P1: FTK/FNV
P2: FQJ/FLK,FFB
Archives of Sexual Behavior [asb]
QC: FTK PL111-335
June 24, 2000
10:25
Style file version Nov. 19th, 1999
Archives of Sexual Behavior, Vol. 29, No. 4, 2000
Self-Administered Questionnaires versus Face-to-Face Interviews in Assessing Sexual Behavior in Young Women Lauren E. Durant, M.S.,1 and Michael P. Carey, Ph.D.1,2
Effects of mode of assessment, person, and situational variables on the accuracy of self-reports of sexual behavior remain uncertain. To evaluate these influences, 190 young women completed measures of erotophilia and social desirability and then monitored their health-related behaviors with a diary for 8 weeks. They returned on two occasions to complete either a face-to-face interview (FTFI) or a self-administered questionnaire (SAQ) regarding their behavior over the same interval. To check the apparent accuracy of participants’ retrospective self-reports, a difference score was calculated by subtracting responses obtained on the FTFI or SAQ from the diary card. Results indicated that both modes of assessment were reliable; reliability did not differ as a function of mode of assessment. However, SAQs elicited less discrepant responses for protected vaginal sex; SAQ and FTFI reports for unprotected sexual behaviors were equivalent. Situational and person variables did not predict accuracy scores, which were impaired at higher frequencies of behavior. Results suggest that both modes of assessment were reliable and SAQs may be more accurate for some sexual behaviors. KEY WORDS: assessment; self-report; sexual behavior; interview; self-administered questionnaire.
INTRODUCTION Behavioral research on human immunodeficiency virus (HIV) and other sexually transmitted infections relies on self-reported data. Although some critics continue to characterize self-reports as unreliable, most investigators focus their efforts on identifying the determinants of accuracy (Weinhardt et al., 1998b). 1 Center
for Health and Behavior, Syracuse University, Syracuse, NY. whom correspondence should be addressed at Department of Psychology, 430 Huntington Hall, Syracuse University, Syracuse, New York 13244-2340.
2 To
309 C 2000 Plenum Publishing Corporation 0004-0002/00/0800-0309$18.00/0 °
P1: FTK/FNV
P2: FQJ/FLK,FFB
Archives of Sexual Behavior [asb]
310
QC: FTK PL111-335
June 24, 2000
10:25
Style file version Nov. 19th, 1999
Durant and Carey
Characteristics of the respondent that might influence the accuracy of selfreport include erotophilia (the tendency to evaluate sexual stimuli in a positive manner) and social desirability (the tendency to respond in a socially desirable manner). For example, Fisher et al. (1988) found that erotophobic participants had more negative reactions to communicating sexual content than did their erotophilic counterparts. Accordingly, erotophilic individuals may find sexual materials more pleasant and provide more accurate reports. Social desirability has been measured in two ways in investigations of sexual behavior: (a) as a general tendency to answer in a socially desirable manner (Crown and Marlowe, 1964) or (b) as a tendency to answer specific types of questions (e.g., sexual behavior) in a socially desirable manner or not at all. Catania et al. (1986) found that participants who were more threatened by sexuality questions were more likely to leave sexuality questions unanswered. Three characteristics of the assessment situation may also influence the accuracy of self-reports (Tourangeau and Smith, 1996). First, privacy (the perception that information will be held in confidence) may increase the accuracy of selfreports. For example, Strecher et al. (1989) found that participants tended to be less honest when they were in situations characterized as high demand. Second, perceived anonymity (the belief that one’s identity will not be known) of the assessment increases participants’ willingness to report sensitive behaviors (Bradburn and Sudman, 1979). For example, the presence of an interviewer may reduce selfdisclosure when respondents believe that the assessor can link their responses with personal identifiers. Alternatively, when participants believe that they cannot be connected with their responses, they are more likely to disclose sensitive information (Czaja, 1987). Third, credibility of the assessor appears to affect participants’ responses (Bradburn and Sudman, 1979). If participants believe that their information will be used in a legitimate prosocial manner, then they are more willing to disclose sensitive information (Weinhardt et al., 1998b). Commonly used self-report methods vary in the extent to which they provide privacy, anonymity, and credibility. The self-administered questionnaire (SAQ) may be the most commonly used mode of self-report. The SAQ affords privacy and allows administration to large numbers of participants economically. Participants can skip questions that are too embarrassing or intrusive, and SAQs are less labor-intensive to administer. However, use of SAQs often exclude participants who have limited literacy skills and result in more missing data. Nevertheless, several investigators have concluded that SAQs result in increased reports of sexual behavior when compared to other self-report modes (e.g., Boekeloo et al., 1994; Kalichman et al., 1997). The face-to-face interview (FTFI) requires that participants have a one-onone interaction with an interviewer. An advantage of the FTFI is that it allows for the clarification of responses and probing, which tend to result in fewer missing responses. However, Tourangeau and Smith (1996) compared two types of SAQs
P1: FTK/FNV
P2: FQJ/FLK,FFB
Archives of Sexual Behavior [asb]
QC: FTK PL111-335
Assessment of Sexual Behavior
June 24, 2000
10:25
Style file version Nov. 19th, 1999
311
(i.e., audio administered and computer administered) with FTFIs and found that FTFIs resulted in fewer reports of sexual behavior. These results may be explained in part by respondents’ privacy concerns. That is, respondents may be reluctant to disclose their sexual behaviors if they are not assured that their responses will be held in confidence by the interviewer. Furthermore, FTFIs are labor-intensive and expensive. However, FTFIs may be perceived by participants as a more legitimate assessment mode than SAQs. Self-monitoring, a third mode of self-report, requires participants to record their behavior contemporaneously over brief intervals. Diaries using ongoing selfmonitoring tend to produce higher reports of socially undesirable behavior that are presumed to be more accurate (e.g., Downey et al., 1995; McLaws et al., 1990; Reading, 1983). That is, in lieu of validation evidence, investigators have adopted the “more is better” standard of self-report. Such records may also minimize problems associated with low literacy skills, recall, and memory problems (e.g., telescoping). Diaries have not been used widely in HIV-prevention research because they require considerable effort, conscientiousness when recording behaviors over time, and they would require participants to keep prospective diaries before beginning an intervention. However, diaries have been used widely in the alcohol, weight reduction, and smoking literatures (e.g., McFall, 1970; Webb et al., 1991). Furthermore, the diary has been used to ascertain data quality (e.g., Downey et al., 1995) because two faulty assumptions have been perpetuated in studies of sexual behavior: (a) respondents always underreport their sexual behaviors, and (b) underreporting is constant across assessment points in time (Miller et al., 1990). To avoid misinterpretations of sexual self-report data, investigators have tried to verify sexual behavior with some type of corroborative evidence. Accordingly, investigators have assumed that self-monitoring yields the most accurate reports of sensitive behaviors (i.e., it is characterized as anonymous, private, and credible). Moreover, the available evidence supports our assumption that the diary should elicit the most accurate reports of sexual behavior in comparison to other modes of administration (e.g., Downey et al., 1995; McLaws et al., 1990), modes that lack one or more of the characteristics associated with increased accuracy (i.e., FTFI and SAQ). Few published studies have compared the accuracy of reports elicited by SAQs and FTFIs (Weinhardt et al., 1998b). In lieu of a “gold standard” to measure accuracy of reports, some investigators have used self-monitoring to assess the accuracy of self-reports. However, no published study has examined the relationships between diary measures and FTFIs or SAQs in the assessment of sexual behavior. Therefore, we sought to determine which of the two most commonly used modes of self-report elicited more accurate reports of sexual behavior by comparing them to data from a prospective diary. We predicted that, compared to FTFI, the SAQ would elicit less discrepant reports. We also predicted that accurate reports may be more likely when participants scored higher on a measure of erotophilia, and less likely when participants reported higher frequencies of sexual behavior.
P1: FTK/FNV
P2: FQJ/FLK,FFB
Archives of Sexual Behavior [asb]
QC: FTK PL111-335
June 24, 2000
10:25
312
Style file version Nov. 19th, 1999
Durant and Carey
METHOD Participants One-hundred ninety women were recruited from psychology courses at Syracuse University, and all agreed to take part in the study. Participants tended to be young (M = 19 yrs, SD = 1.4), Caucasian (75%), and in their first or second year of college (87%). Informed consent was obtained from all women, who received course credit for participating. Measures Marlowe Crown Social Desirability Scale (MC-SDS; Crown and Marlowe, 1964). The MC-SDS, a 33-item scale that assesses respondents’ tendency to answer in a socially desirable manner, was used. Scores for this sample were internally consistent (alpha = .77). Sexual Self-Disclosure Scale (SSDS; Catania et al., 1986). The SSDS, a 19-item measure, was used to assess participants’ comfort in reporting sexual information in a variety of situations. Scores were internally consistent (alpha = .92). Sexual Opinion Survey (SOS; Fisher et al., 1988). The SOS, a 21-item measure, assesses a respondent’s affective disposition to sexual cues. Erotophilia correlates with self-reports of sexual activity, amount of past sexual experience, and number of sexual partners (Fisher et al., 1988). Scores were internally consistent (alpha = .88). Demographic Questionnaire. A 9-item demographic questionnaire obtained information regarding the participant’s age, ethnicity, living, and relationship status. Health Behavior Questionnaire (HBQ). A 56-item survey requested information about nutrition, exercise, stress, alcohol, smoking, and sleep. The purpose of the HBQ was to contextualize the assessment of sexual behavior in a broader health framework. Sexual Behavior SAQ. A 13-item questionnaire obtained information about participants’ sexual history. Respondent’s reported their sexual behaviors in their lifetime and in the previous 2 months. Nonpenetrative (e.g., masturbation, petting) and penetrative (i.e., oral, vaginal, and anal sex) activities were assessed. Retrospective Self-Administered Questionnaire. The self-administered questionnaire requested information about the 8 sexual behaviors monitored on the diary card and 4 health behaviors (i.e., number of alcoholic and caffeinated beverages, number of cigarettes, and minutes of exercise). All sexual behavior questions were phrased as follows: “In the past 2 months, how many times have you ?” Participants were then asked to provide a frequency estimate of how many times they had engaged in each target behavior.
P1: FTK/FNV
P2: FQJ/FLK,FFB
Archives of Sexual Behavior [asb]
QC: FTK PL111-335
June 24, 2000
10:25
Assessment of Sexual Behavior
Style file version Nov. 19th, 1999
313
Retrospective Sexual Behavior Face-to-Face Interview. The FTFI was identical to the sexual behavior SAQ in content except that the FTFI was read aloud to participants by an interviewer. That is, the FTFI contained 8 questions that obtained information about participants’ sexual history in the previous 2 months, including questions about penetrative (i.e., oral, vaginal, and anal sex) and nonpenetrative (e.g., masturbation, petting) sexual behaviors as well as 4 health behaviors (i.e., number of alcoholic and caffeinated beverages, number of cigarettes, and minutes of exercise). Debriefing Questionnaire. An 18-item survey assessed participants’ comfort with the assessment, honesty in reporting, and difficulty in remembering their behaviors. In addition, the survey requested ratings of our credibility as well as the anonymity and privacy of the assessment. For example, one of the questions that measured our credibility asked: “Do you believe that the information that you have disclosed will be used to help develop a better understanding of women’s health behaviors?”
Diary Card Health-Behavior Diary Card A 8.5- × 15-cm card provided participants with a structured way to selfmonitor health behaviors. The card had 7 columns (one column for each day of the week) and 13 rows. The rows listed acronyms for 8 sexual behaviors [vaginal, oral, and anal intercourse with and without a barrier; masturbation; nonpenetrative activities; type of sexual partner (i.e., new, occasional, or main partner)], number of alcoholic and caffeinated beverages, number of cigarettes, and minutes of exercise. The card was designed to be easy and convenient to use, as well as cryptic in the event that it was lost or discovered. The card also included the respondent’s selfgenerated code name.
Interviewers The interviewers were five female research assistants who were trained thoroughly regarding sexual-behavior interviewing, research ethics, and confidentiality. The interviewers and participants in this study were purposefully similar in age and were matched on sex to help foster rapport and allow for comfort in reporting sensitive behaviors (for a review, see Catania, 1999). However, interviewers differed from interviewees in two important aspects: training and academic status. Interviewers were junior or senior psychology majors, whereas participants were freshmen or sophmores. Moreover, interviewers received rigorous training before interviewing participants. Specifically, interviewers first
P1: FTK/FNV
P2: FQJ/FLK,FFB
Archives of Sexual Behavior [asb]
QC: FTK PL111-335
June 24, 2000
10:25
Style file version Nov. 19th, 1999
314
Durant and Carey
observed the principal investigator model the structured interview, then practiced administrations with each other that were observed by the principal investigator, who provided immediate corrective feedback, and finally completed mock interviews until their performance was scored an average of 9 on the following scale: 1 = very poor to 10 = exceptional. The length and type of training was considered adequate because interviewers asked structured questions about behavioral frequencies and allayed concerns about the sensitive nature of the interview. All interviewers were blind to the study’s hypotheses. Procedure Recruitment and Session 1 Women from introductory-level psychology courses (N = 190) received a brief presentation regarding the study’s purpose and rationale during a regularly scheduled lecture time. Specifically, the study was introduced as a study of college women’s health and participants were told that their involvement would help the investigators learn more about women’s thoughts and feelings about important health issues. In addition, participants were told that their involvement would give them the opportunity to become more aware of their health behaviors. Participants were not told the entire purpose of the study in an effort to minimize volunteer bias (i.e., to limit the possibility that those who volunteer to participate may differ from those who do not). Interested women signed up for an overview session, during which they received (a) an explanation of the procedures; (b) provided written consent; and (c) completed the demographic questionnaire, HBQ, MCSDS, SSDS, and SOS. The questionnaires were ordered according to the degree of self-disclosure required (i.e., from least threatening questions to most threatening). After completing the questionnaires, participants received instruction in the self-monitoring procedures. Participants then generated code names to ensure the anonymity of the cards and were instructed to return a card each Monday to a locked drop box in a secure location. Session 2 After 8 weeks of self-monitoring, participants were randomly assigned to receive either the sexual behavior SAQ or an FTFI. The SAQ was completed by each participant independently, and both the SAQ and FTFI were administered in sound-attenuated rooms in a psychological services center. The interviewer read each item to the respondent and response formats were explained for each item. Participants responded verbally to the assessor, who recorded their responses on the survey form. Following the SAQ or FTFI, all participants completed a debriefing form regarding their experiences.
P1: FTK/FNV
P2: FQJ/FLK,FFB
Archives of Sexual Behavior [asb]
QC: FTK PL111-335
June 24, 2000
10:25
Style file version Nov. 19th, 1999
Assessment of Sexual Behavior
315
Final Session Participants returned in 1 week for an identical assessment. At the end of their session, participants were debriefed.
RESULTS Five of 190 participants were dropped from the study due to delinquent diary cards. All participants who were dropped from the study had either dropped or withdrawn from the class from which they were initially recruited. Of the 185 remaining participants, 120 were sexually active. Table I presents descriptive statistics of sexual behaviors monitored over the 8-week interval. As depicted in Table I, oral sex with barrier, anal sex, and anal sex with barrier were reported by an insufficient number of participants and were therefore dropped from subsequent analyses. To be considered sexually active, women had to report at least one incident of oral, vaginal, or anal sex (protected or unprotected) in the 8-week period prior to the study. These criteria were used to increase the likelihood of reports of sexual activity during the 8 weeks women were self-monitoring their health behaviors. Furthermore, sexually active women did not differ from inactive women in age, ethnicity, or any of the exploratory measures (i.e., SOS, SSDS, MC-SDS).
Preliminary Analyses To establish the equivalence of groups, t-tests were run on each sexual behavior variable on the diary cards and on the retrospective mode assessment data to identify any differences in the frequency in which participants engaged in specific Table I. Descriptive Statistics for Sexual Behavior Variables for Sexually Active Participants by Mode of Assessment (N = 120)
Sexual behaviors (over 8 weeks) Oral sex Oral sex with barriera Vaginal sex Vaginal sex with barrier Anal sexa Anal sex with barriera
Self-administered questionnaire (n = 61)
Face-to-face interview (n = 59)
M
SD
Mdn
Range
M
SD
Mdn
Range
4.6 0.2 3.5 4.6 — —
6.9 1.3 6.5 7.7 — —
2 0 0 1 — —
0–40 0–10 0–25 0–40 — —
5.4 0.2 6.2 5.7 — 0.1
6.9 1.2 11.9 9.1 — 0.3
3 0 1 3 — 0
0–35 0–8 0–60 0–60 — 0–2
Note: Table presents reports of frequency of engagement in each target sexual behavior. Higher values indicate more reports of sexual activity. With barrier = condom or dental dam; — = behaviors with no observations. a Behaviors not included in further analyses due to insufficient number of observations.
P1: FTK/FNV
P2: FQJ/FLK,FFB
Archives of Sexual Behavior [asb]
QC: FTK PL111-335
June 24, 2000
10:25
Style file version Nov. 19th, 1999
316
Durant and Carey
sexual activities by mode. There were no differences in reported frequencies for any of the sexual activities. Furthermore, there were no differences in the number of participants who were sexually active by mode of assessment. Primary Analyses Test–Retest Reliability Because distribution of sexual behavior scores were positively skewed (i.e., for each behavior, there was an overrepresentation of participants reporting 0s and 1s in their retrospective assessments) violating the assumption of normality that tests of significance of correlation coefficients are based, bootstrapping analyses (Lunneborg, 1985) were used to evaluate test–retest reliability. Furthermore, test– retest reliabilities for each variable reported in the 8-week period were computed using the intraclass correlation (ρ). There are several characteristics of bootstrapping that make it preferable in the evaluation of the magnitude of correlation coefficients. First, bootstrapping has no underlying assumption of normality. Second, bootstrapping is a nonparametric technique that simulates the effects of repeated sampling of the variables being analyzed from the population of interest. Bootstrapping accomplishes this by repeatedly sampling from the available sample and calculating a correlation from each resample (Efron and Tibshirani, 1993; Weinhardt et al., 1998a). This repeated sampling creates an approximation to the sampling distribution of the correlation. Hence, the magnitude of the observed correlation can be evaluated relative to what would be expected from the sample’s distribution. It is important to note that a key assumption of bootstrapping is that the sample looks like the population distribution, which is not necessarily normal. Both the FTFI and SAQ yielded strong test–retest reliability coefficients for all sexual variables (range ρ = .84 to .96, M ρ = .92) (see Table II). To determine whether the two modes differed, confidence intervals for each mode were calculated. Because the distribution of sexual behavior scores were positively skewed (i.e., violating the assumption of normality that tests of significance of correlation coefficients are based), bootstrapping analyses compared test–retest reliability across modes (Lunneborg, 1985). These analyses revealed no differences between conditions; that is, all confidence intervals for the mode comparison contained zero. Mode Differences in Accuracy A discrepancy score (diary card—retrospective data) was calculated to determine the accuracy of participants’ retrospective data. To evaluate whether there were differences in reporting of sexual behaviors by mode of assessment, a
1.8 1.7 1.9
.96 .84 .85
.55 to 2.0
.71 to 1.7
1.3 to 2.6
1.1 to 1.9
Confidence interval 95%
.95
.94
.94
.92
FTFI
1.3
1.2
2.0
1.5
M Fisher z
1.5 to 2.2
.93 to 2.5
.72 to 2.7
1.2 to 2.0
Confidence interval 95%
−.14 to 1.4
−.43 to 1.4
−1.4 to .99
−.50 to .63
Mode comparison 95% CI
June 24, 2000 10:25
Note: Reliability coefficients are intraclass correlations; SAQ = data obtained with a self-administered questionnaire; FTFI = data obtained in a face-to-face interview; CI = confidence interval; M Fisher z = mean Fisher z transformations; all confidence intervals are z confidence intervals.
