPlatter Information, Inc. (1991, 1992). SPIRS User's Manual SilverPlatter Version 3.1. Norwood, MA: SilverPlatter Information, Inc. University Microfilms International. (1990). ABI/INFORM Ondisc User's Guide for Version 1.5. Ann Arbor, MI: University Microfilms International.
APPENDIXES Appendix 1 contains an illustration of "Proposal Format Guidelines" from a Proposal Kit by the Northeast Regional Center for Rural Development (1991), and Appendix 3 contains an illustration of "Essentials of a Project Outline" from the Hatch and Mclntire-Stennis administration manuals of the Cooperative State Research Service of the U.S. Department of Agriculture (Barr, 1984). Based on these samples of specific proposal guidelines, two sample research proposals entitled "An Economic Analysis of Alternative Sources of Household Income for Farm Families" (Appendix 2) and "An Analysis of Rural Health Services in Louisiana" (Appendix 4) are included for illustration.
This page intentionally left blank
Appendix 1 Proposal Format Guidelines of the Northeast Regional Center for Rural Development (NERCRD) TITLE PAGE 1. 2. 3. 4.
Title: A brief, clear, specific statement of the project. Duration of project, for example, July 1, 1990- June 30, 1991. Names of principal investigator(s) and host institution(s). Abstract or summary (maximum of 1/2 page).
PROBLEM STATEMENT What is the problem? Why is it a problem for the Northeast region? How is the problem related to rural development? Is this problem a research/extension priority for NERCRD? Why is support from the Northeast Regional Center for Rural Development appropriate for this kind of project?
RELATED CURRENT AND PREVIOUS WORK What evidence/documentation is there to support the problem statement? Include a brief literature review which demonstrates your knowledge of the field and indicates where gaps exist. How will the outcome of this project help meet the needs/priorities of NERCRD?
128
Appendix 1
PROGRAM PROPOSED Purpose. What is the stated purpose of your program? What are your objectives? Objectives should be clear, concise, and measurable. Approach. What approach will be used (gathering of primary or secondary data, method of analysis of existing data, workshop, conference, or other)? Describe working plans and methods to attain your objectives. Anticipated outcome. Research paper, conference, conference proceedings, extension materials with regional usefulness. Anticipated audience. Research, extension professionals, both, other.
HUMAN SUBJECT FORM If your proposal involves research on human subjects, include a verification of human subjects review from your host institution. After the proposal is evaluated at NERCRD, if further human subject review is deemed necessary, you will be notified about any additional procedures.
PERSONNEL INVOLVED For each member of the project team, specify the following four items: institution, personnel involved, time required, and responsibilities.
BUDGET Use enclosed form. Include the in-kind costs (salaries, secretary costs, etc.) under "matching institutions" as appropriate. NERCRD looks favorably on proposals which indicate support from the host institutions and other funding sources.
REFERENCES Cite relevant research and extension sources used in the Problem Statement and Related Current and Previous Work sections.
Appendix 1
129
TIMETABLE The timetable should display the projected tasks and who will perform them over the duration of the project. Interim and final report points should be shown.
APPROVAL Include letters of administrative approval from each institution involved with the proposal. Include an acknowledgment waiving all overhead costs.
VITAE Include short vitae for project participants. List addresses and telephone numbers.
REFERENCE Northeast Regional Center for Rural Development (NERCRD). (1991). Proposal Kit. University Park, PA: Pennsylvania State University.
This page intentionally left blank
Appendix 2 Sample Research Proposal 1: An Economic Analysis of Alternative Sources of Household Income for Farm Families January 1992 - December 1992
Tesfa G Gebremedhin, Ph.D. Associate Professor of Agricultural Economics Division of Resource Management College of Agriculture and Forestry West Virginia University Morgantown, West Virginia
Daniel A. Lass, Ph.D. Associate Professor of Agricultural Economics Department of Resource Economics University of Massachusetts Amherst, Massachusetts
132
Appendix 2
ABSTRACT Agriculture is no longer the principal source of family income for many small-scale farm families. In most cases farm income now serves as a supplement to the income from another job rather than the off-farm job supplementing farm income. As agricultural employment is reduced by technological and economic changes, the demand will rise for employment opportunities for farm families in the nonfarm sector. Farm families may be willing to supply more time to off-farm work but they may be constrained by the lack of off-farm work opportunities. Knowledge of alternative sources of income and characteristics of local labor skills can help rural and agricultural development programs to realize the full economic and social vitality of farm families and to strengthen the linkages between farm and nonfarm sectors.
PROBLEM STATEMENT Over the years, rural areas have become less dependent on agriculture as the principal source of family income. In 1960, over 42 percent of farm family income came from nonfarm sources. Today, more than 72 percent of farm family income comes from off-farm employment or nonlabor income (Leuning and Jones, 1989;Gunter and McNamara, 1988). Farm families are more likely than any other occupational class to hold two or more jobs and may participate full-time or part-time in nonfarm labor markets (Ahearn, 1986). Clearly, income from nonfarm sources is critical to the financial wellbeing of farm households and has a significant impact on the economic vitality of rural communities. This indicates that the health of the local nonfarm economy influences the economic vitality of local family farms. However, it is reasonable to expect regional or local differences in the effects of nonfarm labor income. These differences may be due to regional or local variations in the proximity and types of nonfarm employment available and differences in the character and structure of agriculture among the states. Small-scale, low farm income operations are the most dominate farms surrounding many rural communities in the Northeast region. Families operating small-scale farms depend on off-farm employment as a major source of family income more than do other farm operators. On the average, in the Northeast region, over one-half of the farm operators in 1987 were part-time farmers who have worked off-farm, with the highest proportion in New Hampshire (61%) and lowest in Vermont (46%) in 1987 (Mukherjee, Brucker, and Hastings, 1990). It is likely that part-time farming is a
Appendix 2
133
permanent way of life in the Northeast region, not a short-term response to farm financial problems. Consequently, farming has become a secondary occupation for the majority of these households. The farm may serve as a supplement to off-farm income rather than the converse as commonly assumed. Despite the relative meager earnings from the farm, small-scale farm households are in many ways positive and substantial contributors to both the agribusiness and consumer industries of rural communities. However, the general trends in agriculture of increasing farm size and decreasing farm numbers are forcing many farm families to enlarge their operations, get out of farming, and/or seek off-farm work (Gladwin and Zabawa, 1985). As agricultural employment is reduced by technological and economic changes, supply of labor by farm families to nonfarm sectors will expand. Emerging modern technologies in various nonfarm sectors have increased the demand for high-skilled labor and decreased the demand for less-skilled labor. As a result, farm families with limited job skills may find that economic development leaves them in a comparatively worse economic position than before because off-farm employment opportunities are relatively scarce. Farm families willing to supply labor to off-farm markets may be unable to do so because of their lack of appropriate training or skills. Farm families who do participate may also be dissatisfied with their levels of employment due to lack of appropriate skills. These constraints on participation contribute to underemployment of farm families. Types of nonfarm jobs held and levels of employment of farm families can be expected to differ by educational background of the families and characteristics of local labor markets. The human resource is essential for the economic development of rural areas. The economic vitality of rural communities depends on the availability of a high-quality work force. Higher-level education in a rural area enhances the population's chance of meeting economic challenges, knowing how to create or attract new jobs, and competing for higher paying jobs. Farmers may need more training if they are to match skills required by the nonfarm sector. Better understanding of factors that influence off-farm earning capacity is important from a state, regional, and federal policy perspective in designing human capital development programs for community development. Neither off-farm employment nor the structure of the rural economy can be understood apart from the linkages between agriculture and the rest of the community (Kraybill and Johnson, 1986).
134
Appendix 2
RELATED CURRENT AND PREVIOUS WORK A search was completed using the Current Research Information System (CRIS) of ongoing and recently completed research related to the general topic of this study. Review of the literature revealed a series of research studies conducted on the extent and characteristics of off-farm work in the Northeast region by Findeis (1985); Findeis and Reddy (1987); Findeis, Hallberg, and Lass (1987); Hallberg, Findeis, and Lass (1987); Lass (1988); and Lass, Findeis, and Hallberg (1989). These studies indicated that the economies of most rural towns and communities were transformed by offfarm employment opportunities, first in manufacturing and mining and more recently in services. In addition, Larson (1976), Huffman (1980), Summer (1982), Simpson and Kapitany (1983), and Huffman and Lange (1989) examined factors affecting off-farm decisions of farm operators. Coughenour and Christenson (1983) observed that the changing structure of U.S. agriculture is moving toward a dual system, with increasing numbers of large farms (gross farm sales of $200,000 or more annually) at one end of the spectrum and increasing numbers of part-time small-scale farms with heavy dependence on off-farm employment (gross farm sales of less than $40,000 annually) at the other end. The number of medium-sized farms grossing between $40,000 and $200,000 per year is declining. This is not a temporary phenomenon, but a situation in which the economic and natural resource base of farming and rural communities will be changing constantly. One result of this trend is the possible emergence of large commercial farms in traditionally strong agricultural areas and part-time farm operations near off-farm employment centers (Gebremedhin, 1988). A study by Heady and Sonka (1974) showed that rural area employment and community income are likely to be reduced as farm size increases. Since income is an important factor in consumer demand, an agricultural sector characterized by small and part-time farms is expected to have greater demand for consumer goods and services than would be the case with a greater proportion of large farms. Coughenour and Christenson (1983) found part-time farmers and nonfarmers to be more supportive of community growth and economic development than large-scale, full-time farmers. Parttime farmers are linked to the rural economy through the labor market and serve as a stable source of labor for attracting new business ventures. Effects of off-farm employment are of vital concern to the community because of labor, input, and product market linkages of agriculture to the rest of the economy (Carlin and Saupe, 1990). It is evident that the economic viability and future development prospects of small towns and rural
Appendix 2
135
communities depend on the availability of skilled workers able to compete successfully with urban areas for off-farm jobs and business ventures. Deavers (1990) indicated that through most of the 1980s rural areas have not been competitive, especially as measured by employment change, income level, and population retention. Programs that seem most likely to induce rural development include education, training, local leadership development, and entrepreneurship. Even if such programs fail to stimulate local growth and development, they have inherent value to the rural community. Strong basic educational skills impart quality human resources that are vital to the success of local economic and community development and personal earnings. It is important to determine the policies and programs that can facilitate the option of combining farm and off-farm work. What problems do farm families encounter in securing nonfarm employment? What types of marketable skills do farm families have available for off-farm jobs? What kinds of education and training are needed for off-farm employment opportunities? Obviously, if rural communities cannot provide off-farm work, unemployment or underemployment could result. Thus, this study differs from related research due to the increased emphasis on the available skills of farm families and the effects of local labor market characteristics on offfarm earnings.
PROGRAM PROPOSED Purpose The main purpose of the study is to develop socioeconomic information on levels and sources of off-farm income of farm families. The purpose also is to identify measures of off-farm participation and supply as well as the existence of underemployed family labor resources. The results will be useful to research and extension programs devoted to improving skills and identifying sources of off-farm income for part-time farm families.
Objectives The overall objective of the study is to conduct a socioeconomic analysis of off-farm employment and income for farm families. Specific objectives are to:
136
Appendix 2
1. Assess the types of skills farm families have available for off-farm employment opportunities; 2. Determine the effects of socioeconomic factors on off-farm employment participation, supply, and income; 3. Assess levels of underemployment of farm family members in offfarm employment; and 4. Make recommendations from the research findings for viable policy alternatives to improve the economic well-being of farm families.
Approach Survey data will be collected to better understand the promise and problems brought by technological and economic changes to farm families securing off-farm jobs in West Virginia and Massachusetts. The data collection will be part of the Farm Costs and Returns Survey (FCRS) of the Economic Research Service, U.S. Department of Agriculture. Cooperation with researchers of the Farm Sector Financial Analysis Branch, Agricultural and Rural Economy Division of the Economic Research Service, has resulted in several questions important to this study being added to the FCRS. The questions will be structured such that farm family skill levels can be identified as well as off-farm employment, hours worked, and relevant socioeconomic characteristics of the farm families. Questions will be included to evaluate the levels of underemployment for farm families. These questions will be designed to assess the occurrence of underemployment due to lack of employment opportunities or lack of required skills. Additional attitudinal questions on policy alternatives with respect to human capital development programs will also be included. Farm families will be asked if they are willing to participate in skill development programs and to indicate in what programs they are willing to participate. The FCRS is a national survey instrument employed for all states including the Northeast states. The project will compare the characteristics of labor skills and types and sources of off-farm income of farm families in West Virginia, Massachusetts, and other Northeast states. Secondary data will be collected to augment the database and strengthen the statistical analysis. To accomplish the stated objectives, descriptive statistics will be presented and the distribution of nonfarm income of farm families analyzed to identify major sources of total family income. Econometric analysis will be used to determine the effects of various socioeconomic characteristics on off-farm employment decisions. It is well known that off-farm supply functions are
Appendix 2
137
conditional upon the individual's decision to participate in the off-farm market. Recently, research has addressed the issue of joint participation decisions by farm operators and spouses (Lass, Findeis, and Hallberg, 1989; Lass and Gempesaw, 1992) and farm husbands and wives (Huffman and Lange). This research has shown that the structure of an individual's offfarm supply function also depends upon the participation decision of the individual's spouse. These models can be estimated using maximum likelihood techniques or by the two-stage least squares approach of Heckman (1976). The participating decisions cause data on off-farm labor supply to be truncated from below. Off-farm participants may also be constrained from above. They may not be able to find as many hours of work as they desire; as a result, the individual is underemployed. The attitudinal questions will indicate when an individual would prefer to work more hours but faces a constraint. This initial measure of underemployment will be captured by a binary dependent variable. Probabilities of underemployment by farm family members will be modeled using appropriate maximum likelihood techniques (i.e.,probit or logit models). Given that the individual is underemployed, the level of underemployment will be estimated using attitudinal questions about the number of hours the individual would prefer to work. We will be able to use these models to determine socioeconomic characteristics that affect levels of underemployment. We also propose to investigate wages of farm family members who are underemployed to determine whether these individuals also receive lower wages than expected given their stock of human capital. Finally, off-farm labor supply functions are conditional upon decisions by individuals to participate in the labor force and to be underemployed. To accurately estimate supply responses for farm family members, it is important to consider both participation and underemployment constraints. Estimation of participation and supply decisions will allow us to identify factors that affect the off-farm employment income of the farm family. Independent variables in the off-farm labor participation and supply models can be separated into five categories. The first category includes personal characteristics of the individual such as age, sex, and measures of human capital. The effects of these factors on participation and supply are well documented in the literature (see Findeis, Hallberg, and Lass, 1987,and Lass, 1988, for reviews). A second category of independent variables includes characteristics of other family members. Family size and human capital measures of other family members have been found to be important to both participation and supply decisions. Exogenous nonlabor income of the
138
Appendix 2
household is also a factor in both participation and supply decisions. Theoretical work (Huffman, 1980; Sumner, 1982) has established the importance of a fourth category of variables, farm characteristics. Most previous empirical studies included variables such as measures of farm size and farm type to characterize the farm. Use of the FCRS data provides an exciting possibility of estimating profits or cjuasi-rents and including predicted values in the participation and supply functions. Such an empirical procedure is consistent with the theoretical models of off-farm labor supply. The final set of independent variables will be comprised of characteristics of the local economy. Distance variables, either to the nearest town or actual commuting mileage, are typically used with mixed results. This research will attempt to collect measures of the health of the local economy such as employment data by industrial classifications and unemployment rates. Models of underemployment will primarily focus on the impacts of human capital measures and local economic conditions on the probability of underemployment and the level of underemployment. The results will allow us to classify individuals with a high probability of off-farm employment as well as individuals with a high probability of underemployment. Such information will be useful to local policymakers in assessing the existing pool of labor in their areas and the skills available in that labor pool. Results will suggest the types of individuals who are lacking necessary skills for the local labor market and the types of employment training programs that would provide positive effect on the local economy.