1.6
M Fisher z
.91
SAQ
(N = 185)
PL111-335
How many partners have you had sex with (i.e., oral, anal, or vaginal sex)? How many times have you had vaginal sex with a condom? How many times have you had vaginal sex without a condom? How many times have you had oral sex without a barrier (i.e., condom or dental dam)?
Behavior (in the previous 2 months)
(n = 89)
P2: FQJ/FLK,FFB
(n = 96)
Archives of Sexual Behavior [asb]
Table II. Test–Retest Correlation Coefficients and Reliability Comparison, by Mode
P1: FTK/FNV QC: FTK Style file version Nov. 19th, 1999
Assessment of Sexual Behavior 317
P1: FTK/FNV
P2: FQJ/FLK,FFB
Archives of Sexual Behavior [asb]
QC: FTK PL111-335
June 24, 2000
10:25
Style file version Nov. 19th, 1999
318
Durant and Carey
two-sample Wilcoxon rank–sum (Mann–Whitney) test was used. This test was used because (a) the discrepancy scores were positively skewed, and (b) transformations of the data could not normalize the distributions. Only sexually active participants (n = 120) were used in these analyses because participants who did not engage in any sexual behaviors would inflate the number of zero discrepancy scores. Table III shows that participants’ discrepancy scores in the FTFI condition were significantly larger than those in the SAQ condition for vaginal sex with a barrier. Discrepancy scores for oral sex and vaginal sex without a barrier did not differ between modes.
Exploratory Analyses Predicting the Discrepancy Scores Groups did not differ on anonymity, privacy, erotophilia, sexual selfdisclosure, social desirability, or perceptions of our credibility (all ps > .10). To explore whether these variables predicted discrepancy scores (i.e., accuracy in reporting), a negative binomial regression model was used. This analysis was most appropriate given the overrepresentation of 0s, 1s, and 2s for reports of frequency of sexual behaviors (Gardner et al., 1995). The model used the absolute value of the discrepancy scores and was restricted to participants who had participated in each behavior. Three separate multiple regression models were constructed for each of the discrepancy scores (i.e., vaginal sex, vaginal sex with barrier, oral sex). The carriers included assessment mode, SOS, SSDS, MC-SDS, perceived credibility, anonymity, and privacy, which were entered into the regression equation in a hierarchical fashion. Mode (dummy coded) was entered first, followed by anonymity, privacy, credibility, the SOS, SSDS, and MC-SDS. None of the three models accounted for a significant amount of the variance (vaginal sex, [X 2 (7) = 4.80, p = .68], vaginal sex with barrier [X 2 (7) = 3.52, p = .83], and oral sex discrepancy score model [X 2 (7) = 9.18, p = .24]). Discrepancy Score in Relation to Frequency of Behavior To explore whether discrepancy scores increased as frequency of behavior increased, pairwise correlations were calculated between the frequency of each behavior and the participant’s discrepancy score. Absolute discrepancy scores were used, the sample was restricted to sexually active participants, and analyses were conducted separately for oral sex and protected and unprotected vaginal sex. For all three behaviors, as frequency of sexual behaviors increased, so did the discrepancy score, r s = .59, .52, .73 ( ps < .001).
50 25 35
Oral sex Vaginal sex Vaginal sex with barrier
0.3 2.8 −0.2
M 6.0 8.9 5.3
SD 0 0 0
Mdn
n 52 32 39
Range −21 to 22 −6 to 35 −16 to 18 2.1 1.0 2.4
M 6.2 8.9 5.6
SD 1 0 −1
Mdn
−11 to 24 −22 to 27 −21 to 9
Range
−1.38 .21 2.07
z
.17 .83 .03
p
June 24, 2000
Note: Discrepancy scores are differences found between the diary card reports of sexual behavior and mode assessment reports of sexual behavior (i.e., diary card—mode assessment). Positive discrepancy scores indicate underreporting in retrospective mode assessment; negative discrepancy scores indicate overreporting in retrospective mode assessment. With barrier = condom.
n
Discrepancy scores
Discrepancy score comparison
PL111-335
Face-to-face interview (n = 59)
P2: FQJ/FLK,FFB
Self-administered questionnaire (n = 61)
Archives of Sexual Behavior [asb]
Table III. Descriptive Statistics and Mode Comparison of Discrepancy Scores of Sexually Active Participants by Behavior
P1: FTK/FNV QC: FTK 10:25
Assessment of Sexual Behavior
Style file version Nov. 19th, 1999
319
P1: FTK/FNV
P2: FQJ/FLK,FFB
Archives of Sexual Behavior [asb]
QC: FTK PL111-335
June 24, 2000
10:25
320
Style file version Nov. 19th, 1999
Durant and Carey
DISCUSSION Four findings were obtained. First, analyses indicated high levels of test– retest reliability on all sexual behaviors, regardless of mode of administration. These findings converge with those of Kalichman et al. (1997), who found that the SAQ and FTFI were equally reliable methods of obtaining self-reports of sexual behavior. Second, compared to FTFIs, SAQs resulted in a lower discrepancy score for protected vaginal sex; mode effects were not observed for unprotected vaginal or unprotected oral sex. These differences may indicate that SAQs elicit more accurate results than FTFIs. Perhaps SAQs allow participants to proceed at their own pace with the aid of a visual representation of the question on paper, leading to less discrepant responses. The null findings for both unprotected vaginal and unprotected oral sex attenuate confidence in this interpretation. The absence of group differences on these latter variables may result from their lower frequency. Alternatively, patterned social behaviors in young women’s lives (e.g., weekend parties) may make sexual behaviors easier to recall regardless of mode of assessment. Young, well-educated women in a university setting may also be more comfortable responding to questions about sexuality relative to women recruited through community outreach (e.g., Carey et al., 2000). Finally, because participants perceived the SAQ condition and FTFI condition as equally credible, private, and anonymous (i.e., optimal conditions for accurate reports), hypothesized mode differences did not emerge in this context. We hypothesize that participants may have based their ratings on all procedures instead of just their mode of assessment. That is, perhaps participants included the 2-month self-monitoring period (in which participants were identified by code name only) in their ratings, making their perceptions of the assessment modes more similar. Third, reports of “safer sex” (i.e., protected sex) activities were reported less accurately than riskier events. This finding contrasts with those obtained in previous studies with community samples of gay men (Downey et al., 1995) and men and women who have had homosexual, bisexual, or sex-worker partners (Boekeloo et al., 1994). However, women in our sample were not engaging in a number of risky sexual behaviors (e.g., anal sex), which limited the number of high-risk behaviors analyzed. Finally, our efforts to predict self-report accuracy with situational (i.e., privacy, credibility, anonymity) or person (e.g., erotophilia, sexual self-disclosure, social desirability) variables were unsuccessful. The relatively brief period of selfmonitoring may have precluded higher frequencies of reported behaviors in this sample, limiting power. Future research can address this issue. Like Downey et al. (1995), we found that self-report accuracy decreased as frequency of sexual behaviors increased. This suggests that behaviors that were performed more often may be more difficult to recall.
P1: FTK/FNV
P2: FQJ/FLK,FFB
Archives of Sexual Behavior [asb]
QC: FTK PL111-335
June 24, 2000
10:25
Assessment of Sexual Behavior
Style file version Nov. 19th, 1999
321
The findings of this study need to be interpreted in light of its limitations. First, the use of self-monitoring is an imperfect index of accuracy. Second, the assessment interval may have been too brief to elicit sufficient data, or to afford sufficient time for subtle memory decay. It might have been necessary to have participants self-monitor their health behaviors for a longer period of time. Finally, this study was unable to assess reactivity to concurrent self-monitoring on participants’ sexual behavior reports over the 8-week interval. That is, we were unable to assess whether sexual behavior reports were affected by participants’ knowledge that their responses could be compared to self-monitoring data (that covered the same period) collected prior to their retrospective assessment. Despite these limitations, our results suggest that both FTFIs and SAQs obtain stable estimates of sexual behavior over brief intervals, but that SAQs may elicit more accurate responses, at least for some behaviors. Because SAQs are also less expensive and less labor intensive, they may be preferred in many settings. The current study contributes to the literature by attempting to validate participants’ self-reports with contemporaneous self-monitoring. Additional research is needed to identify the situational and individual determinants of self-accuracy. ACKNOWLEDGMENTS Research supported by grants from the National Institute of Mental Health to Michael P. Carey (K02-MH01582 and RO1-MH54929). The authors thank John Gleason and Christopher Gordon for statistical consultation; Jennifer Alvarez, Jennifer Chernowski, Nikia Hearst, Lori Mothersell, and LaToya Shakes, who entered and helped collect data; and Lance Weinhardt for his helpful comments on an earlier draft of this paper. Postal correspondence regarding this article may be sent to Michael P. Carey, Department of Psychology, 430 Huntington Hall, Syracuse University, Syracuse, NY 13244-2940; electronic mail to
[email protected]. REFERENCES Boekeloo, B. O., Schiavo, L., Rabin, D. L., Conlon, R. T., Jordan, C. S., and Mundt, D. J. (1994). Self-reports of HIV risk factors by patients at a sexually transmitted disease clinic: Audio vs. written questionnaires. Am. J. Pub. Health 84: 754–760. Bradburn, N., and Sudman, S. (1979). Improving Interview Method and Questionnaire Design. Washington, DC: Jossey Bass. Carey, M. P., Braaten, L. S., Maisto, S. A., Gleason, J. R., Forsyth, A. D., Durant, L. E., and Jaworski, B. C. (2000). Using information, motivational enhancement, and skills training to reduce the risk of HIV infection for low-income urban women: A second randomized clinical trial. Health Psychol. 19: 3–11. Catania, J. A. (1999). A framework of conceptualizing reporting bias and its antecedents in interviews assessing human sexuality. J. Sex Res. 36: 25–38. Catania, J. A., McDermott, L. J., and Pollack, L. M. (1986). Questionnaire response bias and face-toface interview sample bias in sexuality research. J. Sex Res. 22: 52–72.
P1: FTK/FNV
P2: FQJ/FLK,FFB
Archives of Sexual Behavior [asb]
322
QC: FTK PL111-335
June 24, 2000
10:25
Style file version Nov. 19th, 1999
Durant and Carey
Crowne, D. P., and Marlowe, D. (1964). The Approval Motive: Studies in Evaluative Dependence. New York: Wiley. Czaja, R. (1987). Asking sensitive behavioral questions in telephone interviews. Int. Q. Comm. Hlth. Ed. 8: 23–32. Downey, L., Ryan, R., Roffman, R., and Kulich, M. (1995). How could I forget? Inaccurate memories of sexually intimate moments. J. Sex Res. 32: 177–191. Efron, B., and Tibshirani, R. J. (1993). An Introduction to the Bootstrap. New York: Chapman & Hall. Fisher, W. A., Byrne, D., White, L. A., and Kelley, K. (1988). Erotophobia–erotophilia as a dimension of personality. J. Sex Res. 25: 123–151. Gardner, W., Mulvey, E. P., and Shaw, E. C. (1995). Regression analyses of counts and rates: Poisson, overdispersed poisson, and negative binomial models. Psychol. Bull. 118: 392–404. Kalichman, S. C., Kelly, J. A., and Stevenson, L. Y. (1997). Priming effects of HIV risk assessments on related perceptions and behavior: An experimental field study. AIDS Behav. 1: 3–8. Lunneborg, C. E. (1985). Estimating the correlation coefficient: The bootstrap approach. Psychol. Bull. 98: 209–215. McFall, R. M. (1970). Effects of self-monitoring on normal smoking behavior. J. Consult. Clin. Psych. 35: 135–142. McLaws, M., Oldenburg, B., Ross, M. W., and Cooper, D. A. (1990). Sexual behaviour in AIDS-related research: Reliability and validity of recall and diary measures. J. Sex Res. 27: 265–281. Miller, H. G., Turner, C. F., and Moses, L. E. (1990). Methodological issues in AIDS surveys. In Miller, H. G., Turner, C. F., and Moses, L. E. (eds.), AIDS: The Second Decade. Washington, DC: National Academy Press. Reading, A. E. (1983). A comparison of the accuracy and reactivity of methods of monitoring male sexual behavior. J. Behav. Assess. 5: 11–23. Strecher, V. J., Becker, M. H., Clark, N. M., and Prasada-Rao, P. (1989). Using patients’ descriptions of alcohol consumption, diet, medication compliance, and cigarette smoking: The validity of self-reports in research and practice. J. Gen. Intern. Med. 4: 160–166. Tourangeau, R., and Smith, T. W. (1996). Asking sensitive questions: The impact of data collection mode, question format, and question context. Publ. Opin. Q. 60: 275–304. Webb, G. R., Redman, S., Gibberd, R. W., and Sanson-Fisher, R. W. (1991). The reliability and stability of a quantity-frequency method and a diary method of measuring alcohol consumption. Drug. Alcohol. Depend. 27: 223–231. Weinhardt, L. S., Carey, M. P., Maisto, S. A., Carey, K. B., Cohen, M. M., and Wickramasinghe, S. M. (1998). Reliability of the timeline followback sexual behavior interview. Ann. Behav. Med. 20: 25–30. Weinhardt, L. S., Forsyth, A. D., Carey, M. P., Jaworski, B. A., and Durant, L. E. (1998). Reliability and validity of self-report measures of HIV-related sexual behavior: Progress since 1990 and recommendations for research and practice. Arch. Sex Behav. 27: 155–180.
P1: FVI/FLK
P2: FVI
Archives of Sexual Behavior [asb]
PL111-336
June 17, 2000
11:10
Style file version Nov. 19th, 1999
Archives of Sexual Behavior, Vol. 29, No. 4, 2000
Methodologic Concerns Regarding Estimates of Physical Violence in Sexual Coercion: Overstatement or Understatement? Christine Alksnis,1 Serge Desmarais,1,3 Charlene Senn,2 and Nichola Hunter2
Self-report measures of sexual violence that ask women whether they have experienced threats of physical violence have attracted criticism in recent years; detractors claim that these measures lead researchers to overestimate the prevalence of sexual violence. Our study explored this issue by collecting data on the prevalence of threats versus force in the context of sexual aggression. Female undergraduates at two universities (n1 = 69; n2 = 111) were asked about their experiences with sexual coercion using a revised version of the Sexual Experiences Scale (Koss and Gidycz, 1985). Four of the original items were modified to distinguish between sexual contact that occurred as a result of a perpetrator using physical force and sexual contact that occurred because a perpetrator threatened physical violence. Analyses of the revised items revealed that the use of physical force was at least as likely as threats and that for some types of sexual acts, physical force was actually more likely than verbal threats. Furthermore, prevalence figures for three of the four types of sexual acts considered were not significantly altered by collapsing threat of force with use of force. Implications for future research on women’s experiences of sexual coercion are discussed. KEY WORDS: sexual coercion; force versus threats; Sexual Experience Scale, methodology.
1 University
of Guelph, Guelph, Ontario, Canada. of Windsor, Windsor, Ontario, Canada. 3 To whom correspondence should be addressed at Department of Psychology, University of Guelph, Guelph, Ontario, Canada, N1G 2W1; e-mail:
[email protected]. 2 University
323 C 2000 Plenum Publishing Corporation 0004-0002/00/0800-0323$18.00/0 °
P1: FVI/FLK
P2: FVI
Archives of Sexual Behavior [asb]
324
PL111-336
June 17, 2000
11:10
Style file version Nov. 19th, 1999
Alksnis, Desmarais, Senn, and Hunter
INTRODUCTION When sexual violence first became a legitimate subject for empirical study, it became clear that official estimates of prevalence (e.g., police reports, crime surveys) greatly miscalculated the magnitude of the problem. In the United States, it is estimated that between 10% and 50% of rapes are reported to the police (Koss, 1985). In Canada, a national survey on violence against women indicated that only 6% of rape victims ever reported their experience to police (Statistics Canada, 1993). Furthermore, researchers found that sexual assaults by strangers are overrepresented in official statistics, whereas incidents of sexual violence by acquaintances often go unreported (e.g., Yurchesyn et al., 1992). Given the systematically biased and incomplete nature of official statistics, many researchers have acknowledged the superiority of self-report measures for obtaining prevalence estimates of sexual assault and coercion. The methodologic differences associated with self-report studies likely account for many of the discrepancies between estimates obtained by social scientists and those obtained by legal authorities. The problem of underreporting that occurs in the legal context appears to be reduced in sociological and psychological studies because the latter types of investigations often entail surveying a more representative sample of women regarding their sexual experiences, asking participants about a wide range of coercive sexual experiences (e.g., forced sexual touching; coerced vaginal, oral, and anal sex) and querying about specific acts experienced instead of using emotionally charged descriptors such as rape and sexual assault, terms that carry a connotation of criminality that is frequently resisted by respondents. However, such methodologic changes have not been lauded by all observers; one of the best-known and most widely used self-report measures of sexual coercion, the Sexual Experiences Survey (SES; Koss et al., 1982, 1987), has been the target of numerous critiques in the popular media. According to critics, the SES and other similar instruments: (a) use definitions of sexual violence and coercion that are too broad, (b) were designed to generate high estimates of sexual violence and coercion for the purpose of influencing policy makers, (c) generate estimates of sexual assault that do not agree with official estimates, and (d) promote a conceptualization of sexual coercion that overemphasizes the role of patriarchy while ignoring both the occurrence of female-to-male violence and the existence of alternative explanations for violence (Fekete, 1994; Gilbert, 1993a; Pearson, 1997). Some critics have gone as far as to argue that feminist researchers use the SES for the express purpose of producing biased research findings that in turn will aid feminists’ efforts to distort the facts about sexual violence. For instance, it has been speculated that feminists are motivated to create a “rape crisis” in order to obtain federal funding and grants; Gilbert (1993b) stated that the rape education and prevention services for which feminist groups were lobbying were in fact designed “to deal with an epidemic of date rape that does not really exist” (p. A18). Others have suggested that feminists devised
P1: FVI/FLK
P2: FVI
Archives of Sexual Behavior [asb]
PL111-336
Physical Violence in Sexual Coercion
June 17, 2000
11:10
Style file version Nov. 19th, 1999
325
a rape crisis in an effort to facilitate more widespread acceptance of feminist ideology: “Somebody is finding this rape crisis, and finding it for a reason. Asserting the prevalence of rape lends urgency, authority, to a broader critique of culture” (Roiphe, 1993, p. 55). The argument that social scientists’ definitions of sexual coercion have become too inclusive—resulting in inflated prevalence and incidence figures for sexual assault generally, and rape in particular (Gilbert, 1991; Roiphe, 1993)— provided the impetus for this investigation. It appears that the operational definitions that form the basis of feminist-oriented psychological and sociological investigations of sexual assault are frequently misrepresented; consider the following quotes: “[t]oday’s definition of date or acquaintance rape stretches beyond acts of violence or physical force. According to common definitions of date rape, even verbal coercion or manipulation constitutes rape” (Roiphe, 1993, pp. 66–67) and “with such a sweeping definition of rape, I wonder how many people there are, male or female, who haven’t been date-raped at one point or another. People pressure and manipulate and cajole each other into all sorts of things all the time” (Roiphe, 1993, p. 79). Although it is certainly true that many researchers view some forms of verbal pressure as constituting sexual coercion, the label “rape” has only been applied to a limited number of acts. When communicating their findings, Koss, DeKeseredy, and others repeatedly emphasize that their prevalence figures for rape and attempted rape are based on the subset of survey items that describe acts that reflect the legal definition of rape used in most North American jurisdictions—items focusing on attempted and completed anal, oral, or vaginal sex that involved the use of force, the threat of force, or the inability to consent due to intoxication. A closer look at some of the critics’ objections reveals that their claims are based on either misconstrual or distortion of the research findings. By way of illustration, Koss and colleagues (1987) found that 27.5% of a national sample of college women reported an experience of rape or attempted rape since the age of 14. In relating this finding, one writer paraphrased: “sex is, in one in four cases, against your will” (Roiphe, 1993, p. 21). The use of the word “cases” implies that approximately one-quarter of women’s intercourse experiences are forced. Clearly, such a statement is not equivalent to Koss’s assertion that one in four women experience sexual violence in the form of rape or attempted rape at least once in their lives. However, given that most people learn about the findings of scientific investigations through newspapers and magazines rather than through scientific journals, it is important for researchers to consider how they may unintentionally contribute to having their work misrepresented in the popular media. In the SES, questions designed to tap the occurrence of rape and other forms of sexual violence consider both threats of physical violence and acts of physical violence together (Porter and Critelli, 1992). The goal of this study was to reevaluate this strategy. This paper focuses on the four questionnaire items that combine the use of force or the threat of force—three of these items pertain to attempted and
P1: FVI/FLK
P2: FVI
Archives of Sexual Behavior [asb]
PL111-336
June 17, 2000
11:10
326
Style file version Nov. 19th, 1999
Alksnis, Desmarais, Senn, and Hunter
completed anal, oral, or vaginal sex, whereas the fourth item is concerned with nonpenetrative sex acts. It is interesting to note that in an earlier version of the SES (Koss and Oros, 1982), there were separate items to assess threat of force and the use of force in the context of rape and attempted rape. In later versions (Koss and Gidycz, 1985; Koss et al., 1987) these items were amalgamated, but no clear rationale was provided for this strategy. Many people are inclined to believe that sexually coercive acts can be arranged in a hierarchy of “seriousness,” and that more serious acts, such as physical aggression, are not as frequent as verbal acts, such as threats of violence. Cook (1995) showed that both male and female respondents anticipated that “the use of verbal threats of harm” would be more common than the use of “force to try to obtain or to obtain intercourse” (p. 191). By combining threats of force and actual force in the survey items, Koss may inadvertently have left the door open for critics to assert that estimates of the number of women who experience physical violence have been inflated via confounding of more frequent events and less frequent ones. Yet some researchers have shown that actual physical harm is at least as common as threats of violence, if not more so (e.g., Miller and Marshall, 1987; Muehlenhard and Linton, 1987), suggesting that Koss’s strategy of considering threats and actual violence together may be methodologically sound. This investigation was carried out with an eye toward establishing whether the current version of the SES yields an overstatement or an understatement of the prevalence of physical violence in sexual coercion. Concerns regarding the overestimation of the physical violence that accompanies sexual assault would only be warranted if two conditions were met: (a) women report the experience of threat of force at a higher frequency than they report experiencing force, and (b) the collapsing of threat and actual force increases the number of women who report each of the four sex acts considered.
METHOD Participants Two independent samples of female students from two Southern Ontario universities were recruited to participate in a study of dating and sexual behavior. In the first sample, the women were enrolled in introductory psychology courses and given course credit for their participation. These 69 students ranged in age from 18 to 26 years (M1 = 20.1 years, SD1 = 1.6 years). The 111 students from the second university were recruited either in coed residence meetings (71%) or from introductory psychology courses (29%) and were of comparable age (M2 = 19.8 years, SD2 = 2.1 years). The vast majority of students reported that their sexual orientation was heterosexual, with the exception of two women in sample 2 (one was bisexual and one indicated that she was not sure).
P1: FVI/FLK
P2: FVI
Archives of Sexual Behavior [asb]
PL111-336
June 17, 2000
Physical Violence in Sexual Coercion
11:10
Style file version Nov. 19th, 1999
327
Measures Students in sample 2 were asked about sexually coercive experiences that had occurred since the age of 14 (identical to Koss’s procedure). Students in sample 1 were asked about experiences that had been perpetrated by male dating partners. The vast majority of respondents in sample 1 indicated that they had begun dating after the age of 13, making the data obtained from two sites comparable to Koss’s. The measure of sexual coercion used was a modification of the SES (Koss et al., 1987). The four items in the later version of the scale that combined physical force and the threat of force asked about sexual acts that comprised attempted vaginal rape, vaginal rape, anal/oral sex, and sex play. Each of these four items were split into two new items that separated sexual acts that involved physical force actually being used against the woman from sexual acts occurring as a result of the threat of force. For example, one of the original items asked: “Have you had sexual intercourse when you didn’t want to because a man threatened or used some degree of physical force (twisting your arm, holding you down, etc.) to make you?”4 whereas the two revised items asked: (1) “Have you had sexual intercourse when you didn’t want to because a man used some degree of physical force (twisting your arm, holding you down, etc.) to make you?” and (2) “Have you had sexual intercourse when you didn’t want to because a man threatened to use physical force (twisting your arm, holding you down, etc.) to make you?” The revised scale consisted of 14 items instead of 10. Internal consistency reliabilities for the modified scale were comparable to Koss and Gidycz’s (1985) finding of α = .74 for their national sample (αSample1 = .71, αSample2 = .80; αcombined = .77). Women who experience threat of physical violence and women who experience force do not necessarily comprise independent groups. In this sample, the proportion of women who have experienced both threat and force is 54% for anal/oral sex, 58% for attempted vaginal rape, 82% for vaginal rape, and 83% for sex play. It is not known whether the threat and force experiences reported by any given woman were part of the same sexual assault incident. However, the goal of this paper was to determine the prevalence of the various methods by which women have been sexually coerced, not to discover whether those methods are mutually exclusive or to uncover the number of different sexual assault incidents experienced by each respondent. As a result, the potential for double-counting incidents via revision of the SES items was not perceived as problematic. 4 The
other three items tapped sexual acts that comprised attempted vaginal rape—“Have you had a man attempt sexual intercourse (get on top of you, attempt to insert his penis) when you didn’t want to by threatening or using some degree of force (twisting your arm, holding you down, etc.) but intercourse did not occur?” [emphasis in original]; sexual assaults which did not involve rape—“Have you had sex play (fondling, kissing, or petting, but not intercourse) when you didn’t want to because a man threatened or used some degree of physical force (twisting your arm, holding you down, etc.) to make you?”; and anal/oral sex—“Have you had sex acts (anal or oral intercourse or penetration by objects other than the penis) when you didn’t want to because a man threatened or used some degree of physical force (twisting your arm, holding you down, etc.) to make you?”
P1: FVI/FLK
P2: FVI
Archives of Sexual Behavior [asb]
PL111-336
June 17, 2000
11:10
328
Style file version Nov. 19th, 1999
Alksnis, Desmarais, Senn, and Hunter
Procedure The test materials were administered to respondents individually by a female researcher in groups of no more than six women. Privacy was ensured through spaced seating. Participants were given both a written and an oral description of the study, and then filled out the survey. No time constraints were imposed. Due to the sensitive nature of the questions, participants were provided with information about local support agencies that they could access if taking part in the study caused them to feel distressed. RESULTS Although it seemed profitable to combine the two subsamples in order to increase the statistical power of the analyses conducted, the differences in the recruitment strategies employed at the two sites were seen as potentially problematic, conceivably requiring separate analyses for each group. Chi-square tests were conducted on the responses provided by the two subsamples in an effort to determine whether the apparent differences between them could be attributed to the vagaries of sampling from the same population or whether it was more likely that the samples came from different populations. Each of the four sexual acts was considered under three conditions [(1) threat, (2) physical force, and (3) threat or force], resulting in 12 comparisons. Using Bonferroni’s correction, a more stringent criterion alpha of .004 was adopted in order to account for the multiple χ 2 tests. These tests revealed no statistically significant differences between the sites, indicating that the two samples were drawn from the same population and that it would be appropriate to treat them as a single sample for analytical purposes. For completeness of presentation, data from the two sites were also analyzed separately; however, the reader is cautioned that these subsample analyses may not produce results identical to those obtained with the larger sample. Before focusing on the question of the frequency of threats of force and actual force, it seemed useful to compare the overall rates of all types of sexual victimization obtained in this sample with those found in Koss and coworkers’ (1987) national sample. For the current (combined) sample, frequencies for the modified items were made comparable to Koss’s original items by collapsing across the revised force and threat of force items. Respondents were then classified according to the highest degree of sexzual victimization they had reported using Koss’s five-level categorization: no sexual aggression/victimization, sexual contact, sexual coercion, attempted rape, and rape. Note that for the purposes of this preliminary analysis, the label “rape” refers to a wider range of phenomena than is captured by the individual “vaginal rape” items (i.e., the broader category includes the occurrence of any type of anal, oral, or vaginal sex involving the use of force, the threat of force, or the inability to consent due to intoxication, all of which
P1: FVI/FLK
P2: FVI
Archives of Sexual Behavior [asb]
PL111-336
June 17, 2000
11:10
Style file version Nov. 19th, 1999
Physical Violence in Sexual Coercion
329
Table I. Prevalence Rate Percentages for Sexual Aggression and Victimization Aggression or victimization level
Sample 1 (n = 69)
Sample 2 (n = 111)
Combined sample (n = 180)
Koss’s (1987) national sample
No sexual aggression or victimization Sexual contact Sexual coercion Attempted rape Rape
36.2% (25)
33.3% (37)
34.4% (62)
46.3%∗∗
10.1% (7) 18.8% (13) 11.6% (8) 23.2% (16)
9.0% (10) 18.0% (20) 9.0% (10) 30.6% (34)
9.4% (17) 18.3% (33) 10.0% (18) 27.8% (50)
14.4% 11.9%∗ 12.1% 15.4%∗∗∗
∗ p < .006. ∗∗ p < .001. ∗∗∗ p < .0001.
are customarily included in legal definitions of rape in North American jurisdictions). The prevalence rates for the five categories are presented in Table I. Z -tests revealed that the proportion of women in the current sample whose most serious victimization involved sexual coercion or any form of rape was significantly higher than the proportions reported by Koss (18.3% vs. 11.9%, p < .006 for coercion, and 27.8% vs. 15.4%, p < .0001 for rape), whereas the proportion reporting no sexual victimization was significantly lower (34.4% vs. 46.3%, p < .001). Next, a series of one-sample z-tests were conducted in order to ascertain whether the rates of unwanted sexual contact reported in the current samples on the four specific items of interest were comparable to Koss and coworkers’ (1987) national sample. As before, comparisons were made using the frequencies obtained by collapsing across force and threat of force in the current (combined) sample. The tests presented in Table III revealed that the proportion of women in the current sample who had experienced each act was significantly higher than the proportion reported by Koss. Twenty-three percent of women in the current sample reported sex play occurring after force or threats of force, in contrast to 13% of women in Koss’s sample ( p < .0001). The proportions were 21.8% vs. 15% for attempted vaginal rape ( p < .008), 14.4% vs. 9% for vaginal rape ( p < .008), and 12.2% vs. 6% for anal/oral sex ( p < .0005). Cochran’s Q-tests were then used to ascertain whether the number of respondents reporting threats of violence was significantly higher than the number who reported physical violence. An alpha of .004 was adopted to account for the fact that 12 threat vs. force comparisons were made (i.e., the threat vs. force comparison was made for each of the four acts on the combined sample as well as on each subsample separately). The results presented in Table II reveal that, for the combined sample, the use of physical force was reported by a higher proportion of women than were threats of force for vaginal rape (13.3% and 6.1%, respectively; Q (1) = 9.9, p < .002) and for sex play (21.1% and 12.8%, respectively; Q (1) = 9.8, p < .002), but not for attempted vaginal rape or anal/oral sex. When the groups were considered separately, the pattern of responses on the vaginal rape
P1: FVI/FLK
P2: FVI
Archives of Sexual Behavior [asb]
PL111-336
June 17, 2000
11:10
330
Style file version Nov. 19th, 1999
Alksnis, Desmarais, Senn, and Hunter Table II. Percentage of Sample Endorsing Each Item (Threat vs. Use of Force) Sample 1 (n = 69)
Sample 2 (n = 111)
Combined sample (n = 180)
Sex play—threat of force Sex play—use of force
5.8% (4) 21.7% (15) Q (1) = 11.0∗∗∗
17.1% (19) 20.7% (23) Q (1) = 1.3
12.8% (23) 21.1% (38) Q (1) = 9.8∗∗
Attempted vaginal rape—threat of force Attempted vaginal rape—use of force
13.2% (9) 22.1% (15) Q (1) = 4.5
15.5% (17) 10.8% (12) Q (1) = 1.7
14.6% (26) 15.2% (27) Q (1) = 0.04
Vaginal rape—threat of force Vaginal rape—use of force
2.9% (2) 13.0% (9) Q (1) = 7.0
8.1% ( 9) 13.5% (15) Q (1) = 3.6
6.1% (11) 13.3% (24) Q (1) = 9.9∗∗
Anal/oral sex acts—threat of force Anal/oral sex acts—use of force
1.4% (1) 4.3% (3) Q (1) = 2.0
10.8% (12) 11.7% (13) Q (1) = 0.1
7.2% (13) 8.9% (16) Q (1) = 0.6
Sex act
∗∗ p < .002. ∗∗∗ p < .001.
item failed to reach statistical significance, although the direction of the effect for each group was in the expected direction (i.e., a higher proportion of women experienced force than experienced threats of force). Further inspection of the two subsamples revealed significant differences in the distribution of responses for threat of force and use of force in the context of sex play for sample 1 (21.7% experienced force, whereas 5.8% experienced threats of force; Q (1) = 11.0, p < .001) but not for sample 2.5 In all other comparisons, women were as likely to have been subjected to actual force as to have been threatened with force. Cochran’s Q-tests were also used to compare the number of women who would be categorized as having experienced each act of sexual aggression if threat were included in the criterion (as per the more recent version of the SES) versus the situation in which force alone was the criterion. Again, a conservative alpha of .004 was employed to account for the multiple comparisons of threat and/or force versus force alone. Table III presents the frequency data and results of these analyses. Only for the attempted vaginal rape categorization did the inclusion of threat result in a significantly higher prevalence rate than would have been the case if actual force was the criterion (21.8% vs. 15.2%; Q (1) = 11.0, p < .001). The same pattern of results was found in sample 2 (20.0% vs. 10.8%; Q (1) = 10.1, p < .002), but not in sample 1.6 For the remaining comparisons, the inclusion of threat in the categorization procedure produced no significant increase in the proportion of women classified as having experienced each act. 5 As
alluded to earlier, this apparent difference between the subsamples is attributable to sampling fluctuations; chi-square tests indicate that this difference is not statistically significant. 6 Once again, this apparent difference between the subsamples is attributable to sampling fluctuations.
P1: FVI/FLK
P2: FVI
Archives of Sexual Behavior [asb]
PL111-336
June 17, 2000
11:10
Style file version Nov. 19th, 1999
Physical Violence in Sexual Coercion
331
Table III. Percentage of Sample Meeting Criteria for Each Sex Act when Threats are Included vs. Excluded Sex act Sex play Force only Force or threats Attempted vaginal rape Force only Force or threats Vaginal rape Force only Force or threats Anal/oral sex acts Force only Force or threats
Sample 1 (n = 69)
Sample 2 (n = 111)
Combined sample (n = 180)
Koss’s (1987) national sample
21.7% (15) 21.7% (15) Q (1) = 0.0
20.7% (23) 24.3% (27) Q (1) = 4.0
21.1% (38) 23.3% (42) Q (1) = 4.0
13%∗∗∗
22.1% (15) 24.6% (17) Q (1) = 1.0
10.8% (12) 20.0% (22) Q (1) = 10.1∗∗
15.2% (27) 21.8% (39) Q (1) = 11.0∗∗∗
15%∗
13.0% (9) 13.0% (9) Q (1) = 0.0
13.5% (15) 15.3% (17) Q (1) = 2.0
13.3% (24) 14.4% (26) Q (1) = 2.0
9%∗
4.3% (3) 4.3% (3) Q (1) = 0.0
11.7% (13) 17.1% (19) Q (1) = 6.0
8.9% (16) 12.2% (22) Q (1) = 6.0
6%∗∗∗
∗ p < .01. ∗∗ p < .002. ∗∗∗ p < .001.