Anticipated Outcome and Audience Results of the study will be published as a research report or bulletin. Papers will be prepared for presentation at professional association annual meetings and for publication in professional journals. The research publications will be organized for use by state and county extension offices and other agricultural related agencies. For example, models will be of use in predicting an individual's success in obtaining off-farm employment given constraints such as human capital stock and local labor market conditions. State policymakers and local leaders will have a clearer understanding of what type of economic sectors are most important in employment creation and in influencing economic development success in rural areas. Thus, the research will provide information useful in designing appropriate rural development programs that focus on human capital development and rural industrialization. Such programs affect the availability of off-farm
Appendix 2
139
employment opportunities for farm families to increase their income levels and to remain in their rural communities.
INSTITUTIONS AND PERSONNEL INVOLVED The study will be conducted by the Division of Resource Management, West Virginia University and the Department of Resource Economics, University of Massachusetts. External cooperation and assistance will be sought primarily from the State Agricultural Statistics Office. Dr. Tesfa G. Gebremedhin, associate professor in the Division of Resource Management West Virginia University, and Dr. Daniel A. Lass, associate professor in the Department of Resource Economics, University of Massachusetts, will be the principal investigators of the project (resumes attached). A graduate research assistant at each institution will be included to assist in data collection and completion of the study.
REFERENCES Ahearn, Mary. (July 28, 1986). "Off-Farm Income and Labor Allocation of Farm Operator Households." Paper presented at the American Agricultural Economics Association annual meeting in Reno, Nevada. Carlin, Thomas A., and William E. Saupe. (March 1990). Structural Change in Farming and Its Relationship to Rural Communities. Staff Paper Series No 316. Madison: Department of Agricultural Economics, University of Wisconsin. Coughenour, Milton, and James A. Christenson. (1983). "Farm Structure, Social Class, and Farmers' Policy Perspectives." In David E. Brewster, Wayne D. Rasmussen, and Garth Youngberg, eds., Farms in Transition. Ames: Iowa State University Press. Deavers, Kenneth C. (April 1980). "Rural Development: Income Strategies for Farm and Non-Farm People." In R. G. F. Spitze and A. H. Silvis, eds., Agricultural and Food Policy Issues for the 1990s, USDA/ERS/ARED. No. 4. Findeis, Jill. (1985). "The Growing Importance of Off-farm Income." Farm Economics. University Park: Pennsylvania Cooperative Extension Service. Findeis, Jill L., M. C. Hallberg, and Daniel Lass. (August 1987). Off-Farm Employment: Research and Issues. Staff Paper No. 146. University Park: Department of Agricultural Economics and Rural Sociology, Pennsylvania State University. Findeis, Jill, and Venkateshwar Reddy. (August 1987). "Regional Aspects of Offfarm Income: Distributional Issues and Impacts." Paper presented at the American Agricultural Economics Association annual meeting in East Lansing, Michigan.
140
Appendix 2
Gebremedhin, Tesfa G. (July 1988). "Assessing the Impacts of Technology on Southern Agriculture and Rural Communities." Southern Journal of Agricultural Economics 20:45-52. Glad win, C. H., and R. Zabawa. (1985). "After Structural Change Are Part-Time or Full-Time Farmers Better Off?" Gainesville: Food and Resource Economics Department, IFAS/University of Florida. Gunter, Lewell, and Kevin McNamara. (January 1988). "The Impact of Local Labor Market Characteristics on the Off-Farm Earnings of Farm Operators." FS 88-16. Athens: Division of Agricultural Economics, College of Agriculture, University of Georgia. Hallberg, M. C, J. L. Findeis, and Daniel Lass. (September 1987). "Part-Time Farming In Pennsylvania and Massachusetts: Survey Results." A.E. & R.S. 194. University Park: Department of Agricultural Economics and Rural Sociology, Agricultural Experiment Station, Pennsylvania State University. Heady, Earl O., and Steven T. Sonka. (1974). "Farm Size, Rural Community Income, and Consumer Welfare." American Journal of Agricultural Economics 56:534-42. Heckman, James J. (1976). "The Common Structure of Statistical Models of Truncation, Sample Selection, Limited Dependent Variables, and a Simple Estimator for Such Models." Annals of Economic and Social Measurements 5:475492. Huffman, Wallace E. (1980). "Farm and Off-Farm Work Decisions: The Role of Human Capital." The Review of Economics and Statistics 62:14-23. Huffman, Wallace E., and Mark D. Lange. (October 1989). "Off-farm Work Decisions of Husbands and Wives: Joint Decision Making." The Review of Economics and Statistics 71:471-480. Kraybill, David S., and Thomas G. Johnson. (July 27-30, 1986). "Off-Farm Employment and its Effects in the Community." Paper presented at the American Agricultural Economics Association annual meeting, Reno, Nevada. Larson, Donald K. (1976). "Impact of Off-Farm Income on Farm Family Income Levels." Agricultural Finance Review 36:7-11. Lass, Daniel A. (May 16,1988). "Factors Affecting the Supply of Off-Farm Labor." Paper presented in the Symposium of Multiple Job-Holding among Farm Families in North America at Arlington, Virginia. Lass, Daniel A., and Conrado M. Gempesaw, II. (1992). "The Supply of Off-Farm Labor: A Random Coefficients Approach." American Journal of Agricultural Economics 74:400-411. Lass, Daniel A., Jill L. Findeis, and M. C. Hallberg. (1989). "Off-farm Employment Decisions by Massachusetts Farm Households." Northeastern Journal of Agricultural and Resource Economics 18:149-159. Leuning, Robert A., and Bruce L. Jones. (1989). "Characteristics of U.S. Farm Managers." In Deborah Smith, ed., Farm Management: 1989 Yearbook of Agriculture. Washington, DC: U.S. Government Printing Office.
Appendix
2
141
Mukheijee, Jaideep, Sharon M. Brucker, and Steven E. Hastings. (March 1990). "Factors Affecting the Off-Farm Labor Patterns of Delaware Farm Operators." Paper presented at the Southern Regional Science Association annual meeting, Washington, DC. Simpson, Wayne and Marilyn Kapitany. (1983). "Off-Farm Work Behavior of Farm Operators." American Journal of Agricultural Economics 65:801-805. Sumner, Daniel. (1982). "The Off-Farm Labor Supply of Farmers." American Journal of Agricultural Economics 64:499-509.
142
Appendix 2
NERCRD Proposal Budget Form-Both States Requested funds
In-Kind Funds
Matching
Expenditure
NERCRD
Host Institution
Sources
Salaries
$11,900
$11,200
Total $23,100
— —
... ...
... ...
Co-Investigator
6,135
6,135
12,270
Co-Investigator
5,765
5,065
10,830
700
—
700
500
—
500
Graduate students Faculty
Wages Secretary Wage payroll Programmer Computer time Materials/supplies Office costs Telephone Postage Photocopy Travel expenses
1,500
...
1,500
...
8,800
8,800
...
400
Lodging Meals Mileage, airfare Conference Speakers Facilities Scholarships Other (Indirect costs @ 44%) Publication costs
400
Training materials
...