DISCUSSION There have been numerous responses to the allegations that feminist investigators use overly inclusive definitions of sexual coercion and methodologically flawed studies to advance a hidden agenda (DeKeseredy, 1996; DeKeseredy and MacLeod, 1997; Koss and Cook, 1993). The purpose of this paper was to focus on a specific definitional issue by examining the pattern of women’s experiences of threats of force versus actual force. The findings of the present study suggest that Koss and colleagues may be correct in merging threats of force with use of force. Counter to common wisdom, it appears that collapsing threat and force in research measures does not lead investigators to conclude that force is more common than it is in reality (Muehlenhard and Linton, 1987; see also Rapaport and Burkhart, 1984, who used male self-reports). The first set of analyses revealed that threat of force in coercive sexual interactions was no more common than the actual use of force. In fact, the differences obtained were in the opposite direction for the sex play item (for the combined sample and sample 1) and for the rape item (for the combined sample only). The finding that actual force was more prevalent than threats of force in the context of rape echoes a result obtained by Koss and Oros (1982) with an earlier version of the SES. The results of the second set of analyses showed that, on empirical grounds, the SES items can be collapsed with no likelihood of overestimation of the prevalence
P1: FVI/FLK
P2: FVI
Archives of Sexual Behavior [asb]
332
PL111-336
June 17, 2000
11:10
Style file version Nov. 19th, 1999
Alksnis, Desmarais, Senn, and Hunter
rates of forced sex play, rape completed by force, or forced anal or oral sex. However, the reader is cautioned that collapsing threat of force with use of force may artificially inflate the number of women who are seen to have experienced physical violence during attempted rapes. The finding that inclusion of threat increases the prevalence estimates for this particular item is not surprising when one considers the unique nature of the attempted rape item. This question requires the respondent to make an inference about the perpetrator’s intent (i.e., that his goal is sexual intercourse and not other forms of sexual contact). No such inference is required for the other three items, each of which deals with completed sexual acts. Thus the possibility of an unwanted outcome (intercourse) constitutes an implicit threat and the pattern of responses observed may be due to women’s reactions to the element of threat inherent to this item. The issue is further complicated by the fact that the behaviors that are supposed to signal the perpetrator’s intent to engage in sexual intercourse are confounded with physical force—this item instructs the respondent to consider whether the attempt involved the perpetrator trying to “get on top of you” while simultaneously furnishing “holding you down” as an exemplar of force. The finding that including threats increases prevalence estimates for the attempted rape item may have very limited meaning in the real world of women experiencing the psychological effects of abuse. It is important to note that the experience of attempted rape cannot be assumed to be less stressful because a perpetrator threatened to hurt the woman, but refrained from actually using physical force. As pointed out earlier, the legal case for attempted rape can be made in both cases (Koss, 1992). This finding might carry weight in the minds of the critics of sexual coercion research who are already suspicious of self-report measures and feminist-oriented investigations. The finding that the local samples described here reported more sexual violence than did Koss’s nationally representative sample may be explained in a number of ways. One possible explanation for our higher numbers is that our recruitment strategies did not preclude the operation of self-selection factors. However, the focus of the current study was advertised as “dating and sexual behavior,” not “sexual coercion,” thus we have no reason to believe that assaulted women were drawn to our study in inordinate numbers. In addition, awareness of the problem of sexual coercion has increased considerably in the years since Koss’s original prevalence estimates were obtained, thanks in part to the press coverage given to subsequent research investigations (e.g., DeKeseredy and Kelly, 1993; Statistics Canada, 1993). It is possible that women in our sample perceived the societal climate as more conducive to disclosure of sexually coercive experiences, counteracting the tendency toward underreporting described earlier. In conclusion, we hope that misinterpretation of research findings will be reduced if investigations are conducted in a self-reflective fashion. By merging the categories of threat and physical force, Koss may have inadvertently provided some ammunition for critics who claim that feminists inflate sexual violence
P1: FVI/FLK
P2: FVI
Archives of Sexual Behavior [asb]
PL111-336
June 17, 2000
11:10
Physical Violence in Sexual Coercion
Style file version Nov. 19th, 1999
333
prevalence figures via the inclusion of less serious forms of coercion. Our strategy of separating the threat and force components of the SES’s items may better serve women and the social movement to end rape. This strategy would show that the results do not follow the intuitively appealing hierarchical pattern with respect to prevalence. It would also help dispel the notion that descriptive statistics obtained using the SES are overestimates resulting from imprecise wording or overly inclusive categories. Because there is no theoretical or empirical basis for keeping threat and force experiences separate, the women with both experiences (there is already substantial overlap) could then be combined for further analyses. Other fruitful avenues of research could include asking participants if they are describing the same incident more than once so that it can be ascertained whether threats are usually a prelude to actual force and surveying men in order to determine if their perceptions of the frequency with which force is used is comparable to women’s. Although research on sexual violence is frequently subject to intense scrutiny by the popular media, the issues raised here are relevant for a wide range of socially relevant research questions that deal with topics that people find threatening, especially research that suggests that one group in society oppresses another (e.g., studies investigating the roots of racism, poverty). The writings of Roiphe, Gilbert, and others illustrate how minor ambiguities in research design may result in an entire investigation being dismissed. One defense against critics’ claims that social scientists are willing to sacrifice methodologic rigor in favor of bolstering a “feminist agenda” is to ensure that research is meticulously conducted and that findings cannot be rejected on the basis of item construction choices that were made in the interests of data simplification. Another potent means of defense involves engaging in an activity that has been traditionally resisted by academics—using the popular media in order to challenge critics and present a more accurate picture of their own research, an approach referred to as “newsmaking criminology” by Barak (1988; cited in DeKeseredy and MacLeod, 1997). ACKNOWLEDGMENTS We acknowledge the support provided by the Social Sciences and Humanities Research Council of Canada. We also thank Mary P. Koss and four anonymous reviewers for their helpful comments and suggestions on an earlier draft of this article. REFERENCES Cook, S. L. (1995). Acceptance and expectation of sexual aggression in college students. Psychol. Wom. Quart. 19: 181–194. DeKeseredy, W. (1996). The Canadian national survey on woman abuse in university/college dating relationships: Biofeminist panic transmission or critical inquiry? Can. J. Crim. 38: 81–104.
P1: FVI/FLK
P2: FVI
Archives of Sexual Behavior [asb]
334
PL111-336
June 17, 2000
11:10
Style file version Nov. 19th, 1999
Alksnis, Desmarais, Senn, and Hunter
DeKeseredy, W., and Kelly, K. (1993). The incidence and prevalence of woman abuse in Canadian university and college dating relationships. Can. J. Soc. 18: 137–159. DeKeseredy, W., and MacLeod, L. (1997). Woman Abuse: A Sociological Story. Toronto: Harcourt Brace. Fekete, J. (1994). Moral Panic: Biopolitics Rising. Toronto: Robert Davies Publishing. Gilbert, N. (1991, June 27). The campus rape scare. Wall Street Journal, p. A14. Gilbert, N. (1993a). Examining the facts: Advocacy research overstates the incidence of date and acquaintance rape. In Gelles, R. J., and Losesk, D. R. (eds.), Current Controversies on Family Violence, Thousand Oaks, CA: Sage. Gilbert, N. (1993b, June 29). The wrong response to rape. Wall Street Journal, p. A18. Koss, M. P. (1985). The hidden rape victim: Personality, attitudinal, and situational characteristics. Psychol. Wom. Quart. 9: 193–212. Koss, M. P. (1992a). Defending date rape. J. Interper. 7: 122–126. Koss, M. P. (1992b). The underdetection of rape: Methodological choices influence incidence estimates. J. Soc. Issue 48: 61–75. Koss, M. P., and Cook, S. L. (1993). Facing the facts: Date and acquaintance rape are significant problems for women. In Gelles, R. J., and Losesk, D. R. (eds.), Current Controversies on Family Violence, Thousand Oaks, CA: Sage. Koss, M. P., and Gidycz, C. A. (1985). Sexual Experiences Survey: Reliability and validity. J. Consult. Clin. Psychol. 53: 422–423. Koss, M. P., Gidycz, C. A., and Wisniewski, N. (1987). The scope of rape: Incidence and prevalence of sexual aggression and victimization in a national sample of higher education students. J. Consult. Clin. Psychol. 55: 162–170. Koss, M. P., and Oros, C. J. (1982). Sexual Experiences Survey: A research instrument investigating sexual aggression and victimization. J. Consult. Clin. Psychol. 50: 455–457. Miller, B., and Marshall, J. C. (1987). Coercive sex on the university campus. J. Coll. Stud. 28: 38–47. Muehlenhard, C. L., and Linton, M. A. (1987). Date rape and sexual aggression in dating situations: Incidence and risk factor. J. Coun. Psyc. 34: 186–196. Pearson, P. (1997). When She Was Bad. Toronto: Random House of Canada Limited. Porter, J. F., and Critelli, J. W. (1992). Measurement of sexual aggression in college men: A methodological analysis. Arch. Sex Behav. 21: 525–542. Rapaport, K., and Burkhart, B. R. (1984). Personality and attitudinal characteristics of sexually coercive college males. J. Abn. Psych. 93: 216–221. Roiphe, K. (1993). The morning after: Sex, fear and feminism on campus. Boston: Little, Brown. Statistics Canada. (1993). The Violence Against Women Survey. The Daily, November 18, 1993, Catalogue No. 11-001E. Ottawa: Statistics Canada. Yurchesyn, K. A., Keith, A., and Renner, K. E. (1992). Contrasting perspectives on the nature of sexual assault provided by a service for sexual assault victims and by the law courts. Can. J. Beh. S. 24: 71–85.
P1: FOM Archives of Sexual Behavior [asb]
PL111-337
June 17, 2000
11:19
Style file version Nov. 19th, 1999
Archives of Sexual Behavior, Vol. 29, No. 4, 2000
Nurse–Patient Sexual Contact in Psychiatric Hospitals Kurt M. Bachmann, M.D.,1,3 Jeannette Bossi, Ph.D.,2 Franz Moggi, Ph.D.,1 Frances Stirnemann-Lewis, Ph.D.,1 Ruth Sommer, B.A.,1 and Hans D. Brenner, M.D., Ph.D.1
Sexual contacts between nurses and patients in psychiatric hospitals have not been investigated systematically. The aim of the present study was to determine the frequency of nurse–patient sexual relationships and their prominent characteristics on the one hand and the nurses’ attitudes towards these contacts on the other. A questionnaire was mailed to 714 nurses employed at two psychiatric hospitals. Although 94% of the 279 respondents considered sexual contact (defined as “physical contact between a patient and a nurse, in which sexual arousal occurred in the nurse”) to be inappropriate, 17% of the male and 11% of the female responding nurses reported having had such contacts with patients. KEY WORDS: sexual contact; nurse; patient; psychiatric hospitals.
INTRODUCTION In the late 1960s, reports on sexual contact between therapists and their patients began to appear (Shepard, 1971; Van Einde Boas, 1966; Dahlberg, 1970) in the literature. In the meantime, several studies have been published on sexual boundary violations involving physicians in various speciality boards (Kardener, 1973; Hankins et al., 1994), psychologists (Pope, 1986; Wincze et al., 1996), psychiatrists (Gartrell, 1986; Leggett, 1994), social workers (Gechtman et al., 1985; Bernsen et al., 1994; Seto, 1995), and counselors (Thoreson et al., 1995; Hoffmann, 1995; Case et al., 1997). Although some research has been conducted 1 Psychiatric Services of the University of Berne, Bolligenstrasse 111, CH-3000 Berne 60, Switzerland. 2 Alderstrasse 3 To
21, CH-8008 Zurich, Switzerland. whom correspondence should be addressed. 335 C 2000 Plenum Publishing Corporation 0004-0002/00/0800-0335$18.00/0 °
P1: FOM Archives of Sexual Behavior [asb]
336
PL111-337
June 17, 2000
11:19
Style file version Nov. 19th, 1999
Bachmann, Bossi, Moggi, Stirnemann-Lewis, Sommer, and Brenner
on the characteristics of psychiatric nurses who have had sexual contact with patients in outpatient treatment settings and its impact on patients (Schoener et al., 1989; Pope, 1990, 1994), there is a paucity of literature or case reports that refer to the situation of inpatients (Kirstein, 1978). A study on sexual exploitation of patients in the hospital setting was published in 1989 by a group at the Menninger hospital (Averill et al., 1989) that reviewed reported incidents in the institution during the previous decade. A national survey by Berland and Guskin (1994) showed that 36% of 255 psychiatric units reported patient allegations of sexual abuse by a staff member during 1985–1991. The study showed that public hospitals were more likely to have experienced allegations of sexual abuse than were community or private hospitals. As very few data about the frequency of nurse–patient sexual relationships have been published, no data about the consequences of these contacts can be found in the literature. Considering the fact that nurse–patient relationships are substantially different from other professional relationships, it is surprising that sexual contact between nurses and their patients has not been the subject of systematic research. Nurses spend considerably more time with their patients—up to several hours a day—than other treating staff. Body contact is often a necessary part of interpersonal relationships that are characterized by involvement and intimacy. This closeness makes it more difficult for nurses than for practitioners of other disciplines to maintain clear roles and boundaries in the relationships they have to patients. The fact that psychiatric nurses and their patients live with each other in a way that resembles family life to a much greater extent than other treatment settings suggests that many aspects of intrafamilial relationships might be present as well. Therapist–patient sexual relationships and incest have been shown to reveal many similarities, both in their dynamics (Smith, 1989) and in their consequences (Apfel et al., 1985). So nurse–patient sexual contacts could also give some model-like insight into intrafamilial sexual contacts. In other therapeutic professions, these consequences have been widely described and are known as the patient sex syndrome (Seto, 1995; Moggi and Brodbeck, 1997). It can be hypothezised that similar syndromes occur in patients having had sexual contact with their nurses during psychiatric hospitalisations. The hypothesized patient sex syndrome may be even more severe in hospitalized patients than in nonhospitalized patients: On the one hand, hospitalized patients are in general severely ill compared to ambulatory patients; on the other hand, sexual abuse as a risk factor for sexual abuse during therapy and as a reinforcement factor for the patient sex syndrome may be found more often in hospitalized patients than in the ambulatory treated population (Mueser et al., 1998; Read, 1998; Read and Fraser, 1998). The aim of the present study was threefold: first, to determine if psychiatric nurses have sexual contact with their patients, and if so, to examine nurses’
P1: FOM Archives of Sexual Behavior [asb]
PL111-337
June 17, 2000
Nurse–Patient Sexual Contact
11:19
Style file version Nov. 19th, 1999
337
attitudes to this issue; third, to determine the frequency of these contacts and their characteristics. METHOD A 35-item questionnaire was mailed to all male and female nurses employed at two psychiatric hospitals. It contained questions in the following domains: sex; attitude toward sexual contact in nurse–patient relationships; characteristics of sexual contact with their own patients (if any); nurses’ own history of child sexual abuse; number of colleagues known to have had sexual contact with patients; and the need to provide help for nurses who have had sexual contact with patients. Questionnaire 1. Are you 2.
3.
4. 5.
6.
7. 8.
male female Have you been victim of sexual abuse in your own childhood? yes no don’t know If yes, by whom: father mother other person in the family other person not belonging to the family If yes, do you feel psychologically impaired today as a consequence of these events? Are there any therapeutic situations in which sexual contacts between a nurse and her/his patients are opportune according to your opinion? yes no Do you believe that sexual contacts between a nurse and a patient may have negative consequences for the patient? yes, sometimes yes, mostly no From how many nurse-colleagues do you know for sure that they had sexual contacts with their patients? Please give an estimated number: Have you ever had sexual contacts except intercourse with patients of yours? (Sexual contacts defined as physical contacts in which sexual arousal occurred in yourself) yes no
P1: FOM Archives of Sexual Behavior [asb]
338
11. 12.
13. 14.
15.
16.
17.
18.
19.
20. 21.
22. 23. 24. 25.
June 17, 2000
11:19
Style file version Nov. 19th, 1999
Bachmann, Bossi, Moggi, Stirnemann-Lewis, Sommer, and Brenner
9. If yes: 10.
PL111-337
with patients of your sex with patients of the opposite sex If yes: with one patient with more than one patient If yes: during the hospitalization after the hospitalization Did a love relationship exist at the time of the sexual contact to the patient? yes no Can you describe shortly the exact nature of the sexual contact? Do you know which diagnosis the patient had been given? diagnosis I don’t know Did you take the active role during these sexual contacts? yes no Did you have the opportunity to discuss these sexual contacts with other nursecolleagues? yes no How old were you during these sexual contacts? less than 30 years more than 30 years How long had you been working as a nurse at this point of time? less than 5 years more than 5 years Would you act in the same way in a similar situation as you did at that moment? yes no Do you have any remarks to questions 8–19? Did you have sexual intercourse with a patient? yes no If yes: with patients of your sex with patients of the opposite sex If yes: with one patient with more than one patient If yes: during the hospitalization after the hospitalization Did a love relationship exist at the time of the sexual contact to the patient? yes no
P1: FOM Archives of Sexual Behavior [asb]
PL111-337
Nurse–Patient Sexual Contact
June 17, 2000
11:19
Style file version Nov. 19th, 1999
339
26. Do you know which diagnosis the patient had been given? diagnosis I don’t know 27. Did you take the active role during these sexual contacts? yes no 28. Did you have the opportunity to discuss these sexual contacts with other nursecolleagues? yes no 29. How old were you during these sexual contacts? less than 30 years more than 30 years 30. How long have you been working as a nurse at this point in time? less than 5 years more than 5 years 31. Would you act in the same way in a similar situation as you did at that moment? yes no 32. Do you have any remarks to questions 21–31? 33. Would you appreciate a counselling or assistance service for nurses who have had sexual contacts with their patients? yes no 34. In case that you had sexual contacts yourself with your patients: Would you have used such a service? yes no 35. Remarks to the entire questionnaire. Respondents who reported having had sexual contact with patients were asked to answer questions about the specific activities engaged in: the characteristics are shown in Table I. Sexual contact was defined as “physical contact, in which sexual arousal occurred in the nurse.” Most of the questions were forced-choice responses (agree/disagree). Space was provided to describe the sexual contact and for additional comments. Enclosed in this questionnaire was a prestamped return envelope and a cover letter (see Appendix 1) that explained the purpose of the study, guaranteed complete anonymity, specified a hand-in deadline, and described the procedures to be used. This material was mailed to the 714 (447 female and 267 male) nurses employed at two public psychiatric hospitals. The two psychiatric hospitals with a total of 841 beds serve a catchment area of 600,000 inhabitants. The year the
P1: FOM Archives of Sexual Behavior [asb]
PL111-337
340
June 17, 2000
11:19
Style file version Nov. 19th, 1999
Bachmann, Bossi, Moggi, Stirnemann-Lewis, Sommer, and Brenner Table I. Characteristics of the Nurses Affirming Having Sexual Contacts with Patients All nurses Characteristics
F
%
Female nurses F
%
Male nurses F
%
Had sexual encounters with more than one patient
8 25 (n = 32)
4 24 (n = 17)
4 27 (n = 15)
n.s.
Would repeat the sexual contact in similar situations
12 38 (n = 32)
4 25 (n = 16)
8 50 (n = 16)
n.s.
Initiated the contact
10 30 (n = 33)
2 13 (n = 16)
8 47 (n = 17)
X 2 = 4.66 df = 1 p < 0.05
Was in love with the patient
7 21 (n = 34)
4 24 (n = 17)
3 18 (n = 17)
n.s.
Same-gender contact
8 24 (n = 34)
4 24 (n = 17)
4 24 (n = 17)
n.s.
Sexual contact occurred during hospitalization of the patient Was under 30 years old
20 57 (n = 35) 29 88 (n = 33) 25 81 (n = 31) 30 88 (n = 34) 33 94 (n = 35)
10 59 (n = 17) 16 94 (n = 17) 14 88 (n = 16) 16 89 (n = 18) 17 94 (n = 18)
10 56 (n = 18) 13 81 (n = 16) 11 73 (n = 15) 14 88 (n = 16) 16 94 (n = 17)
n.s.
26 77 (n = 34)
15 88 (n = 17)
11 65 (n = 17)
n.s.