Other (Data Collected) Totals
5,000 $20,000
5,000 $20,000
$40,000
143
Appendix 2 NERCRD Proposal Budget Form-West Virginia
Expenditure Salaries Graduate students Faculty Co-Investigator
Requested funds NERCRD
In-Kind Funds Host Institution
$6,135
$6,135
— —
... ...
6,135
6,135
Wages Secretary Wage payroll
700
...
250
...
750
...
...
...
—
4,943
Programmer Computer time Materials/supplies Office costs Telephone Postage Photocopy Travel expenses Lodging Meals Mileage, airfare Conference Speakers Facilities Scholarships Other (Indirect costs @ 44%) Publication costs
400
...
3,000
_._
Training materials Other (Data Collected) Totals
$11,235
$11,078
Matching Sources
Total
144
Appendix 2
NERCRD Proposal Budget Form-Massachusetts
Expenditure Salaries
Requested funds
In-Kind Funds
Matching
NERCRD
Host Institution
Sources
$5,765
$5,065
5,765
5,065
Graduate students Faculty Co-Investigator Wages Secretary Wage payroll Programmer Computer time Materials/supplies
250
Office costs Telephone Postage Photocopy Travel expenses
750
Lodging Meals Mileage, airfare Conference Speakers Facilities Scholarships Other (Indirect costs @
—
3,857
2,000
-~
$8,765
$8,922
44%) Publication costs Training materials Other (Data Collected) Totals
Total
Appendix 3 Essentials of a Project Outline for Hatch, Mclntire-Stennis, or State Matching Funds The format for the outline for proposals to be funded from Hatch, MclntireStennis, or State Matching Funds should follow the list below (specified in the Administration manual for the Hatch or Mclntire-Stennis Acts).
TITLE A brief, clear, specific designation of the subject of the research. The title, used by itself, should reflect the objectives and scope of the project.
JUSTIFICATION Present (1) the importance of the problem to agriculture and rural life of the state or region, (2) reasons for doing work (such as the needs the project will fill) and doing it at this time, and (3) ways in which public welfare or scientific knowledge will be advanced.
PREVIOUS WORK AND PRESENT OUTLOOK A brief summary of previous research (citing important publications), status of current research, and the additional knowledge needed that the project is expected to provide. (Literature citations may be listed at the end of the project outline.)
146
Appendix 3
OBJECTIVES A clear, complete, and logically arranged statement of the specific objectives of the project. PROCEDURE A statement of the essential working plans and-methods to be used in attaining each of the stated objectives. Procedures should correspond to the objectives and follow the same order. Phases of the work to be undertaken currently should be designated. Location of the work and the facilities and equipment needed and available should be indicated. Wherever appropriate, the procedure should produce data suitable for statistical analysis. The procedure should reflect careful planning and should provide flexibility for changes if changes become necessary. PROBABLE DURATION An estimate of the maximum time likely to be required to complete the project and publish results. Whenever a material change in the objectives of a project is advisable, a new or revised project outline should be prepared and submitted. A major change in procedure might also necessitate a revision of the project outline. FINANCIAL SUPPORT Estimated annual allotments (by funds) to (1) salaries and (2) maintenance based on analysis of requirements for labor, equipment, supplies, travel, and other operating expenses. Or, as an alternative, the estimated total scientist-years (SY) effort proposed for the project. PERSONNEL The leader or leaders and other technical workers assigned.
Appendix 3
147
INSTITUTIONAL UNITS INVOLVED Each subject-matter unit in the Agricultural and Forestry Experiment Station and any other units of the institution contributing essential services or facilities. The responsibilities of each should be indicated. If there is an advisory, coordinating, or directing committee for the project, the official title of the committee should be listed.
COOPERATION A statement listing the U.S. Department of Agriculture or other stations, institutions, or agencies expected to cooperate formally or informally on the projects. If project is part of a Regional effort, list Regional research Project Number.
REFERENCE Barr, A. L. (1984). Procedures for Preparing and Processing Proposals, Reports, and Manuscripts. Morgantown, WV: College of Agriculture and Forestry Experiment Station, West Virginia University.
This page intentionally left blank
Appendix 4 Sample Research Proposal 2: An Analysis of Rural Health Services in Louisiana JUSTIFICATION Health services are essential components of social infrastructure in a growing rural community (Hanson, 1972). Part of the strategy for economic development in Louisiana rural areas is to assure the availability of adequate health services to improve the health and productivity of the labor force and to attract business ventures. The maintenance of the productivity of the labor force and the life and well-being of the society depends on the availability of health services of a high quality and in sufficient quantity to meet needs. Poor health adversely affects educational attainment, labor force participation, and earning ability. Poverty causes disease and disease causes poverty. This interaction between poor health and poverty often creates a vicious cycle, reducing living standards and well-being of people and communities (Roemer, 1968; Luft, 1978). Though the relationship operates in all types of geographic and social settings, the problem is more intense in rural areas than in urban areas partly because the resources for preventing and treating sickness are less developed in the rural areas (Roemer, 1968; Doherty, 1970). The delivery of rural health services is a special problem because of rural people's relative inaccessibility to professional health services and their greater exposure to inferior socioeconomic conditions as compared to urban dwellers. The need to improve health services is greater in rural areas than in urban areas. Rural areas have disproportionate numbers of both older and poorer people-groups having more health problems and needing more health services than the general population (Schmidt, 1976). In rural areas, income is generally lower and health services are generally poorer and less accessible than in urban areas. At the same time, the incidence of illness and accidents
150
Appendix 4
limiting work activity and reducing incomes increases with both rurality and poverty. Low incomes and sparse population prevent rural areas from competing effectively in the market for medical services. Consequently, deficiencies exist in both quantity and quality of rural-located medical personnel and facilities. Distance poses a problem for the rural poor who lack transportation, finance, or the information necessary to obtain needed services (Doherty, 1970; Schmidt, 1976). Despite the disparity in the distribution of health services and health problems between the rural and urban areas, people in the rural areas are by no means without health care. The problem is that the rural population finds quality health services inaccessible. Relatively few physicians and facilities are found in the rural areas. Doctors tend to concentrate in metropolitan areas, where the opportunities are greater for high incomes, modern medical technology, and community amenities. The expectation is that as metro areas become saturated, physicians will find it more advantageous to locate innonmetro areas and increase their base of potential patients and sources of income ("Study Reveals Rural Areas Attracting More Physicians," 1985). Louisiana's health system has experienced similar developments and differences in the socioeconomic characteristics and health services between the rural and urban areas. According to the Louisiana Department of Health and Human Resources (1982), Louisiana's overall health status falls short of the nation's health status. Louisiana ranks among states with the greatest number of persistent low-income areas. Forty-one of the state's 64 parishes are classified as poverty areas, having more than 22 percent of the population with incomes below the poverty level. The current state physician-topopulation ratio is 136.1 per 100,000 population, compared to 165.3 physicians per 100,000 population for the nation. Louisiana's age-adjusted death rate in 1975 was 726 per 100,000 population, or 1.8times the U.S. rate for 1975. The rate increased to 890deaths per 100,000 white population and to 980 per 100,000 black population in 1978. The infant death rate (deaths of persons under one year of age excluding fetal death) was 1,740 per 100,000 population in 1978. The black infant death rate in 1978 was 2,450 per 100,000 population, or 1.9 times the white infant death rate (U.S. Department of Commerce, 1982; Department of Health and Human Resources, 1979). This research will provide timely and precise analysis of health problems of rural areas and ascertain how these problems differ in severity and type from those of the urban areas. Information on health services provided by this study will assist in the development of health plans and programs for
Appendix 4
151
improving the well-being of the rural community. The information generated by this research will be of importance for decision making by planners and public and private leaders in agriculture, industry, and government. Rural communities will benefit through improved government health programs and rural development policies made possible by the information from this and related studies.
REVIEW OF PREVIOUS WORK Much information is available on general community services, including health, housing, education, sewer, water, and solid waste disposal (Jones and Gessaman, 1974; Jones and Murdock, 1978; Maki, 1974; Eddleman, 1974; Day, 1968; Woods, Doeksen, and Nelson, 1983; Summers and Clemente, 1976). An examination of the literature and the Current Research Information System (CRIS) revealed several ongoing and recently completed research efforts related to the general public health services. The CRIS review revealed no current research projects studying the impact of social and economic changes on the delivery of health services in the rural areas. Several studies (Beale, 1975 and 1985;Ploch, 1978) have noted the growth of population in some rural and small towns as a result of the search for quality of life and for better economic opportunities. Another process is increasing urbanization and industrialization of the rural areas, bringing with it a need for greater efforts to maintain a healthy environment and for health services. Rural industrialization is adding new businesses and industries in some rural areas (Summers and Hirschel, 1985). A large influx of retirees with health care needs and improved incomes has augmented spending in rural areas. This income flow can lead to the creation of employment opportunities and income for the rural population in the same way that industrial payrolls generate jobs. The disparities between needs for health services in rural areas and the actual delivery of these services are documented (Matthews, 1974;Dubov and Deaton, 1977; Davis and Schoen, 1978). In evaluating the trends in the use of health services by the rural communities where poor people are heavily concentrated, a significant narrowing is noted of the gap in the opportunity to receive health services. However, some observers (Bice, Eichharn, and Fox, 1972; Lefcowitz, 1973; Schmidt, 1976; Feldstein, 1973; Council of Economic Advisors, 1976) conclude that health service programs for the poor have been so successful that there is no longer cause for special concern.
152
Appendix 4
Others cite instances in which the poor now surpass others in the use of health services and suggest that perhaps the poor use them excessively. Neither of these conclusions is warranted by the evidence gathered so far. Greater use of services by the poor than by others can be attributed to the greater incidence of illness and injury among them. Access to a regular source of health service is important from an economic perspective. Illness prevention is considered cost-effective compared with the alternative of medical treatment after illness. Because health insurance programs such as medicaid and medicare for the low-income population are publicly funded, access to health service for this group can significantly affect public intervention costs for health service (Makuc, 1981). Rising health care costs may exclude persons from purchasing needed health services. Purchasing needed health services may severely limit their purchase of other essential items such as food and clothing. Thus, the recent social and economic changes in the structure and organization of American society have caused concern among federal, state, and local area decision makers over how to finance and deliver health and other community services to rural people in the face of rising per capita service costs and a declining or slowly rising tax base (Eddleman, 1974). Evidence suggests no clearly discernible, separate rural health delivery systems nor general organizations or program centered on rural health care (Doherty, 1970; Hanson, 1972). The central question is whether rural areas warrant special policy and program attention in the delivery of health services. Although rural and urban areas have many health care problems in common, the different demographic and socioeconomic characteristics, emergency care difficulties, accident fatality, and spatial variation in rural areas call for unique and different health care delivery systems in rural areas (Matthews, 1974). Not only do rural residents living in poverty face adverse health effects, but their communities often have fewer resources to assist them in bettering their conditions (Ahearn and Fryar, 1985). Many rural residents properly rely on urban health care services. The most pressing research need is to determine the economic feasibility and delivery effectiveness of alternative systems of emergency medical services (EMS). EMS systems must encompass sufficient area, economic base, and people to assure that all the necessary components of a system can be contained and supported within the area, including well-staffed and equipped transport systems and emergency treatment centers. Because of the widely scattered population in rural areas and the scarcity of health personnel, well-developed emergency medical service systems are rare. Death rates from accidents are far higher in rural areas than in urban areas.
Appendix 4
153
Ambulances sometimes must travel long distances to rural patients for emergency medical services (Schmidt, 1976). Rural roads are in poorer condition than urban roads. Because of such problems, in Louisiana an EMS is more expensive per capita in rural areas than in urban areas. For example, EMS in East Baton Rouge serves residents at an operating cost of about $8.86per person compared to $13.30per person for residents in Arcadian rural areas (Dunne, 1984; Mid-Louisiana Health Systems Agency, 1980). In the past, several researchers (Bierman and Powers, 1970; Daberkaw and King, 1977; Duncan, 1976;Doeksen, Frye, and Green, 1975) have noted the relative lack of empirical research on providing emergency medical services in rural areas. Their studies analyzed the cost and capabilities of possible options to provide emergency medical services in rural areas. The studies formulated procedures to conduct emergency medical service systems and measure feasibility based on estimated demand and acceptable effectiveness. The internal organizational management and operation of the system, demographic and socioeconomic characteristics of rural community, and the system linkages with other community services differ from region to region. Thus analysis is needed for specific areas and cannot be generalized.
OBJECTIVES The overall objective of this study is to analyze the delivery and utilization of health services and facilities in the rural population of Louisiana. Specific objectives are to: 1. Assess the demographic and socioeconomic characteristics and structure of the rural areas important to the delivery of health services; 2. Relate rural health problems to the availability, cost, and utilization of health services in the rural areas; 3. Design and analyze a general emergency medical service (EMS) system that is economically feasible and appropriate to a sparsely populated rural area; and 4. List policy and organizational alternatives for improving the delivery of adequate health services in the rural areas.
154
Appendix 4
METHOD OF ANALYSIS The research project will utilize primary and secondary data on the demographic and socioeconomic characteristics and the availability and utilization of health services and facilities in the rural areas of Louisiana. Secondary data will be compiled from the Census of Population, U.S. Census of Agriculture, periodic reports of the U.S. Department of Agriculture, Louisiana Department of Health and Human Resources, U.S. Department of Transportation, and the Louisiana Medical Society. Descriptive, quantitative primary data on the delivery of health services and socioeconomic characteristics will be obtained by personal interviews of a randomly selected sample of rural areas. The field data used in this study will be drawn from a selected state planning district. Health professionals, agency executives, community association leaders, and a representative sample of the people who suffer adversities and inadequate health care will be surveyed. Procedures for specific objectives are listed below:
Objective 1 Assess the demographic and socioeconomic characteristics and structures of the rural areas important to the delivery of health services. Procedure: Secondary and primary data will be used to assess and analyze the demographic and socioeconomic characteristics of the rural community that relate to health problems. Information on family size, household income, education, employment, and population density and distribution by age, sex, and race will be collected from the rural area. A traditional population model will be developed to project the rural population. A cohort-survival population predictor will be utilized. Using an initial population, birth rates, death rates, and migration rates for each specific cohort, the future population of the study area will be predicted annually by age and sex categories.
Objective 2 Relate the patterns of health problems to the availability, cost, and utilization of health services in the rural areas. Procedure: The study will estimate underutilization of or deficiencies in health services and the factors affecting the distribution of health services in
Appendix 4
155
the rural areas based on needs as determined by the social and economic characteristics of the rural population. A descriptive statistical analysis on the availability, distribution patterns, and utilization of health services will be made for a selected typical rural town and surrounding community area consisting of a small rural hospital or health center within the selected state planning district area. To determine the total cost #nd cost efficiency of health services in the study areas, a survey form will be developed to collect cost data by service categories. Measures of the supply of health services include occupancy rate, utilization rate, patient-days of service extended by age categories, patient-day cost per service category, health care cost per capita, expenses by service categories, hospitalization rates, admission rates, general hospital beds, nurse- and physician-to-population ratios, and the various resources and facilities available. Relationship between social and economic characteristics and availability of health services in the study area will be analyzed by multivariate statistical techniques. Measures of rurality will be added for possible contribution to variations in the availability of health services. Relationships found from this analysis may indicate deficiencies that can be overcome to alleviate health service scarcity. Ordinary least square linear regression will use health services as the dependent variable and socioeconomic and rurality variables as independent variables.