6 24 (n = 25)
6 46 (n = 13)
0 0 (n = 12)
X 2 = 7.29 df = 1 p < 0.025
Less than 5 years of practical working experience as a nurse Thinks that sexual contacts are contraindictive in therapy Thinks that nurse–patient sexual contacts can have negative consequences for patients Thinks that special consultation for offenders would be necessary Would have sought consultation after having engaged in sexual contacts with patients
n.s. n.s. n.s. n.s.
n = Total number of responses for each question (varied because some responders did not answer all questions). F = Number of affirmative responses.
present study was conducted, 2986 adult patients covering the whole spectrum of DSM-IV diagnoses were treated. The investigated hospitals represent typical Swiss psychiatric clinics. Most of the patients are accommodated in double-bed rooms. The majority of patients stay longer than 1 month in the clinics. Duties involving intimate contacts are usually performed by a nurse having the same gender as the patient. The responses were key-punched and tabulated (SPSSX software). Chi-square tests were used as a test of significance in the event that the expected cell frequencies were sufficiently high (n > 5). If not, Fisher’s exact test (two-tailed) was performed. When n exceeded 40, a continuity correction was applied.
P1: FOM Archives of Sexual Behavior [asb]
PL111-337
June 17, 2000
11:19
Style file version Nov. 19th, 1999
Nurse–Patient Sexual Contact
341
RESULTS Characteristics of the Respondents Questionnaires were returned by 279 of the 714 nurses (39%). The return rate was not related to the respondent’s gender: 107 male (40%) and 172 females (39%) sent the questionnaire back. Although 279 questionnaires were returned, not all questions were answered by all respondents. Thus the number of responses for specified questions ranged from 229 to 279. As a consequence of very strict anonymity of the respondents which is requested by this study, no further information about the differences between respondents and nonrespondents can be given. Nurses’ Attitudes Ninety-four percent of the male and female respondents considered it inappropriate for nurses to have sexual contact with patients, and 92% thought that sexual intimacies would have negative effects on patients in the long run. There was a significant difference between the sex of the nurses who denied that sexual intimacies would have negative effects on patients: 3 of the 102 male respondents (3%) and 18 of the 162 female nurses (11%) believed that sexual contact with patients would not lead to negative consequences for patients (X 2 = 5.7, df = 1, p < 0.025). Similar results were obtained by Bernsen et al. (1994), who reported that female therapists were more likely than male therapists to report that their attraction to a patient had never been harmful. Although 78% of all the respondents felt that a special counselling program or a self-help group should be available for nurses who have sexual contact with their patients, significantly more of 141 female nurses (84%) thought so than their 88 male counterparts (68%) (X 2 = 8.35, df = 1, p < 0.005). Fifty-two percent of the sample knew of at least one colleague who had had sexual contact with patients. Sexual Contact with Patients In our sample, 18 (17%) male and 18 (11%) female respondents admitted having had sexual contact with patients. Twenty-five (12 female and 13 male nurses) answered the question about the nature of the activities engaged in. The characteristics of the nurses who revealed having had sexual contact with patients are summarized in Table I. Although none of the male nurses had had intercourse with their patients, 4 of the 18 female nurses reported they had. Other sexual activities were divided into two subgroups: activities in subgroup 1 implicated sexual arousal during nursing activities involving the patient’s genitals, hugging, and hand-holding. Subgroup 2 was defined by activities like kissing, touching genitals, petting, and intercourse. Although 8 of the nurses who answered questions
P1: FOM Archives of Sexual Behavior [asb]
342
PL111-337
June 17, 2000
11:19
Style file version Nov. 19th, 1999
Bachmann, Bossi, Moggi, Stirnemann-Lewis, Sommer, and Brenner
about the nature of the activities had engaged in subgroup 1 activities only, 17 reported subgroup 2 activities (characteristics are shown in Table II). Gender was not predictive of the type of activities engaged in. Nurses with History of Being Sexually Abused Nineteen female (12%) and 9 male (9%) respondents revealed having been sexually abused as children. The characteristics of the nurses who had been victims of childhood sexual abuse and its consequences are shown in Table III. Five of the nine men who had been subjected to sexual abuse during their childhood reported having had sexual contact with their patients, compared to 13% (n = 93) of the 93 nurses who had not been sexually abused as children (Fisher test = 10.75, df = 1, p < 0.025). Three of the 19 female nurses who had been sexually exploited as children had sexual contact with patients, compared to 10% (n = 146) of female nurses who had not been victims of childhood sexual abuse (n.s.). More than Table II. Characteristics of the Sexual Contacts All nurses (n = 25)
Female nurses (n = 12)
Male nurses (n = 13)
Characteristics
F
%
F
%
F
%
Sexual arousal during nursing activities involving genitals of the patient Hugging, handholding Kissing Touching genitals Petting Intercourse
4
16
0
—
4
31
4 10 1 2 4
4 40 4 8 16
1 6 0 1 4
8 50 — 8 33
3 4 1 1 0
23 31 8 8 —
n = Total number of responses for each question (varied because some respondents did not answer all questions). F = Number of affirmative responses. Table III. Characteristics of the Childhood Sexual Abuse Experienced All nurses (n = 28)
Female nurses (n = 19)
Male nurses (n = 9)
Characteristics
F
%
F
%
F
%
Offender Father Another immediate family member Somebody outside the family Respondent feels the damage done cannot be redressed
3 9 16 8
11 32 57 29
3 6 10 6
16 32 53 32
3 6 2
33 67 22
n = Total number of responses for each question (varied because some respondents did not answer all questions). F = Number of affirmative responses.
P1: FOM Archives of Sexual Behavior [asb]
PL111-337
June 17, 2000
11:19
Nurse–Patient Sexual Contact
Style file version Nov. 19th, 1999
343
half (5 of 9) of the male nurses who had been subjected to sexual abuse had had sexual contact with patients, compared to 13% (12 of 93) of the nurses who had not been sexually abused as children (Fisher test = 10.75, df = 1, p < 0.0259). Approximately 16% (3 of 19) of the female nurses who had been sexually exploited as children had sexual contact with patients, compared to approximately 10% (14 of 146) of female nurses who had not been victims of childhood sexual abuse (n.s.). DISCUSSION Sexual intimacy between therapists and their patients is well known and has been investigated empirically for several years and in different groups of professions, but not in nurses. Those intimacies may differ in some way between nurses and their patients compared to other health professions. The nurse–patient relationship is less formal and structured than the traditional psychotherapist– patient relationship. Although the therapeutic intensity and the power imbalance are at least similar to those of other professional–patient relationship, hospitalized patients are often more sick than patients in ambulatory settings, making the power imbalance even stronger. The nature of nurse–patient relationship is not as clearly defined as the relationship in other psychotherapeutic settings: nurses must shift this relationship in a complex manner from apparently social in nature (e.g., taking a meal together with their patients), taking physical care, setting boundaries, and making psychotherapeutic interventions. Yet every contact between a nurse and a patient should be considered as being therapeutic in the hospital setting. Nurses stay a longer time together with their patients, usually several hours a day over a period of many days. Within this complex setting, sexual attraction between patients and nurses does occur. It may be difficult for a nurse to recognize this attraction as a transference–countertransference mechanism and to deal with it appropriately. This is on one hand due to the complex setting, and on the other hand due to the training of the nurses in which psychotherapy and its pitfalls are not in the center. Other more regressive or counterregressive mechanisms may play a more important role than in other settings. Little is known about the motivation of the nurses to extend their roles into open sexual activities. These motivations may differ from those of other professions. As an analogy from knowledge gained from other helping professions, the existence of a patient harm–syndrome may be probable in nurse–patient settings. However, this harm-syndrome needs to be investigated and proved. Systematic research has not been conducted on sexual contact in nurse–patient relationships. This may be related to the fact that it is illegal for a nurse to have sexual contact with a patient, and this offence has both severe legal and professional (immediate discharge) consequences. Furthermore, this issue might be even more taboo in nursing than in other therapeutic professions. One reason for this might be that nurse–patient relationships have a great deal in common with intrafamilial relationships, which means that the possibility to get sexually involved with patients is deniable to a similar degree as incest.
P1: FOM Archives of Sexual Behavior [asb]
344
PL111-337
June 17, 2000
11:19
Style file version Nov. 19th, 1999
Bachmann, Bossi, Moggi, Stirnemann-Lewis, Sommer, and Brenner
Several special measures were taken in this study in an effort to overcome the nurses’ deep-set fear of answering questions about sexual contact and to encourage participation. First, utmost anonymity had to be assured. As a consequence, the number of questions on demographic data had to be very limited. The return rate (39%) is comparable to those reported in studies investigating other professions that ranged from 26% (Gartrell et al., 1986), to 45% (Bernsen et al., 1994) to 70% (Holroyd, 1977). Second, a very broad definition of sexual contact was used. A similar definition for sexual contact was used as in the studies of Gartrell et al. (1986) and Herman et al. (1987): “contact which was intended to arouse or satisfy sexual desire in the patient, therapist, or both.” Also, the fact that this is the first study to be conducted on nurses working in psychiatric hospitals justifies using a broad definition of sexual contact in order to be able to compare different disciplines within the field of therapy. Further research must examine various categories of sexual contacts so that they can be compared in a much more accurate manner also to other professions working in therapy. Nurses have sexual contact with patients more frequently than other professional groups. Whereas 16.8% of a sample of male nurses reported having had sexual contact with patients, a somewhat smaller proportion (13.7%) of male psychologists revealed having had such contact with patients (Forer, 1980). A higher percentage of female nurses (11%) reported having had sexual contact with patients than other female professionals (ranging from 0% to 3%; Forer, 1980; Gartrell, 1986). This might be related to the fact that nurses have a different relationship with patients than do other therapists. It might be much more difficult to maintain clearcut boundaries in a relationship that necessarily involves body contact and a certain degree of intimacy. This assumption is supported by the fact that sexual contact sometimes occurs during nursing activities or involves more “harmless” activities like hugging and handholding. However, even if this aspect is taken into consideration, female nurses working at psychiatric hospitals still have sexual contact with their patients relatively often. It may also indicate a shift in attitudes or in behavior of female therapists. In the present study, there was general agreement between the male and the female nurses that sexual contact in nurse–patient relationships is not indicated in therapy. In respect to behavior, more male nurses reported having taken an active part in the sexual contact than did their female colleagues. This finding is congruent with accepted social behavior regarding sexual contact. More than half of the sexual intimacies (57%) reported in this survey took place while the patient was in hospital and were terminated on discharge, 23% of the sexual contacts took place during and after hospitalization, and 20% after discharge. Only 21% of the nurses who had had sexual contact with patients said they had been in love with the patient. This suggests that there are other motives involved, as likewise observed by Pope and Bouhoutsos (Pope et al., 1986). Nurses who disclosed having had sexual contact with patients were relatively young: although 53% of the total nurse population recruited for the study is younger
P1: FOM Archives of Sexual Behavior [asb]
PL111-337
Nurse–Patient Sexual Contact
June 17, 2000
11:19
Style file version Nov. 19th, 1999
345
than 30, 88% of the sample surveyed was in this age group. Young respondents are more likely to admit sexual contact, and might be more open to disclose such contacts. Young nurses can be assumed to be at higher risk for sexual contacts with patients. In accordance to this fact, most nurses who had had sexual contact with patients had less than 5 years of work experience in their profession. Even though sexual abuse in childhood seems to be a further risk factor for sexual contact in professional relationships, only a relatively small number of the respondents in our study reported having been sexually abused as children. This may be accounted for by reliance on a single screen question to elicit a history of sexual abuse. There is a definite connection between the number of questions asked on this topic and the prevalence rates reported (Peters, 1986; Moggi, 1991). Significantly more male nurses who had been victims of abuse became sexually involved with their patients than male nurses who had not been. There was no such difference in the group of female nurses. These findings are similar to observations on sexually abused children—especially boys (Porter, 1986), who have been shown to subsequently inflict harm on other children. Although 77% of the nurses who had sexual contact with patients thought that a special counselling program should be provided for professionals who have this problem, only 46% of the female nurses and none of the male nurses would have turned for professional help themselves. This shows the importance of providing primary prevention of sexual abuse in professional relationships that involves training nurses to deal with this aspect of their relationship to patients. The Menninger study (1989) emphasized that clear hospital structure and orientation of new staff members about inappropriate interactions between staff members and patients, frequent educational staff meetings, supervision, and a tactful investigation when taking patients’ histories about childhood incest will aid prevention. A national survey of social workers (Bernsen et al., 1994) and psychologists (Pope et al., 1986) concluded that the greater the amount of relevant graduate training reported, the fewer the reported sexual involvements would be. The study showed that 51% of staff members reported no education about attraction to clients, 26% reported very little training, 13% reported some education, and only 10% reported that their education was adequate. This was despite the findings that 81% of social workers and 87% of psychiatrists reported experiencing sexual attraction to at least one client. In summary, our survey showed that nurses working in psychiatric hospitals do have sexual contact with patients. The nurses who reported having had sexual contact with their patients had less than 5 years working experience in their profession. Male nurses who were victims of childhood sexual abuse are at particularly high risk of becoming sexually involved with their patients. A special counselling program was considered to be necessary by a majority of the respondents, but only a few of the nurses who had sexual contact with patients would have actually turned for professional help. This underlines the importance of undertaking prevention. First and foremost, education on the
P1: FOM Archives of Sexual Behavior [asb]
346
PL111-337
June 17, 2000
11:19
Style file version Nov. 19th, 1999
Bachmann, Bossi, Moggi, Stirnemann-Lewis, Sommer, and Brenner
issue of sexual contact with patients would have to be improved in nursing training. Second, the basic ethical principles of the international council of nurses (Henderson, 1977) would have to be expanded and refined so as to incorporate the possibility of sexual boundary violations. Further studies must extend and integrate our results into the whole domain of sexual involvement between therapists and patients.
REFERENCES Apfel, R. F., and Simon, B. (1985). Patient–therapist sexual contact. 1. Psychodynamic perspectives on the cause and results. Psychother. Psychosom. 43: 57–62. Averill, S. C., Beale, D., Benfer, B., Collins, D. T., Kennedy, L., Myers, J. A., Pope, D., Rosen, I., and Zoble, E. (1989). Preventing staff–patient sexual relationships. Bull. Menn. Clin. 53: 384– 393. Berland, D. I., and Guskin, K. (1994). Patient allegations of sexual abuse against psychiatric hospital staff. Gen. Hosp. Psych. 16: 335–339. Bernsen, A., Tabachnick, B. G., and Pope, K. S. (1994). National survey of social workers’ sexual attraction to their clients: Results, implications and comparison to psychologists. Eth. Behav. 4(4): 369–388. Case, P. W., McMinn, M. R., and Meek, K. R. (1997). Sexual attraction and religious therapists: Survey findings and implications. Couns. Val. 41(2): 141–154. Dahlberg, C. C. (1970). Sexual contact between patient and therapist. Contemp. Psychoanal. 6: 107– 124. Forer, B. (1980). The therapeutic relationship. Paper presented at the annual meeting of the California State Psychological Association. Pasadena, California. Gartrell, N., Herman, J., Olarte, S., Feldstein, M., and Localio, R. (1986). Psychiatrist–patient contact: Results of a national survey, I: Prevalence. Am. J. Psych. 143(9): 1126–1131. Gechtman, L., and Bouhoutsos, J. C. (1985). Social workers’ attitudes and practices regarding erotic and nonerotic physical contact with their clients. Paper presented at the Annual conference of the California Society for Clinical Social Work, and the National Federation of Societies for Clinical Social Work, Universal City, CA. Hankins, G. C., Vera, M. I., Barnard, G. W., and Herkov, M. J. (1994). Patient–therapist sexual involvement: A review of clinical and research data. Bull. Am. Acad. Psych. Law 22(1): 109–126. Henderson, V. (1977). Basic principles of nursing care. International Council of Nurses. Geneva. Herman, J. L., Gartrell, N., Olarte, S., Feldstein, M., and Localio, R. (1987). Psychiatrist–patient sexual contact: Results of a national survey, II: Psychiatrists’ attitudes. Am. J. Psych. 144(1): 164– 169. Hoffman, R. M. (1995). Sexual dual relationships in counseling: Confronting the issues. Couns. Val. 40(1): 15–23. Holroyd, J. C., and Brodsky, A. N. (1977). Psychologist’s attitudes and practices regarding erotic and nonerotic physical contact with patients. Amer. Psychol. 32(10): 843–849. Kardener, S. H., Fuller, M., and Mensh, I. N. (1973). A survey of physician’s attitudes and practices regarding erotic and nonerotic contact with patients. Am. J. Psych. 130: 1077–1081. Kirstein, L. (1978). Sexual involvement with patients. J. Clin. Psychol. 39: 366–368. Leggett, A. (1994). A survey of Australian psychiatrists’ attitudes and practices regarding physical contact with patients. Aust. NZ J. Psych. 28(3): 488–497. Moggi, F. (1991). Sexuelle Kindesmisshandlung: Definition, Pr¨avalenz und Folgen. Ein Ueberblick. Z. Klin. Psychol. Psychopath. Psychother. 39: 323–335. Moggi, F., and Brodbeck, J. (1997). Risk factors and consequences of sexual abuse in psychotherapy. Z. Klin. Psych. Forsch. Praxis 26(1): 50–57. Mueser, K. T., Goodman, L. B., Trumbetta, S. L., Rosenberg, S. D., Osher, F. C., Vidaver, R., Auciello, P., and Foy, D. W. (1998). Trauma and posttraumatic stress disorder in severe mental illness. J. Consult. Clin. Psych. 66(3): 493–499.
P1: FOM Archives of Sexual Behavior [asb]
PL111-337
Nurse–Patient Sexual Contact
June 17, 2000
11:19
Style file version Nov. 19th, 1999
347
Peters, S. D., Wyatt, G. E., and Finkelhor, D. (1986). Prevalence. In Finkelhor, D. (ed.), A Sourcebook on Child Sexual Abuse, London, Sage, pp. 15–59. Pope, K. S. (1990). Therapist–patient sexual involvement: A review of the research. Clin. Psych. Rev. 10: 477–490. Pope, K. S. (1994). Sexual involvement with therapists: Patient assessment, subsequent therapy. Forensics. Washington DC: American Psychological Association. Pope, K. S., and Bouhoutsos, J. (1986). Sexual intimacy between therapists and patients. New York, Praeger. Pope, K. S., Keith-Spiegel, P., and Tabachnick, B. G. (1986). Sexual attraction to clients. Amer. Psychol. 41(2): 147–158. Porter, E. (1986). Treating the young male victim of sexual assault: Issues and intervention strategies. Sater Society Press, Syracuse, New York. Read, J., and Fraser, A. (1998). Abuse histories of psychiatric inpatients: To ask or not to ask? Psych. Serv. 49(3): 355–359. Read, J. (1998). Child abuse and severity of disturbance among adult psychiatric inpatients. Child Abuse Neg. 22(5): 359–368. Schoener, G. R. (1989). Psychotherapists’ sexual involvement with clients: Intervention and prevention. Minneapolis, Minnesota: Walk-In Counseling Center. Seto, M. C. (1995). Sex with therapy clients: Its prevalence, potential consequences, and implications for psychology training. Can. Psych. 36(1): 70–86. Shepard, M. (1971). The love treatment: Sexual intimacy between patients and psychotherapists. New York, Peter H. Wyden. Smith, S. (1989). The seduction of the female patient. In Gabbard, G. O. (ed.), Sexual Exploitation in Professional Relationship, American Psychiatric Press, Washington, DC. Thoreson, R. W., Shaughnessy, P., and Frazier, P. A. (1995). Sexual contact during and after professional relationships: Practices and attitudes of female counselors. J. Coun. Dev. 74(1): 84–89. Van Einde Boas, C. (1966). Some reflections on sexual relations between physicians and patients. J. Sex Res. 2: 215–218. Wincze, J. P., Richards, J., Parsons, J., and Bailey, S. (1996). A comparative survey of therapist sexual misconduct between an American state and an Australian state. Prof. Psych. Res. Prac. 27(3): 289–294.