Objective 3 Design and analyze a general emergency medical service (EMS) system that is economically feasible and appropriate to the sparsely populated rural area. Procedure: The providers of ambulance services will be identified from the study area. From these sources alternative arrangements for EMS will be identified. An analysis and estimation procedure of income receipts and costs associated with specific standards or services for each of the major alternative delivery systems will be developed. Data will be collected from ambulance operators on the number of calls made. Those include highway accidents involving automobiles or other vehicles, transfer calls that include the movement of patients between hospitals, and other medical calls. Data will be collected on age of patients, fees charged or income generated, and the percentage of bills collected. Supply data on medical expenses and vehicle operation expenses will be collected from the ambulance operators. Estimates by dealers of ambulance
156
Appendix 4
equipment, communication equipment, and buildings are pooled to supply capital cost information, including cost of various types of vehicles and medical equipment. Having collected the necessary information, the analyst will then determine the distribution of costs and benefits under alternative arrangements for financing emergency medical services. Alternatives will be compared to help decision makers determine what system would be most appropriate for the rural area.
Objective 4 List policy and organizational alternatives for the delivery of adequate health services in the rural areas. Procedure: A first step is to identify the existing federal, state, and local government organizations and structure that influence the delivery of health services in the rural areas of Louisiana. Analysis will investigate the nature and scope of the interorganization decision-making process associated with the delivery of health services and the degree to which government health programs meet the health care needs of the rural area. Based on the analysis of demographic and socioeconomic changes and supply of health services in the rural areas, policy recommendations will be made to decision makers to serve in planning, financing, and regulating health services in rural areas of Louisiana. Policy recommendations will be made based on the findings of the study. Results of the study will be published in research reports, research bulletins, and professional journals.
INSTITUTIONAL UNITS INVOLVED The institutional unit involved will be the College of Agriculture and Home Economics, Agricultural Experiment Station, at Southern University. However, implementation of the project calls for extensive involvement and contribution of essential services and facilities of the Department of Agricultural Economics. The full support and assistance of the Department of Sociology and Social Works, the School of Nursing, and Allied Health Programs of the university will be needed. The research project will operate under the auspices of Hatch/USDA, which will be the principal source of funds.
Appendix 4
157
COOPERATION This research study will also be conducted in close cooperation with the State Department of Health and Human Resources, Highway Traffic Safety Administration, Emergency Medical Service agencies, Local Hospitals and Medical Society, State and Local Government Offices, and Cooperative Agricultural Extension Services.
PERSONNEL The principal investigator will be Dr. Tesfa G. Gebremedhin, associate professor of agricultural economics (resume attached) in the College of Agriculture and Home Economics, Southern University. Suitable support personnel, including a graduate research assistant, will assist in the study.
PROBABLE DURATION Estimated duration to collect, compile, and analyze the data will be at least three years from the date of funding.
BUDGET EXPLANATION Adequate funding is essential to complete the study. The total estimated annual cost of the proposed research project is $40,000. Funds allocated for salaries will be used to cover a portion of the principal investigator, graduate research assistant, and a clerical typist. Travel funds will be used for collecting data and for attending professional association meetings. Operating expense is mainly for computer service and publications. Basic materials and supplies include offices supplies and communication materials. The 10 percent overhead cost covers university facilities used to conduct the research.
158
Appendix 4
Budget
ITEMS
DOLLARS
SALARIES Principal Investigator (two months)
28,800 10,600
Graduate Research Assistant (12 months)
9,400
Clerical Typist (three months)
4,000
Benefits @ 20%
4,800
TRAVEL
3,000
Out-of-State Travel
1,250
In-State Travel
1,750
OPERATING EXPENSES
2,700
Computer Service
1,500
Printing, Duplication, etc.
1,200
MATERIALS & SUPPLIES
1,500
OVERHEAD COSTS @ 10%
3,000
TOTAL
40,000
REFERENCES Ahearn, Mary C, and Michelle D. Fryar. (March 1985). Physicians in Nonmetro Areas During the Seventies. Rural Development Research Report No. 46. Washington, DC: Economic Research Service, U.S. Department of Agriculture. Beale, C. L. (1975). The Revival of Population Growth in Nonmetropolitan American. Research Report No. 605. Washington, DC: Economic Research Service, U.S. Department of Agriculture. Beale, Calvin L. (February 1985). "U.S. Population Trends Break with Past." Rural Development Perspectives 1:4-7. Bice, Thomas W., Robert L. Eichham, and Peter D. Fox. (May-June 1972) "Socio-economic Status and Use of Physician Services: A Reconsideration." Medical Care 10:261-271. Bierman, Leland, and Mark J. Powers. (April 1970). Ambulance Services in Northwest South Dakota. Bulletin 569. Brookings: Agricultural Experiment Station, South Dakota State University.
Appendix
4
159
Council of Economic Advisors. (January 1976). Economic Report of the President. Washington, DC: U.S. Government Printing Office. Daberkaw, A. G., and J. A. King. (1977). "Response Time and the Location of Emergency Medical Facilities in Rural Areas: A Case Study." American Journal of Agricultural Economics 59:466-477. Davis, Karen, and Cathy Schoen. (1978). Health and the War on Poverty: A Ten Year Appraisal. Washington, DC: Brookings Institutions. Day, Lee M. (December 1968). "Community Facilities and Services: An Economic Framework for Analysis." American Journal of Agricultural Economics 50:1195-1205. Department of Health and Human Resources. (January-December 1979). Louisiana Vital Statistics Report. Provisional Statistics. Baton Rouge, LA: Office of Health Services and Environmental Quality. Department of Health and Human Resources, Division of Health Planning and Development. (1982). 1982-87 Louisiana State Health Plan. Baton Rouge, LA: Office of Management and Finance. Doeksen, Gerald A., Jack Frye, and Bemal Green. (1975). Economics of Rural Ambulance Service in the Great Plains. AER 308. Washington, DC: Economic Development Division, U.S. Department of Agriculture. Doherty, Neville. (1970). Rurality, Poverty, and Health: Medical Problems in Rural Areas. Agricultural Economic Report. Washington, DC: Economic Research Service, U.S. Department of Agriculture. — . (November 1971). Efficiency in the Distribution and Utilization of Hospital Services: A Case Study in Rural Michigan. ERS-492. Washington, DC: Economic Research Service, U.S. Department of Agriculture. Dubov, Irving, and Brady J. Deaton. (February 9, 1977). "Rural Health Care: The Economists' Perspective." Paper presented to Rural Sociology Section, Southern Agricultural Scientists Association annual meeting, Atlanta, Georgia. Duncan, Marvin R. (1976). "An Analytical Model for Multi-county Health Planning Areas." Pp. 214-229 in Edward W. Hassinger and Larry P. Whiting, eds., Rural Health Services: Organization, Delivery, and Use. Ames: North Central Regional Center for Rural Development, Iowa State University. Dunn, James W., and Gerald A. Doeksen. (1980). "A Health Care System Model for Nonmetropolitan Areas." American Journal of Agricultural Economics 62:58-65. Dunne, Mike. (April 30, 1984). "Arcadian Ambulance Offers Options if EMS Tax Fails." Morning Advocate. Baton Rouge, Louisiana. Eddleman, B. R. (December 1974). "Financing Public Services in Rural Areas: A Synthesis." American Journal of Agricultural Economics 56:959-963. Feldstein, Martin S. (September 1973). "The Medical Economy." Scientific American 229:151-172.