P1: FNN/FLK
P2: FNN
Archives of Sexual Behavior [asb]
PL111-338
June 24, 2000
10:30
Style file version Nov. 19th, 1999
Archives of Sexual Behavior, Vol. 29, No. 4, 2000
Birth Order in a Contemporary Sample of Gay Men David W. Purcell, J.D., Ph.D.,1 Ray Blanchard, Ph.D.,2 and Kenneth J. Zucker, Ph.D.3,4
The birth order of a contemporary North American sample of 97 gay men was quantified using Slater’s Index. For the 84 probands with at least one sibling, the results showed a late mean birth order compared with the expected value of .50. Additional birth order indices derived from Slater’s Index suggested that the mean later birth order was accounted for more strongly by the proband’s number of older brothers than by his number of older sisters. The present findings constitute a replication of a series of recent studies and add to the growing body of evidence that birth order is a reliable correlate of sexual orientation in males. KEY WORDS: birth order; sex ratio; sexual orientation; homosexuality; males.
INTRODUCTION Identification of correlates of sexual orientation provides possible clues regarding its origins and genesis. Perhaps the most consistent and reliable correlate of sexual orientation pertains to patterns of childhood sex-typed behavior. On average, both gay men and lesbians recall more cross-gender behavior in childhood than do heterosexual men and women (e.g., Bailey and Zucker, 1995; Bell et al., A version of this article was presented at the meeting of the International Academy of Sex Research, Stony Brook, New York, June 1999. 1 Department of Psychology, Emory University, Atlanta, Georgia. Now at Centers for Disease Control and Prevention, Atlanta, Georgia. 2 Clinical Sexology Program, Centre for Addiction and Mental Health—Clarke Division; Toronto, Ontario; Department of Psychiatry, Faculty of Medicine, University of Toronto. 3 Child and Adolescent Gender Identity Clinic, Child Psychiatry Program, Centre for Addiction and Mental Health—Clarke Division; Toronto, Ontario; Department of Psychiatry, Faculty of Medicine, University of Toronto. 4 To whom correspondence should be addressed at Child and Adolescent Gender Identity Clinic, Child Psychiatry Program, Centre for Addiction and Mental Health—Clarke Division; 250 College St., Toronto, Ontario M5T 1R8 Canada; e-mail: Ken
[email protected]. 349 C 2000 Plenum Publishing Corporation 0004-0002/00/0800-0349$18.00/0 °
P1: FNN/FLK
P2: FNN
Archives of Sexual Behavior [asb]
PL111-338
June 24, 2000
10:30
350
Style file version Nov. 19th, 1999
Purcell, Blanchard, and Zucker
1981) and one controlled, prospective study of very feminine boys provides convergent evidence for this association (Green, 1987). Since the early 1990s, a second variable—birth order—has emerged as a reliable and consistent correlate of sexual orientation in men, but not in women (for review, see Blanchard, 1997). Based on several samples of gay men studied by the second and third authors, gay men were found to be born later in their sibships than heterosexual men (Blanchard and Bogaert, 1996a,b; Blanchard and Zucker, 1994; Blanchard et al., 1998; Zucker and Blanchard, 1994), thus replicating two earlier studies by Slater (1962) and Hare and Moran (1979). Following these reports, there have been three independent replications by other investigators (Bailey et al., 1999; Bogaert, 1998; Fedoroff et al., 1999). Moreover, four other studies have reported a similar finding of a later birth order in transsexual men with a homosexual sexual orientation (Blanchard and Sheridan, 1992; Blanchard et al., 1996; Green, 2000; Tsoi et al., 1977). In some samples that were large enough, it was discerned that the birth-order effect is largely in relation to one’s brothers, and not sisters, which Blanchard (1997) characterized as a fraternal birth order effect (for review, see Jones and Blanchard, 1998). Given the erratic history of replication research on topics pertaining to sexual orientation and its possible precursors (see Byne and Parsons, 1993), the present study examined birth order in a fresh contemporary sample of gay men, originally recruited by Purcell (1995) and evaluated with regard to a variety of adjustmentrelated topics. Information on the subjects’ birth orders had been obtained in the original study, although it had not previously been analyzed. METHOD Subjects The participants were 97 gay men (M age, 33.5 yrs; SD = 7.2; range, 19–51) whose mean year of birth was 1959.8 (SD = 7.2; range, 1942–1974). They were recruited in several ways: after receiving a letter sent to the mailing list of the gay, lesbian, and bisexual organization at a private university in the southeastern United States (42%); newspaper advertisements in two general circulation newspapers (21%); an ad in a weekly newspaper oriented to a gay readership (5%); a radio story based on the newspaper ad (4%); and from friendship networks of prior participants (28%). Regarding ethnicity, 92% were Caucasian and the remainder were of various ethnic minorities. Overall, the participants were well educated: 21% had attended some college (or were in college) and 76% had completed at least an undergraduate college degree. Sexual orientation of the participants was self-rated in two ways: (1) on the 7-point Kinsey scale with regard to a “lifetime” appraisal, in which 0 = exclusively heterosexual and 6 = exclusively homosexual (Kinsey et al., 1948) and (2) on a
P1: FNN/FLK
P2: FNN
Archives of Sexual Behavior [asb]
PL111-338
June 24, 2000
Birth Order in Gay Men
10:30
Style file version Nov. 19th, 1999
351
continuous scale asking about percentage of current same-sex and opposite-sex fantasies. The participants were also asked to classify their current sexual identity. On the Kinsey scale, self-rating of sexual orientation was as follows: 38% were a “6,” 55% were a “5,” and 7% were a “4.” Regarding current sexual fantasies, 98.4% were directed toward other men and 1.6% were directed toward women. Self-designation of sexual identity was as follows: 96% were gay or homosexual; 3% were “bisexual, but mostly gay”; and 1% was “bisexual, equally gay and straight.” Measures The participants recorded on a form their numbers of older brothers, older sisters, younger brothers, and younger sisters. Maternal half-siblings were included, but paternal half-siblings were not. Several biodemographic variables were calculated from the foregoing data. The variable, sibling sex ratio, is the ratio of brothers to sisters reported collectively by a given group of probands. In white populations, the ratio of male live births to female live births is close to 106:100 (Chahnazarian, 1988; James, 1987). The ratio of brothers to sisters reported by any group of persons drawn at random from the general population should therefore approach 106 (brothers per 100 sisters). It should be noted that the calculation of the sibling sex ratio does not include the proband himself. In the computation of inferential statistics, the sibling sex ratio is more conveniently expressed as the proportion of males rather than the ratio of males to females—that is, .5146 (106/206). Birth order was quantified using the birth-order index introduced by Slater (1958, 1962). Slater’s Index equals the proband’s number of older siblings divided by his total number of siblings. This index cannot be calculated for only children; for all other individuals, regardless of their number of siblings, it expresses birth order as a quantity between 0 and 1, where 0 corresponds to firstborn and 1 corresponds to last born. In a hypothetical stable population, the expected value of Slater’s Index for samples drawn at random would be .50, and one can determine whether some group’s birth order is significantly early or late by comparing its mean on Slater’s Index with this theoretic value. Birth orders relative to brothers and relative to sisters were calculated separately using indices derived from Slater’s Index (Jones and Blanchard, 1998). The fraternal birth order index equalled the proband’s number of older brothers divided by his total number of brothers. Similarly, the sororal birth-order index was calculated as the proband’s number of older sisters divided by his total number of sisters. It is important to be clear on the distinction between sibling sex ratio and fraternal birth order. Sibling sex ratio concerns a proband’s number of brothers compared with his number of sisters; fraternal birth order concerns his number of older brothers compared with his number of younger brothers. The two parameters
P1: FNN/FLK
P2: FNN
Archives of Sexual Behavior [asb]
PL111-338
June 24, 2000
10:30
352
Style file version Nov. 19th, 1999
Purcell, Blanchard, and Zucker
can vary independently. A man who has a low fraternal birth order (e.g., no older brothers and two younger brothers) might have either a high sibling sex ratio (no sisters) or a low sibling sex ratio (five sisters). RESULTS The probands had 101 brothers and 79 sisters. This equals a sibling sex ratio of 128:100, or a proportion of .5611. The observed proportion of brothers was compared with the expected proportion (.5146) using the z approximation to the binomial test. The result was not significant ( p = .12, one-tailed), which is not surprising because of the large sample requirements to achieve adequate statistical power for this measure (Moore and Gledhill, 1988; Suarez and Przybeck, 1980). On Slater’s Index, the mean birth order of the 84 probands was .63 (SD = .41). This differed significantly from the expected value of .50, t(83) = 2.81, p = .003, one-tailed, thus indicating that these men had a higher (later) than expected birth order. The next phase of the analysis was examining the fraternal and sororal birth order indices in order to determine whether the high birth order of the subjects was driven primarily by older brothers. For reasons previously explained by Jones and Blanchard (1998), we restricted this analysis to those 37 subjects who had at least one brother and at least one sister. The mean sororal index was .53 (SD = .50), whereas the mean fraternal index was .61 (SD = .47), showing that those subjects who had siblings of both sexes were born later among their brothers than they were among their sisters. The fraternal index did not differ from the expected value of .50, t(36) = 1.47, p = .08, one-tailed. The sororal index also did not differ from the expected vale of .50, t(36) = 0.33, p = .37, one-tailed. Because of prior research indicating that birth order effects are seen more clearly in larger families and that this can be achieved by weighting cases according to family size (Blanchard et al., 1995; Zucker et al., 1997), we weighted each case (not score) by the quantity, s × (37/94), where s was the total number of siblings for that case and 37/94 (the number of probands divided by the number of siblings) was a constant required to bring the degrees of freedom back to their original value. In the weighted sample, the mean fraternal index was .63 (SD = .45), and the mean sororal index was .55 (SD = .49). The fraternal index was significantly greater than .50, t(36) = 1.76, p = .04, one-tailed; but the mean sororal index was not, t(36) = 0.66, p = .26, one-tailed. DISCUSSION The results showed that gay men tend to be born late in their sibships and also showed that this phenomenon primarily reflects a higher than expected number of
P1: FNN/FLK
P2: FNN
Archives of Sexual Behavior [asb]
PL111-338
Birth Order in Gay Men
June 24, 2000
10:30
Style file version Nov. 19th, 1999
353
older brothers. In the behavioral sciences in which the field is littered with new and exciting “findings” that subsequently proved chimerical, there is almost no such thing as too much replication. The present findings, therefore, represent a further addition to the cumulative evidence regarding the fraternal birth order of gay men. This study did not include a heterosexual comparison group, and this aspect of its methodology necessitates some further discussion. If one is drawing probands at random from a population that is neither increasing nor decreasing in total size or in average family size, then one would have an equal probability of selecting subjects who are early-born, middle-born, or late-born within their individual sibships. The average proband, in other words, would have an equal number of older and younger siblings. In this circumstance, as previously indicated, the expected value of Slater’s Index for samples drawn at random is .50, and one can determine whether some group’s birth order is significantly early or late by comparing its mean on Slater’s Index with this theoretic value. Several theorists and researchers have shown, however, that the probabilities of selecting early-, middle-, and late-born probands are altered in complex ways when total population size or average family size or both are changing during the years when a sample of probands is being born (Berglin, 1982; Birtchnell, 1971; Cobb, 1914; Hare and Price, 1969, 1974; Jagers, 1982; Price and Hare, 1969). The extent and even direction in which the expected birth order of a given group of probands will be shifted by such demographic factors is, in practice, incalculable, although it has been argued that this can, in principle, be done (Berglin, 1982, 1985). For this reason, the expected value of Slater’s Index may depart from .50, and thus the p values from one-sample tests comparing observed data with this theoretical mean may be inaccurate. There are two reasons why the present findings may still be regarded as valid, despite the foregoing problem. The first is that the available empirical evidence indicates that one-sample tests comparing the observed mean score of homosexual men with a theoretical mean of .50 would tend to be, if anything, too conservative. Blanchard (1997) presented the mean Slater’s Index for 12 heterosexual control samples totaling 4866 subjects. The mean Slater’s Index for one sample was .50; for the other 11 samples it was less than .50, and in most cases substantially less. The data therefore suggest that an assumed mean of .50 may be too high for one-sample tests involving gay men recruited during the 20th century. The second reason is even more important. Perturbations in the expected birth order produced by secular changes in the number of total births, family size, and so on, would affect fraternal and sororal birth order equally. Such demographic factors cannot explain a differential shift in fraternal birth order, as was found in the present study. In summary, these considerations, especially the latter, argue that the present findings represent a genuine confirmation of conclusions derived from studies with heterosexual control groups.
P1: FNN/FLK
P2: FNN
Archives of Sexual Behavior [asb]
354
PL111-338
June 24, 2000
10:30
Style file version Nov. 19th, 1999
Purcell, Blanchard, and Zucker
In light of the apparent reliability of the fraternal birth order effect, some serious explanatory work is in order. The most highly articulated biologic theory of this phenomenon was advanced by Blanchard and Bogaert (1996a). It was conjectured that the high fraternal birth order of gay men may reflect a maternal immune reaction, which is provoked only by male fetuses, and which becomes stronger after each pregnancy with a male fetus. This hypothesis was based partly on the argument that a woman’s immune system would appear the biologic system most capable of “remembering” the number of male (but not female) fetuses that she has previously carried and of progressively altering its response to the next fetus according to the current tally of preceding males. It should be stressed that Blanchard and Bogaert (1996a) did not hypothesize that maternal immune reactions are the only, or the most important, cause of homosexuality in men. Blanchard and Bogaert (1996a) further theorized that the relevant fetal antigen might be one of the male-specific, Y-linked, minor histocompatibility antigens, often referred to collectively as the H-Y antigen. Various lines of indirect evidence supporting the hypothesis that maternal antibodies to H-Y might influence sexual orientation have been summarized by Blanchard and Klassen (1997). The most popular psychosocial explanation of the fraternal birth-order effect is the hypothesis that sexual interaction with older males increases a boy’s probability of developing a homosexual orientation, and that a boy’s chances of engaging in such interactions increase in proportion to his number of older brothers (Jones and Blanchard, 1998). Although this hypothesis may seem intuitively plausible—at least to some people—there are little empirical data to recommend it. In the first place, correlations between same-sex sexual experiences in childhood and homosexuality in adulthood represent very weak evidence. Such correlations might mean that sexual experiences with older boys can cause a younger boy to develop a permanent homosexual orientation, but they can just as easily mean that a prehomosexual boy (whose orientation was already determined in utero) is simply more interested in, or less averse to, sexual interaction with other males. A similar point was made by Gebhard et al. (1965, pp. 329, 457). In the second place, a high birth order has been demonstrated in a large sample of probably prehomosexual boys with a mean age of only 8.46 years (Blanchard et al., 1995). It seems unlikely that a substantial proportion of boys in a sample this young had already been conditioned to homosexuality by sex play with older brothers or anyone else. In the third place, there exist survey data that argue directly against this explanation. Wellings et al. (1994, pp. 204–206) found that men who had attended all-male boarding schools were more likely than men who had not attended such schools to report some homosexual experience, but there was no difference between these groups in the amount of homosexual experience in later life. This suggests that homosexual sex-play in childhood is not an important determinant of sexual orientation in adulthood.
P1: FNN/FLK
P2: FNN
Archives of Sexual Behavior [asb]
PL111-338
June 24, 2000
10:30
Birth Order in Gay Men
Style file version Nov. 19th, 1999
355
In conclusion, there is no obviously correct or even “best-bet” explanation of the fraternal birth-order phenomenon. It is likely that finding the correct explanation is going to be a difficult and lengthy endeavor. For example, decisive studies that might locate the fraternal birth-order mechanism in the prenatal or the postnatal environment (e.g., studies of homosexual and heterosexual adoptees who know their number of biologic as well as adoptive older brothers) are likely to prove difficult for practical reasons (e.g., scarcity of appropriate subjects). The present data, by further reinforcing the reliability of the birth-order phenomenon, show that no matter how difficult this theoretic problem may be, it must be solved in a comprehensive account of sexual orientation. REFERENCES Bailey, J. M., Pillard, R. C., Dawood, K., Miller, M. B., Farrer, L. A., Trivedi, S., and Murphy, R. L. (1999). A family history study of male sexual orientation using three independent samples. Behav. Genet. 29: 79–86. Bailey, J. M., and Zucker, K. J. (1995). Childhood sex-typed behavior and sexual orientation: A conceptual analysis and quantitative review. Dev. Psychol. 31: 43–55. Bell, A. P., Weinberg, M. S., and Hammersmith, S. K. (1981). Sexual Preference: Its Development in Men and Women, Indiana University Press, Bloomington. Berglin, C.-G. (1982). Birth order as a quantitative expression of date of birth. J. Epidemiol. Comm. Health 36: 298–302. Berglin, C.-G. (1985). Male antigenicity and parity. Behav. Brain Sci. 8: 442–443. Birtchnell, J. (1971). Birth rank and mental illness. Nature 234: 485–487. Blanchard, R. (1997). Birth order and sibling sex ratio in homosexual versus heterosexual males and females. Ann. Rev. Sex Res. 8: 27–67. Blanchard, R., and Bogaert, A. F. (1996a). Homosexuality in men and number of older brothers. Am. J. Psychiatry 153: 27–31. Blanchard, R., and Bogaert, A. F. (1996b). Biodemographic comparisons of homosexual and heterosexual men in the Kinsey interview data. Arch. Sex. Behav. 25: 551–579. Blanchard, R., and Klassen, P. (1997). H-Y antigen and homosexuality in men. J. Theor. Biol. 185: 373–378. Blanchard, R., and Sheridan, P. M. (1992). Sibship size, sibling sex ratio, birth order, and parental age in homosexual and nonhomosexual gender dysphorics. J. Nerv. Ment. Dis. 180: 40–47. Blanchard, R., and Zucker, K. J. (1994). Reanalysis of Bell, Weinberg, and Hammersmith’s data on birth order, sibling sex ratio, and parental age in homosexual men. Am. J. Psychiatry 151: 1375–1376. Blanchard, R., Zucker, K. J., Bradley, S. J., and Hume, C. S. (1995). Birth order and sibling sex ratio in homosexual male adolescents and probably prehomosexual feminine boys. Dev. Psychol. 31: 22–30. Blanchard, R., Zucker, K. J., Cohen-Kettenis, P. T., Gooren, L. J. G., and Bailey, J. M. (1996). Birth order and sibling sex ratio in two samples of Dutch gender-dysphoric homosexual males. Arch. Sex. Behav. 25: 495–514. Blanchard, R., Zucker, K. J., Siegelman, M., Dickey, R., and Klassen, P. (1998). The relation of birth order to sexual orientation in men and women. J. Biosoc. Sci. 30: 511–519. Bogaert, A. F. (1998). Birth order and sibling sex ratio in homosexual and heterosexual non-white men. Arch. Sex. Behav. 27: 467–473. Byne, W., and Parsons, B. (1993). Human sexual orientation: The biologic theories reappraised. Arch. Gen. Psychiatry 50: 228–239. Chahnazarian, A. (1988). Determinants of the sex ratio at birth: Review of recent literature. Soc. Biol. 35: 214–235. Cobb, J. A. (1914). The alleged inferiority of the first-born. Eug. Rev. 5: 357–359.