160
Appendix 4
Hanson, Ivan R. (July 1972). "Public Health Services." Pp. 111-137 in Research Applications in Rural Economic Development and Planning. Agricultural Experiment Station Research Report P-665. Stillwater: Oklahoma State University. . (1976). "Comprehensive Health Planning Issues for Rural Health Researchers." Pp. 188-199 in Rural Health Services: Organization, Delivery, and Use. Ames: North Central Regional Center for Rural Development, Iowa State University. Jones, Lonnie L., and Paul H. Gessaman. (1974). "Public Service Delivery in Rural Areas: Problems and Decision." American Journal of Agricultural Economics 56:936-944. Jones, Lonnie L., and Steve H. Murdock. (December 1978). "The Incremental Nature of Public Service Delivery: Implications for Rural Areas." American Journal of Agricultural Economics 60:955-960. Lefcowitz, Myron J. (March 1973). "Poverty and Health: A Reexamination." Inquiry 10:3-13. Luft, Harold. (1978). Poverty and Health: Economic Causes and Consequences of Health Problems. Cambridge, MA: Ballinger. Maki, Wilbur R. (December 1974). "Local Funding of Rural Public Services." American Journal of Agricultural Economics 56:946-952. Makuc, Diane. (December 1981). "Changes in Use of Preventive Health Services." Health United States, 1981. (PHS) 82-1232. Washington, DC: National Center for Health Statistics, U.S. Department of Health and Human Services. Matthews, Teresa H. (April 1974). Health Services in Rural America. Agricultural Information Bulletin No. 362. Washington, DC: Rural Development Service, Economic Research Service, U.S. Department of Agriculture. Mid-Louisiana Health Systems Agency. (September 1980). Health Systems Plan: A Draft 1981-85. Baton Rouge, Louisiana. Ploch, L. A. (1978). "The Reversal of Migration Patterns - Some Rural Development Consequences." Rural Sociology 43:293-303. Powers, Mark J., and Leland G. Bierman. (1970). Supply and Demand of Medical Services in Northwest South Dakota. Bulletin 568. Brookings: Department of Economics, Agricultural Experiment Station, South Dakota State University. Roemer, Milton. (May 1968). "Health Needs and Services of the Rural Poor." Rural Poverty in the United States. Report by the President's National Advisory Commission on Rural Poverty. Washington, DC: U.S. Government Printing Office. Schmidt, Herman. (September 1976). Health of and Health Services for Rural People. Washington, DC: Congressional Research Service, Education and Public Welfare Division.
Appendix
4
161
"Study Reveals Rural Areas Attracting More Physicians." (June 1985). Southern Rural Development Capsules. Vol. 5, No. 6. Starkville: Mississippi State University. Summers, G. F., and F. Clemente. (1976). "Industrial Development, Income Distribution, and Public Policy." Rural Sociology 41:248-268. Summers, G. F., and Thomas A. Hirschel. (February 1985). "Retirees as a Growth Industry." Rural Development Perspectives 1:13-16. U.S. Congress. (1972). Public Law 89-749. Stat. 1180. . (1980). Rural Development Policy Act of 1980. PL 96-355, 94 Stat. 1171. U.S. Department of Commerce, Bureau of the Census. (1982). State and Metropolitan Area Data Book: A Statistical Abstract Supplement. Washington, DC: U.S. Government Printing Office. Woods, Mike D., Gerald A. Doeksen, and James R. Nelson. (December 1983). "Community Economics: A Simulation Model for Rural Development Planners." Southern Journal of Agricultural Economics 15:71-77.
This page intentionally left blank
Index ABI/Inform, 122 Abstract, 76-77. See also Summary Agricola, 123 Analysis, method of, 35, 48. See also Procedures Andrew, Chris, 3, 6, 31, 36, 43, 45, 5152, 64 Ary, Donald, 3, 10, 15, 31, 36, 40, 45, 53, 64 Audiovisuals, 105 Barrass, Robert, 1, 31, 40, 60, 64, 9395, 106 Bates, Jefferson, 93, 106 Best, John W, 3-4, 6, 10-12, 31, 5455, 64 Beveridge, W. I. B., 35, 64, 95, 106 Bibliography, 63-64. See also References Blake, Gary, 74, 91, 92, 94-106 Blaug, Mark, 3, 31, 33 Bly, Robert W., 74, 89, 92, 94-106 Bolsky, Morris I, 93-94, 106 Bonnen, James, 31, 33 Borg, Walter R., 45, 54-55, 64 Budget (proposal), 84, 87
Causation, 14-15 Cohen, Louis, 3, 9, 13-14, 31, 53-54, 64 Collins, Sarah H , 92, 106 Communication: general, 91; oral presentation, 102-106; symbolic, 59; verbal, 59. See also Writing. Composition, fallacy of, 15 Computer, library search, 121. See also Library search Conclusion, 35, 60-61. See also Summary Cooperation (proposal), 83, 86-87 Correlation, 14-15; spurious, 16 Cover letter (proposal), 75 Cross-disciplinary research, 24 Cybernetics, 29 Data: assembly, 4-5, 12, 14, 35, 51, 80; cross-sectional, 52, 80; experimental, 52; primary, 21, 51, 56, 80; secondary, 21, 51, 80; survey, 52; time series, 52, 80. See also Survey Database, 121 Day, Robert A., 60-61, 65, 93, 96-106
164 Deduction, 22, 23. See also Reasoning Descriptive economics, 20 Diagnosis, 7. See also Research Diagnostic hypothesis, 14, 46 Dillman, Don, 56, 65 Disciplinary research, 20 Discussion (results), 58-59 Dowling, Carol L , 122, 123 EconLit, 123 Economics: descriptive, 20; normative, 19-20; positive, 19-20, 23; predictive, 20; prescriptive, 20 Electronic search, 121, 122. See also Library search Emberger, Meta Riley, 3, 31 Empiricism, 14, 21, 22 Error: measurement, 2, 12, 48; sampling, 2, 48; specification, 48. See also Statistics Ethics, 96 Evaluation (proposal), 85-98. See also Literature Exercises, 32-33, 66-68, 88-89, 106107, 115 Experimentalism, 22, 23 External worth, 15, 18 Feedback, 14 Footnotes, 62-63 Format (proposal), 127-129, 145-147 Friedman, Milton, 64-65 Fuguitt, G., 42, 65 Fujimoto, Isao, 42, 65 Gilreath, Charles I., 119-21, 123 Goldschmidt, Walter, 42-43, 65 Goldstein, Inge, 3, 31 Goldstein, Martin, 3, 31 Guidelines, for research proposals, 127; speeches, 102; research, 35; writing, 96
Index Hall, Marian Ross, 3, 31 Hawthorn effect (Heisenberg principle), 18 Hayes, Michael, 43, 65 Hayes, Robert, 53, 65, 93, 106 Heady, Earl, 42, 65 Henderson, David, 38, 42-44, 47, 65 Henry, W.M., 120-22, 124 Hildebrand, Peter, 3, 6, 31, 36, 44-45, 51-52, 65 Humanities, 4 Hypothesis: advanced, 47; conceptual, 46; definition of, 44-45, 79, 86, 110; diagnostic, 46; formulation, 4448; general, 1-2, 6, 10, 12-13, 35, 40; null, 12, 47; primitive, 47; remedial, 46; test, 14 Induction, 21-23. See also Reasoning Institutionalism, 22-23 Institution capability (proposal), 83-85, 87, 88-89 Interdisciplinary research, 25 Internal validity, 15-16, 18 Introduction (proposal), 80; Jacobs, Chester, 10, 15, 31, 36, 40, 45, 53, 65 Johansen, H., 42, 65 Johnson, Glenn, 13, 24, 31 Journal article, review, 109 Kahn, James V., 3-4, 6, 10-12, 32, 5657, 67 Knowledge (scientific), 2-5, 13, 40 Kuhn, Thomas, 26-28, 31 Ladd, George W., 93, 106 Law, 2-4, 48. See also Theory Leedy, Paul. D., 3, 7-9, 13, 31, 40, 59, 65, 118, 124 Lehnert, Thor, 122, 124