P1: FNN/FLK
P2: FNN
Archives of Sexual Behavior [asb]
356
PL111-338
June 24, 2000
10:30
Style file version Nov. 19th, 1999
Purcell, Blanchard, and Zucker
Fedoroff, J. P., Jacques, T., Kazazic, S., and Peever, C. (1999). An Internet survey of sadomasochism: birth order and gender effects. Poster presented at the meeting of the International Academy of Sex Research, Stony Brook, NY. Gebhard, P. H., Gagnon, J. H., Pomeroy, W. B., and Christenson, C. V. (1965). Sex Offenders: An Analysis of Types, Harper & Row, New York. Green, R. (1987). The “Sissy Boy Syndrome” and the Development of Homosexuality, Yale University Press, New Haven, CT. Green, R. (2000). Birth order and ratio of brothers to sisters in transsexuals. Psychol. Med. 30 (in press). Hare, E. H., and Moran, P. A. P. (1979). Parental age and birth order in homosexual patients: A replication of Slater’s study. Br. J. Psychiatry 134: 178–182. Hare, E. H., and Price, J. S. (1969). Birth order and family size: Bias caused by changes in birth rate. Br. J. Psychiatry 115: 647–657. Hare, E. H., and Price, J. S. (1974). Birth order and birth rate bias: Findings in a representative sample of the adult population of Great Britain. J. Biosoc. Sci. 6: 139–150. Jagers, P. (1982). How probable is it to be first born? and other branching-process applications to kinship problems. Mathemat. Biosci. 59: 1–15. James, W. H. (1987). The human sex ratio. Part 1: A review of the literature. Hum. Biol. 59: 721–752. Jones, M. B., and Blanchard, R. (1998). Birth order and male homosexuality: Extension of Slater’s Index. Hum. Biol. 70: 775–787. Kinsey, A. C., Pomeroy, W. B., and Martin, C. E. (1948). Sexual Behavior in the Human Male, W.B. Saunders, Philadelphia. Moore, D. H., and Gledhill, B. L. (1988). How large should my study be so that I can detect an altered sex ratio? Fertil. Steril. 50: 21–25. Price, J. S., and Hare, E. H. (1969). Birth order studies: Some sources of bias. Br. J. Psychiatry 115: 633–646. Purcell, D. W. (1995). The effects of parent–child relationships on the relation between gender nonconforming behavior in childhood and psychological adjustment in adulthood. Unpublished doctoral dissertation, Emory University, Atlanta, GA. Slater, E. (1958). The sibs and children of homosexuals. In Smith, D. R., and Davidson, W. M. (eds.), Symposium on Nuclear Sex, Heinemann Medical Books, London, pp. 79–83. Slater, E. (1962). Birth order and maternal age of homosexuals. Lancet i: 69–71. Suarez, B. K., and Przybeck, T. R. (1980). Sibling sex ratio and male homosexuality. Arch. Sex. Behav. 9: 1–12. Tsoi, W. F., Kok, L. P., and Long, F. Y. (1977). Male transsexualism in Singapore: A description of 56 cases. Br. J. Psychiatry 131: 405–409. Wellings, K., Field, J., Johnson, A., and Wadsworth, J. (1994). Sexual Behaviour in Britain: The National Survey of Sexual Attitudes and Lifestyles, Penguin Books, London. Zucker, K. J., and Blanchard, R. (1994). Re-analysis of Bieber et al.’s 1962 data on sibling sex ratio and birth order in male homosexuals. J. Nerv. Ment. Dis. 182: 528–530. Zucker, K. J., Green, R., Coates, S., Zuger, B., Cohen-Kettenis, P. T., Zecca, G. M., Lertora, V., Money, J., Hahn-Burke, S., Bradley, S. J., and Blanchard, R. (1997). Sibling sex ratio of boys with gender identity disorder. J. Child Psychol. Psychiatry 38: 543–551.
P1: FMN/FTT/FOP
P2: FQJ
Archives of Sexual Behavior [asb]
PL111-339
June 24, 2000
10:34
Style file version Nov. 19th, 1999
Archives of Sexual Behavior, Vol. 29, No. 4, 2000
Heterosexual Anal Intercourse: An Understudied, High-Risk Sexual Behavior Janice I. Baldwin, Ph.D.,1 and John D. Baldwin, Ph.D.1
Differences between heterosexuals who have or have not engaged in anal intercourse were analyzed. Though anal intercourse is widely recognized as an activity that greatly increases the risks for HIV transmission, it has received little attention in heterosexual populations. A questionnaire was mailed to a random sample of university students, a population in which many people engage in vaginal intercourse with several partners each year. The three largest minorities were randomly oversampled in order that all four major ethnic/racial groups could be statistically evaluated for possible differences. Almost 23% of nonvirgin students had engaged in anal intercourse. Regression analysis indicated that people who had participated in anal intercourse were more likely than people without anal experience to have been younger at first vaginal intercourse, to be older when the data were collected, to have engaged in vaginal intercourse in the last three months before data collection, to be more erotophilic, to use less effective contraceptive methods, and to have used no condom at last coitus. Overall, people who engage in anal intercourse take more sexual risks when engaging in vaginal intercourse than do people without anal experience. No major ethnic/racial differences were detected. Sexologists have not explored anal sex in much detail, hence we have been weak in educating those 20 to 25% of young adults who are not reluctant or (inhibited about) exploring anal intercourse. As young adults use condoms less for anal than vaginal intercourse, they have not learned enough about the risk of anal sex. KEY WORDS: college students; sexual behavior; anal intercourse; risky sexual behavior; sensation seeking.
1 Department of Sociology, University of California—Santa Barbara, Santa Barbara, California 93106;
e-mail:
[email protected]. 357 C 2000 Plenum Publishing Corporation 0004-0002/00/0800-0357$18.00/0 °
P1: FMN/FTT/FOP
P2: FQJ
Archives of Sexual Behavior [asb]
PL111-339
June 24, 2000
10:34
358
Style file version Nov. 19th, 1999
Baldwin and Baldwin
INTRODUCTION The human immunodeficiency virus (HIV) is more easily contracted by anal intercourse than vaginal or oral intercourse (Lazzarin et al., 1991; Mayer and Anderson, 1995; Silverman and Gross, 1997). Although anal intercourse among gay men and bisexuals has received a great deal of attention, considerably less research has focused on anal intercourse among heterosexuals, even though the prevalence of heterosexual anal activities have been well documented (Hunt, 1974; Laumann et al., 1994). Using survey data, Voeller (1991) estimated that 25% of American women and their male partners engage in anal intercourse occasionally, and 10% do so on a somewhat regular basis. This paper examines numerous aspects of anal intercourse from a random sample of university students. This is an important population to study because many of these young adults are sexually active with multiple partners, and 25% of people who contract HIV do so before 20 years of age. In their college study, Johnson et al. (1994) found that the students who were infected with HIV were almost three times more likely to have engaged in anal intercourse than those students who were not infected. Most researchers and sex educators do not pay much attention to the topic of anal intercourse. Polite conversation and Victorian values lead most people to leave this sexual practice among the “unspoken” aspects of social life. Many studies of college students have simply ignored anal intercourse (Belcastro, 1985; Earle and Perricone, 1986; Huang and Uba, 1992; Keller et al., 1982; Murstein and Mercy, 1994; Prince and Bernard, 1998; Robinson et al., 1991; Roche et al., 1993; Turner et al., 1993). Several researchers have commented briefly on the incidence of anal intercourse among college students (Cochran et al., 1991; DeBuono et al., 1990; Fleuridas et al., 1997; Gilbert and Alexander, 1998; Hsu et al., 1994; Hutchison, 1994, 1996; Kotloff et al., 1991; MacDonald et al., 1990; Reinisch et al., 1992; Shapiro et al., 1999; Taylor et al., 1997; Weinberg et al., 1998), but few have analyzed it in depth. Even researchers who include questions about anal intercourse for heterosexuals do not specifically report on activities such as condom use during anal intercourse. Because sexologists have not studied heterosexual anal sex in much detail or collected much data on it, sex educators are ill-prepared for teaching adolescents and adults who are interested in trying anal intercourse. Many people who engage in anal sex use condoms less for anal than vaginal intercourse, suggesting that they have not learned enough about the risks associated with anal intercourse. The goals of this study are not only to examine the prevalence of anal intercourse among college students, but also to analyze various behaviors related to anal activities, including the degree to which individuals take precautions about anal and vaginal intercourse.
P1: FMN/FTT/FOP
P2: FQJ
Archives of Sexual Behavior [asb]
PL111-339
June 24, 2000
10:34
Style file version Nov. 19th, 1999
Heterosexual Anal Intercourse
359
MODEL AND HYPOTHESES A literature review suggested several hypotheses relevant to anal intercourse. The first derives from the research on erotophobia and erotophilia (Fisher, 1988; Fisher et al., 1983), which shows that sex-phobic people tend to avoid more sexrelated activities than people without erotophobia. Given erotophobes’ general discomfort with sex, we hypothesize that they are less likely than erotophiles to engage in anal intercourse. The second hypothesis relates to religion. People with serious religious commitments might be expected to abstain from anal intercourse due to scriptural proscriptions. After all, they are more restrained in their vaginal activities than are less religious people (Mahoney, 1980; Thornton and Camburn, 1989). Thus, we hypothesize that people who attend services frequently will be less likely to engage in anal intercourse than are people who attend services infrequently. Hypothesis three centers on ethnicity and race. Even though studies on college students give us little reason to predict any differences in anal intercourse among ethnic or racial groups (Belcastro, 1985; Cochran et al., 1991; Huang and Uba, 1992), there have been too few studies on ethnicity, race, and anal intercourse to leave this issue unexamined. Thus, we predict a relationship may exist, even though the research to date does not suggest one. Hypothesis four is that women are more likely than men to engage in anal intercourse. Research on college students suggests that women are slightly more likely to report engaging in anal intercourse than men, although this difference is not statistically significant (Cochran et al., 1991; Hsu et al., 1994; Kotloff et al., 1991; MacDonald et al., 1990; Reinisch et al., 1992). Following a logic similar to that given for hypothesis three, we explore hypothesis four rather than assuming that prior studies have proved no gender differences in anal intercourse. The fifth hypothesis is that students coming from low-income families are more likely to engage in anal intercourse than their peers from high-income families. MacDonald et al. (1990) reported that students from lower socioeconomic classes were more likely to engage in anal intercourse than students from higher socioeconomic classes. Sixth, we hypothesize that the present age of the respondents is positively related to engaging in anal intercourse: Older people have simply had more years and opportunities than younger people have had to explore a variety of sexual behaviors, including anal intercourse. Seventh is a series of closely related hypotheses derived from the literature on risk taking and sensation seeking (Aggleton et al., 1994; Seidman et al., 1994; Small and Luster, 1994; Valois et al., 1997; Zuckerman, 1974, 1976, 1979, 1994; Zuckerman et al., 1976). Some of these hypotheses when taken alone are of limited power, but taken together as a whole, they open an important topic of inquiry.
P1: FMN/FTT/FOP
P2: FQJ
Archives of Sexual Behavior [asb]
360
PL111-339
June 24, 2000
10:34
Style file version Nov. 19th, 1999
Baldwin and Baldwin
Zuckerman (1979) presents data showing that people have significant individual differences in the need for novel stimulation and high levels of physiological arousal. This suggests that there may be a generalized personality trait for seeking or avoiding exciting and dangerous activities: Zuckerman (1994) describes sensation seeking and risk taking as “a trait defined by the seeking of varied, novel, complex, and intense sensations and experiences, and the willingness to take physical, social, legal and financial risks for the sake of such experience” (p. 27). Zuckerman suggests that risk takers would be more inclined than others to engage in unsafe sexual activities. Anal intercourse would qualify as an unsafe sexual activity. Hovell et al. (1994) provide independent documentation that youths who engage in risky nonsexual behaviors also engage in unsafe sexual behavior. An early lead for the development of the risk-taking hypotheses was the finding that seatbelt use, a nonsexual behavior, is correlated with sexual and other risk-taking behaviors (Cl´ement and Jonah, 1984; Harvey and Spigner, 1995; Hersch and Viscusi, 1990). Thus, one subset of hypothesis seven is that people who use seatbelts the least are the most likely to engage in anal intercourse. Another nonsexual behavior that has been found to be associated with risktaking behavior is drug use: The literature shows that people who like sensory stimulation and risk taking tend to explore numerous drugs (Zuckerman, 1974, 1994). This and research by Graves and Leigh (1995) and Rotheram-Borus et al. (1994) suggest another subset of hypothesis seven: Those people who have used alcohol, marijuana, cocaine, and other recreational drugs will be more likely to engage in anal intercourse than those who have not used such drugs. People who seek high levels of sensory stimulation often throw caution to the winds, taking risks that others would not. MacDonald et al. (1990) reported that students who engaged in anal intercourse had more vaginal partners than those who had not. Zuckerman et al. (1972, 1976) found that high sensation seekers had more sexual partners and engaged in more sexual activities than low sensation seekers. It is also reasonable to predict that high sensation seekers will be more likely to begin vaginal intercourse at an earlier age. Therefore, we hypothesize that people who engage in anal intercourse begin vaginal intercourse at an earlier age, have more vaginal partners in their lives, and are “currently” sexually active (having had intercourse in the previous 3 months). In a related vein, we predict that people who are risk takers are not only more likely to engage in anal intercourse than others, but also are less likely to show caution during vaginal intercourse—using birth control and condoms less frequently than people who limit their risks. The literatures on sexual behavior and risk taking suggest the hypothesis that people who engage in anal intercourse would not be more worried than their peers about contracting HIV and would not see themselves to be at higher risk (Aggleton et al., 1994; Overby and Kegeles, 1994). If anything, risk-takers might be expected to be less fearful or worried than people who were not used to taking risks. The last subset of hypothesis seven is that knowledge about modes of HIV transmission—including anal intercourse—is as good among people who have
P1: FMN/FTT/FOP
P2: FQJ
Archives of Sexual Behavior [asb]
PL111-339
June 24, 2000
10:34
Heterosexual Anal Intercourse
Style file version Nov. 19th, 1999
361
engaged in anal intercourse as people who have not. Horvath and Zuckerman (1993) found that people high on sexual risk taking did not differ significantly from people low on sexual risk taking in their knowledge of activities that transmit AIDS. Gold et al. (1994) found that male homosexuals who engaged in unsafe anal intercourse were not less knowledgeable about the HIV risks associated with this behavior than were gay men who do not engage in unsafe anal intercourse. METHODS Instrument The six-page questionnaire used in this research was adapted from questionnaires used in three prior studies. It was designed to examine the thoughts, emotions, and activities of college students regarding sexual behavior, along with their attitudes and knowledge about AIDS (Baldwin and Baldwin, 1988; Baldwin, 1990, 1992). The methods used in developing and testing these questionnaires are explained in early publications (Baldwin et al., 1990, 1992). The behavior variables had a test–retest reliability of 0.88. The present questionnaire included the revised form of the Sexual Opinion Survey (SOS) measuring erotophobia– erotophilia (Fisher, 1988). Respondents and Procedure We mailed copies of the questionnaire to a random sample of approximately 11% of the undergraduate population of a medium-size university. To ensure that the racial/ethnic minorities on campus were adequately represented for statistical analysis, the sample was stratified in order to oversample the minorities, although these individuals were still selected randomly from within each group. The questionnaire was mailed via third-class mail to a total of 1779 undergraduate students, and 893 students responded, yielding a response rate of 50.2%. Because many of these students change addresses yearly or more often and third-class mail is not forwarded or returned if undeliverable, some 5% to 10% of students may not have received the questionnaire we mailed, making the estimated response rate approximately 53% to 55%. These data have only been used for the current study of anal intercourse. Before the data were analyzed, all students 30 years of age and older were omitted from the sample. This was done to make the present study more easily comparable with prior studies on college students, which tend to omit older students because their marital status, number of sexual partners, sexual behavior, and drug use are often substantially different from the majority of younger undergraduates. Also omitted from the analyses were all students who had never engaged in vaginal intercourse, because they did not provide information relevant to several practices
P1: FMN/FTT/FOP
P2: FQJ
Archives of Sexual Behavior [asb]
PL111-339
June 24, 2000
10:34
362
Style file version Nov. 19th, 1999
Baldwin and Baldwin
related to the core hypotheses of this study. The analysis included 8 homosexual and 23 bisexual students who had engaged in both vaginal and anal intercourse. Their sexual activities are relevant to the heterosexual population, including the heterosexual spread of HIV, because they have engaged in sexual intercourse with heterosexuals. As a result, the following report is based on the 647 nonvirgin undergraduate students under 30 years of age. RESULTS Descriptive Statistics Descriptive statistics were calculated for the final sample of nonvirgins used in the analysis and for selected variables of general interest (Table I). The sample was 37.2% male, and the average age of the respondents was 20.8 years. The ethnic/racial composition was 8.5% African American, 19.8% Asian American, 21.3% Latino, and 51.4% white. The students were predominantly from middleclass families, with an average annual income of $50,000 to $74,999. Frequency data are relevant because the literature on anal sexual behavior is deficient in these, especially for heterosexuals. Stepping back to view the total data set, including virgins and nonvirgins, 78% of the students reported having had vaginal intercourse, and 18% anal intercourse. Among our sample of nonvirgins, 22.9% reported that they had also engaged in anal intercourse, and the average reported age for first vaginal intercourse was 16.8 years, whereas the age of first anal intercourse was 18.8 years. Reported condom use during anal intercourse was low, averaging 20.9% of the time in the previous 3 months, compared with 42.9% of the time for vaginal intercourse in the same time period. Whereas 42.6% of people reported having never used condoms for vaginal intercourse in the previous 3 months, 76.1% reported no condom use for anal intercourse in that period of time. Most important, low rates of condom use during vaginal intercourse in the previous 3 months were reported more commonly by people with anal experience than those without it: 65.4% of the people who engaged in anal intercourse used condoms 25% of the time or less, compared to 45.2% of people who did not engage in anal intercourse. The average percentage of time that students reported condom use for vaginal intercourse was 30.1% for the people who engaged in anal intercourse and 47.2% for people who did not engage in anal intercourse. People reporting that they engaged in anal intercourse were also more likely to report having had at least one sexually transmitted disease (STD) and to have been tested for HIV than did people who did not report engaging in anal intercourse: 20.3% of people who engaged in anal intercourse reported having at least one STD and 41.1% had been tested for HIV, compared to 12.0% and 27.6%, respectively, for people who did not engage in anal intercourse.