165
Index Leigh, J. A., 120-122, 124 Lobao, Linda, 42-43, 65 Local area computer network (LACN), 117 Locke, Lawrence E., 74, 88 Logic, 8-10, 16, 21. See also Reasoning Logical positivism, 48 Library search (literature): computer, 117, 119-23; general, 117-124; manual, 118-119. See also Literature Literature, review of, 39-43, 79, 86; search, 117, 122 Machine-readable cataloging (MARC), 117 Manion, Lawrence, 3, 9, 13-14, 31, 5354, 64 Manuscripts: enhancing acceptance, 100-102; preparing, 96-100; reviewing, 109-115 Marousek, Gerald, 42, 65 McCloskey, Donald, 107-108 Meador, Roy, 74-75, 77-78, 80-81, 85, 89 Meites, Louis, 112-113, 115 Methods: research, 2-3, 28, 27, 35-36, 86, 91; scientific, 1-5, 12-13, 14, 35. See also Research; Science Miller, Roger LeRoy, 14, 31 Mlay, Gilead, 23, 31 Model: definition of, 14, 48; general, 5, 13, 26, 29; linear programming, 48; regression, 4, 47, 49. See also Methods; Research Moral imperatives (ethics), 96 Muller, Jerry, 31, 33 Multidisciplinary research, 25 Multivariate statistical technique, 4. See also Statistics Myth (belief), 7
Nelson, Robert, 19, 31, 33 Neway, Julie M., 118, 124 Normative economics, 19 Objective science, 16 Objectives (research), 5, 35, 43-44, 79, 88-89, 110 Occam's Razor, 16 Olmstead, Alan, 43, 65 On-line public access catalogs (OPAC), 117 Oral presentation, 102 Paradigms, 26-28 Parsimony, principle of, 16 Particularistic premise, 9 Personnel capabilities (proposal), 8182, 84, 86-87 Peterson, Martin S., 3, 5, 31, 40, 61, 65 Philosophy, 4 Popper, Karl, 31, 33 Population, 53. See also Sampling Positive economics, 19 Positivism (logical), 48 Postscript on research planning, 29 Pragmatism, 14, 21, 22 Predictive economics, 20; science, 64 Premises, 8-9. See also Reasoning Prescriptive economics, 20 Probability, 2, 12, 23 Problem: definition of, 14, 36; general, 1-2, 6-8, 29, 30, 35; statement of, 36-39, 78 Problem-solving research, 13, 25 Procedures (proposal), 80, 86. See also Methods Prognosis, 7. See also Research Program evaluation and review technique (PERT), 30 Project outline, example, 145 Proposal (grant): composition and organization of, 74; duration, 83;
166 evaluation, 85; example, 131; general, 73-88; solicited, 74; unsolicited, 74-75, 77 Rationalism, 14, 21, 22 Razavieh, Asghan, 10, 15, 31, 40, 45, 53, 64 Reasoning (logical): deductive, 8-12, 14; inductive, 10-12, 14; types of, 8 Recommendations, 61 References, 63-64; style, 68-72 Reification, 21 Relativism, 22, 23 Reliability, 16 Remedial hypothesis, 14, 46 Remediation, 7. See also Research Request for proposal (RFP), 74 Research: applied 5-7; basic, 5-7; collaborative and disciplinary, 2426; defined, 4; problem solving, 25, 47; subject matter, 24. See also Methods; Science Research libraries information networks (RLIN), 117 Research proposal, example 131, 149 Results (findings), analysis, 14, 35, 5860; dissemination of, 83, 85, 87 Review (manuscript): accuracy, 4, 113; checklist, 109; comments, 109112; explicitness of, 114; helpfulness of, 114; objectivity, 113; relevance of, 114; responsibility, 112-114; thoroughness of, 114 Review of literature, 39-43 Sample research (proposal), 131-144; 149-161 Sampling: cluster, 55; error, 48; general, 2, 6, 12, 53; random, 54; size, 55-56; stratified, 54; systematic, 54. See also Survey Samuels, Marilyn Schauer, 62, 65, 93,
Index 106 Santayana, George, 11, 31 Schreiner, Dean, 43, 65 Science: biological, 6; definition of, 3, 10, 12, 16, 18, 27; physical, 3, 4; predictive, 64; social, 3-4, 6, 9, 28, 52. See also Methods; Research Scientific revolution, 27; writing, 92 Shaffer, R., 43, 65 Significance (proposal), 81, 86 Silverman, Stephen J., 74, 89 SilverPlatter Information, Inc., 122-124 Skees, Jerry, 42, 65 Soil Science Society of America, 109, 113, 115 Sonka, S., 42, 65 Speech, checklist, 102 Spirduso, Waneen Wyrick, 74, 89 Stage fright, 103 Statistics: bias, 15; inference, 2, 9, 1112, 23. See also Error Stock, Molly, 6, 31, 102, 106 Structured science, 64 Strunk, William, Jr., 107-108 Student t-test, 11. See also Statistics Subject matter research, 24 Summary, 60-61. See also Conclusion Supportive services (proposal), 83, 8987 Survey: collection, 52, 56, 80; mail, 56; questionnaire, 56-58; telephone, 56. See also Data; Sampling Swanson, L., 42, 65 Syllogism, 8-9. See also Reasoning Symbolic communication, 59 Systems analysis, 28-29 Tautology (truism), 9, 11 Tedd, L. A., 24-26, 124 Theology, 4 Theory (principles), 2-5, 10, 13, 21, 2627. See also Law; Model.
Index Thinking: creative 1; ordered, 3, 7; rational, 7-8 Thought, schools of, 21 Title page (proposal), 76 Travers, Robert M. W., 15-16, 31, 36, 48, 56, 66 t-test, 11 Tuckman, Bruce W., 2-6, 15, 31, 36, 56, 59-61, 66 Turabian, Kate L , 93, 106 Tuttle, Frederick B., Jr., 92, 106 Tweeten, Luther, 3, 10, 12, 15, 23-24, 31, 42-47, 65-66 UNDESA (UN Department of Economic and Social Affairs), 74, 88 UNESCO, 74, 88 United Nations (UN). See UNDESA, UNESCO Universal premise, 9
167 University Microfilms International, 123, 124 Utilitarianism, 96 Validity, 13, 15. See also Internal Validity Verbal communication, 59 Visuals, planning, 105 White, E. B., 107-108 Wilken, Carl, 16 Williams, P. W., 120-122, 124 Wimberly, Ronald, 43, 66 Woods, Mike, 43, 65 Writing (technical): accuracy, 93, 96; checklist, 96; clarity, 95-96; completeness, 93, 96; ethics, 96; general, 2, 91-106; impartiality, 9394, 96; objectivity, 93-94, 96; order, 94, 96; requirements, 92; simplicity, 95-96. See also Communication; Manuscript
This page intentionally left blank
About the Authors TESFA G. GEBREMEDHIN is an Associate Professor of Agricultural Economics in the Division of Resource Management, College of Agriculture and Forestry at West Virginia University. He was previously an Assistant and Associate Professor at Southern University. He has written book chapters, a number of journal articles and published papers, experiment station bulletins, research reports, and extension publications. LUTHER G. TWEETEN is the Anderson Professor of Agricultural Marketing, Policy, and Trade in the Department of Agricultural Economics and Rural Sociology at The Ohio State University. Dr. Tweeten was previously Regents Professor at Oklahoma State University. He has published eight books and over 400 journal and miscellaneous articles, experiment station bulletins, and research reports.