P1: FMN/FTT/FOP
P2: FQJ
Archives of Sexual Behavior [asb]
PL111-339
June 24, 2000
10:34
Style file version Nov. 19th, 1999
Heterosexual Anal Intercourse
363
Table I. Descriptive Statistics for Coitally Experienced Respondents Variable
Range
Mean
Erotophobia–erotophiliaa Religious events frequency African-American Asian-American Latino Gender Parental income Age Seatbelt use Anal intercourse AIDS knowledge Semen–blood risky Anal sex risky Transmission risk rate Condom protection Believe self at risk Worry contract HIV Tested for HIV Drugs ever usedb Age first vaginal sex Age first anal sex Total number of vaginal partners Had coitus previous 3 months Anal sex previous 3 months Safety of birth control method used last coitus Mean condom use for coitus previous 3 months Mean condom use for anal sex previous 3 months Condom last coitus Had an STDc
19–124 (erotophobic to erotophilic) 1–6 (never–once a week or more) 0–1 (all others–African American) 0–1 (all others–Asian American) 0–1 (all others–Latino) 0–1 (female–male) 1–5 (under $25,000–$100,000 or more) 18–29 (18–29 years) 0–100 (0–100% of the time) 0–1 (no–yes) 0–19 (0–19 answers correct) 0–1 (not transmit–transmit) 0–1 (not transmit–transmit) 1–9 (99% or more–1% or less) 1–4 (no protection–total protection) 1–5 (very unlikely–very likely) 1–4 (not at all–great deal) 0–1 (no–yes) 0–9 (0–9 drugs) 8–24 (8–24 years) 8–25 (8–25 years) 1–99 (1–99 partners) 0–1 (no–yes) 0–1 (no–yes) 1–4 (85% faliure rate–3% or less)
78.587 2.600 0.085 0.198 0.213 0.372 3.272 20.786 92.776 0.229 16.847 0.960 0.937 3.640 2.760 2.249 2.136 0.307 2.482 16.828 18.772 6.616 0.763 0.301 3.262
0–100 (0–100% of time) 0–100 (0–100% of time) 0–1 (no–yes) 0–4 (0–4 STDs)
42.905 20.870 0.470 0.162
a The
erotophobia–erotophilia score was based on the 21 SOS statements, which were scored on a 7-point Likert-type scale ranging from strongly agree to strongly disagree. The summary score was created in the standard way (see Fisher, 1988). b The nine drugs recorded were alcohol, marijuana, cocaine or coke, amphetamines, designer drugs, barbiturates, tranquilizers, heroin, or other narcotics. c Respondents could select any of seven choices—chlamydia, herpes, gonorrhea, nongonoccocal urethritis, trichomoniasis, syphilis, and genital warts.
All nonvirgens were knowledgeable about the means of HIV transmission. Both groups—students who had participated in anal intercourse and those who had not—were similarly knowledgeable about the role of semen and anal intercourse in transmission. Both groups also knew that condoms provide good protection from HIV. Nonvirgens who had engaged in anal intercourse and those who had not engaged in anal intercourse had similar, although incorrect knowledge about the risk of HIV transmission from one act of vaginal intercourse with an infected partner: The majority of both groups believed their chances of contracting HIV
P1: FMN/FTT/FOP
P2: FQJ
Archives of Sexual Behavior [asb]
PL111-339
June 24, 2000
10:34
364
Style file version Nov. 19th, 1999
Baldwin and Baldwin
from one act of coitus with an infected partner was 75% or higher—although the actual risk is closer to 0.2%. Correlation Analysis Analyses of correlations among variables used in the regression analysis revealed that, as hypothesized, people reporting that they engaged in anal intercourse were more likely to be erotophilic than the other nonvirgins (r = 0.206; p < .0001). The correlation data did not support hypothesis two or three relevant to religiosity and ethnicity/race. Hypothesis four was also not supported. Indeed, it was contradicted: Gender yielded a significant finding that was unpredicted: Males were more likely to engage in anal intercourse than females (r = 0.106; p < .01). Contrary to hypothesis five, socioeconomic status did not correlate with anal intercourse. Supporting hypothesis six, anally experienced individuals tended to be older than nonvirgins who did not report engaging in anal sex (r = .244; p < .0001). Correlation data partially supported the seventh hypothesis about sensation seeking and risk taking. People who engaged in anal intercourse were more likely to be younger at first vaginal intercourse (r = −0.117; p < 0.01) and to have had more vaginal intercourse partners than nonvirgins who had not engaged in anal sex (r = 0.220; p < .0001). They also were more likely to have engaged in vaginal intercourse in the previous 3 months (r = 0.149; p < .001). However, contrary to hypothesis seven, people who engaged in anal intercourse were not significantly less likely to use seatbelts (r = −0.066). Also contrary to prediction, people reporting engaging in anal intercourse did not use riskier methods of contraception than those not reporting anal sex (r = −0.064). However, they took other risks, reporting less condom use at last vaginal intercourse (r = −0.166; p < .0001): Only 31.8% of the people who engaged in anal sex reported using condoms the last time they had vaginal intercourse, compared with 51.5% of the people who engaged in anal sex. As predicted, people who indicated that they engaged in anal sex were more likely to have used a variety of drugs than people who did not indicate such experience (r = 0.183; p < .0001). As hypothesized, students who reported engaging in anal intercourse did not worry more about contracting HIV (r = −0.016) or judge themselves to be at greater risk for contracting HIV (r = 0.037), nor were they less knowledgeable about HIV transmission than the other nonvirgins (r = 0.011). Regression Analysis A simultaneous regression analysis was used to evaluate the unique contribution of each variable related to anal intercourse because it could partial out the variance accounted for by the other variables. Because the dependent
P1: FMN/FTT/FOP
P2: FQJ
Archives of Sexual Behavior [asb]
PL111-339
June 24, 2000
10:34
Style file version Nov. 19th, 1999
Heterosexual Anal Intercourse
365
variable—having engaged in anal intercourse—is a dichotomous variable, logistic regression analysis was used. The regression analysis included all people who had engaged in vaginal intercourse, not just heterosexuals or people who had done so in the previous 3 months. A total of 18 independent variables were included in the regression equation to examine their possible effects on participation in anal intercourse. Because the hypotheses developed in this study came from such disparate sources, we have no rationale for justifying any particular ordering for conducting a stepwise regression. The regression analysis indicated that the dependent variable, engaging in anal intercourse or not, was significantly affected by the following: erotophobia– erotophilia, current age, age at first vaginal intercourse, having engaged in vaginal intercourse in previous 3 months, effectiveness of contraception used, and condom used at last coitus (Table II). The people who had participated in anal intercourse were more likely to be more erotophilic than people who had not engaged in anal sex, to be older now, to have begun vaginal intercourse at an earlier age, to have engaged in vaginal intercourse in the previous 3 months, to have used less effective methods of birth control, and not to have used a condom at last vaginal intercourse. The first hypothesis about erotophobia–erotophilia was supported. In fact, 14.3% of the erotophobes (who are defined as measuring less than 66 on the SOS scale) had engaged in anal intercourse, compared with 34.4 % of the erotophiles Table II. Predictors for Engaging in Anal Intercourse Variable
Parameter estimate
Erotophobia–erotophilia Religious events frequency African-American Asian-American Latino Gender Parental income Age Seatbelt use Drugs ever used Age first vaginal sex Total number of vaginal partners Had coitus previous 3 months Safety of birth control method used last coitus Condom last coitus Worry contract HIV Believe self at risk AIDS knowledge Constant
0.0204a 0.1145 0.0529 0.3424 0.1520 0.0794 −0.0641 0.2606b −0.0019 0.0742 −0.1332c 0.0128 1.0758a −0.2593c −0.5865c −0.0752 0.0472 0.0394 −6.9250b
Chi-square for covariates p = 0.0001. < 0.001. < 0.0001. < 0.05.
ap bp cp
P1: FMN/FTT/FOP
P2: FQJ
Archives of Sexual Behavior [asb]
PL111-339
June 24, 2000
10:34
366
Style file version Nov. 19th, 1999
Baldwin and Baldwin
(who are defined as measuring more than 95 on the SOS scale). The second hypothesis was not supported: The frequency of attending religious services did not affect anal intercourse. Relevant to the third and fourth hypotheses, neither race/ethnicity nor gender was related to anal intercourse. Of the people who reported engaging in vaginal intercourse, anal intercourse was reported by 25.4% of Latinos, 23.4% of Asian Americans, 22.1% of whites, and 20.9% of African Americans. The regression analysis did not find an effect of gender on engaging in anal intercourse, even though the correlation analysis did. Contrary to our fifth hypothesis, the socioeconomic status of one’s parents (measured by income) was not related to anal intercourse. In support of the sixth hypothesis, older students were more likely to report engaging in anal intercourse. Anal intercourse was reported among 6.9% of those younger than 19 and 48.0% of those older than 24. Several subsets of the seventh hypothesis about sensory stimulation were supported. People who like and seek out novel sensations would be expected to begin vaginal intercourse at an earlier age than those who do not, and 33.1% of people reporting anal experience began vaginal intercourse at age 15 or younger, compared with 20.8% of those not reporting engaging in anal sex. Also statistically significant was the finding that, during the previous 3 months, people who engaged in anal sex were more likely than those who did not to have engaged in vaginal intercourse: 87.8% of those reporting anal intercourse reported having vaginal intercourse in the previous 3 months, compared to 72.8% of those not reporting engaging in anal intercourse. As predicted, people who engaged in anal sex took more risks with vaginal intercourse, being less likely to use contraception or using less effective means of contraception. However, contrary to prediction, the people reporting engaging in anal intercourse did not have more vaginal intercourse partners in the regression analysis (although they did in the correlation analysis). It was predicted that those who had used more drugs would be more likely to engage in anal intercourse, but the regression analysis did not support this. The data on a nonsexual form of risk taking—namely, seatbelt use—did not support the hypothesis that some people have a generalized tendency to take risks. The predictions related to beliefs and worry about contracting HIV were supported. People reporting engaging in anal sex did not worry more than those who did not about contracting HIV as a result of their sexual activity, nor did they believe they were at greater risk for contracting the virus than others. Also as predicted, students who had engaged in anal intercourse were not less knowledgeable about HIV transmission than the other students. DISCUSSION The limitations to this study must be considered before drawing conclusions from the statistical analyses. Although we attempted to obtain a random sample of
P1: FMN/FTT/FOP
P2: FQJ
Archives of Sexual Behavior [asb]
PL111-339
Heterosexual Anal Intercourse
June 24, 2000
10:34
Style file version Nov. 19th, 1999
367
students, the estimated response rate was 53% to 55%; hence, the randomness of the sample is not assured. Bogaert (1996), Catania et al. (1990), and others found that more liberal and sexually experienced people are more likely to volunteer for studies of sexual behavior than are conservative people, which could bias our results to show more sexual activity than would have been found if all students had responded. There are at least four other problems: First, all self-reported data can be held suspect because people can construct these data without the constraints of needing to provide empirical proof. Second, respondents tend to underreport illegal or stigmatized behavior such as drug use and anal intercourse (Mensch and Kandel, 1988). Third, it is difficult to know how far our findings generalize to other college campuses, much less to noncollege populations. Laumann et al. (1994) found in their national sample of 18- to 59-year-old people that, among heterosexuals, 15.8% of males and 16.2% of females in the 18–24 age group reported ever engaging in anal intercourse, which is lower than the percentages typically reported in college studies. Fourth, our study did not explore the possibility that some students may engage in anal intercourse as a method of birth control without considering that it might be risky for other reasons. Perhaps, some couples turn to anal intercourse as a way to have sexual relations while assuring the female remains a vaginal virgin, or a “technical virgin.” In our sample, there were no vaginal virgins who reported engaging in anal intercourse. Nevertheless, some females may have begun anal intercourse for this reason, then later added vaginal activities. In spite of these limitations, the present study provides important information about an underresearched topic—heterosexual anal sexual behavior and its correlates. The present study reveals a substantial amount of risky anal intercourse because 22.9% of the nonvirgins reported having tried it. This percentage is in line with those given by various other studies on college campuses (Cochran et al., 1991; Hsu et al., 1994; Hutchison, 1994, 1996; Kotloff et al., 1991; MacDonald et al., 1990; Reinisch et al., 1992) and from young noncollege populations (Catania et al., 1989; Heffernan et al., 1996; Langille et al., 1994; Norris et al., 1996; Stanton et al., 1994). However, most of those studies did not examine anal intercourse in detail. Our first hypothesis was supported: People who engaged in anal sex scored higher on erotophilia. However, the direction of the causal arrows cannot be determined: Erotophilia might lead a person to try many sexual activities that erotophobes would not. However, a peer group that rewarded individuals for exploring anal intercourse and various other sexual practices might foster erotophilia. In addition, both erotophilia and anal experience could result from other causes, such as high sensation-seeking behavior, which results in high risk taking. For example, Zuckerman (1994) reports that high stimulation seekers experiment with a larger number of different kinds of sexual activities than do low stimulation seekers.
P1: FMN/FTT/FOP
P2: FQJ
Archives of Sexual Behavior [asb]
368
PL111-339
June 24, 2000
10:34
Style file version Nov. 19th, 1999
Baldwin and Baldwin
Hypothesis two was not supported: A high frequency of attendance at religious services was not associated with low incidence of anal intercourse in either the correlation or regression analyses. Davidson et al. (1995) found that greater frequency of religious attendance did not decrease participation in anal intercourse. Similarly, Daugherty and Burger (1984) found little relationship between the sexual values taught at church and students’ sexual and contraceptive behaviors. As we anticipated when creating the regression model, hypotheses three and four were not supported. None of the traditional background variables—ethnicity, race, or gender—was associated with anal intercourse. Although males were more likely than females to engage in anal intercourse according to the correlation statistics, there was no gender difference for engaging in anal intercourse when other variables were controlled for in the regression analysis. In our study, 19.5% of the nonvirgin females and 28.6% of the nonvirgin males reported engaging in anal intercourse. This is somewhat different from other college studies: In a convenience sample of college students in the Midwest, Reinisch et al. (1992) reported that 19% of nonvirgin men and 22% of nonvirgin women had engaged in anal intercourse. In a large nationally representative survey of first-year college students across Canada, MacDonald et al. (1990) found that 19% of the nonvirgin women and 14% of the nonvirgin men had engaged in anal intercourse. Hypothesis five was not supported: Although parents’ income was predicted to affect anal intercourse, it may be that the sample population was too homogeneous on this variable to find an effect. Alternatively, people who attend college may be similar, regardless of parental income; or the college environment may produce similar student behavior, regardless of economic background. Hypothesis six was supported: Being older increased the likelihood of engaging in anal intercourse, even when controlling for the age that people began vaginal intercourse. This suggests that merely having more years of life experience—not specifically coital experience—results in increased anal intercourse. Seventh was a series of hypotheses related to sensation seeking and risk taking: Some were supported, others not. People reporting engaging in anal sex took more risks than others, although not all possible risks. People who engaged in anal sex were less likely to use contraception and used less reliable means of contraception than people not reporting engaging in anal sex (regression analysis only). They were less likely to have used condoms during the last act of vaginal intercourse than people who did not engage in anal sex: 68.2% of people reporting engaging in anal sex did not use a condom at last coitus, compared to 48.5% of the people not reporting this behavior. People who engaged in anal sex also began vaginal intercourse earlier than those without it; and in the previous 3 months, they were more likely to report engaging in vaginal intercourse. These findings indicate that people who report engaging in anal intercourse take risks in several important sectors of their sexual lives. Zuckerman (1974, 1994) reports data indicating that people who like high levels of sensory stimulation tend to explore sex and drugs—and take more risks
P1: FMN/FTT/FOP
P2: FQJ
Archives of Sexual Behavior [asb]
PL111-339
Heterosexual Anal Intercourse
June 24, 2000
10:34
Style file version Nov. 19th, 1999
369
with these activities—than do low sensation seekers. Our study finds that people who had engaged in anal intercourse reported more vaginal sexual partners and used more drugs than people who did not report engaging in this behavior, but both variables were only significant in the correlation analysis, without controls for confounding variables. Zuckerman (1994) argues that genetics play some role in making people high or low on sensation seeking and risk taking, but nature and nurture both make their contributions. For example, positive reinforcement from peers for doing exciting things increases a person’s chances of seeking and enjoying high sensory stimulation, whereas failure, injury, ridicule, or other forms of punishment associated with sensation-seeking behavior can temper or inhibit it (Baldwin and Baldwin, 2001). In our study, people reporting engaging in anal sex displayed an astonishing lack of concern for the transmission of HIV via anal interactions, judging by the small percentage of time they used condoms during anal intercourse. Although anal intercourse is riskier than vaginal intercourse, condom use averaged only 20.9% of the time for anal intercourse, compared with 42.9% for vaginal intercourse, in the previous 3 months. Only 15.2% of individuals who engaged in anal sex always used condoms for anal intercourse, and 76.1% never did so. In a study of college students, Hutchison (1994, 1996) found that of those who engaged in anal intercourse in the previous 3 months, 26% used condoms 90% to 100% of the time for anal sex, whereas 68% did not use condoms during anal intercourse. Although most students know intellectually that anal intercourse can transmit HIV and that condoms provide good protection from transmission, they behave as if they were ignorant of this. College students can create “illusions of safety” when engaging in risky behavior (Thompson et al., 1996). They can do this by forming reassuring ideas: “I’m in a monogamous relationships so I won’t contract HIV”; “I know my partner’s sexual history”; or “I’d know it if I met a person who was HIV positive and I’d never sleep with that individual.” Such comforting thoughts allow people to rationalize that they are safe, even though the logic is flimsy. If people believe that scientific risk data do not apply to them, why should they worry? We suspect that standard AIDS education has failed to convince many students that they are at risk during anal intercourse because most AIDS education for heterosexuals fails to include open and honest talk about anal sex, based on scientific evidence. The frequency of condom use during anal intercourse is an important variable, but most other studies on college samples do not specifically collect the relevant data. Some research on noncollege populations finds low condom use for anal intercourse. In a study of young men across the United States, 9% of those who were sexually active engaged in heterosexual anal intercourse in the previous year, but only 20% always used a condom and 60% never did so (Ku et al., 1993). In a household probability sample of California adults, Erickson et al. (1995) found that 60% of people who participated in anal intercourse at least once a month in the previous year had never used condoms. In their study of adult clients at an STD
P1: FMN/FTT/FOP
P2: FQJ
Archives of Sexual Behavior [asb]
PL111-339
June 24, 2000
10:34
Style file version Nov. 19th, 1999
370
Baldwin and Baldwin
clinic, Baker et al. (1995) also found condom use during anal sex to be low: About 25% of people reported engaging in anal intercourse in the previous 3 months, but only 11% always used a condom, and 67% never did so. In the present study, people reporting engaging in anal intercourse used condoms less frequently during vaginal intercourse in the previous 3 months than did other people: 65.3% of people engaging in anal sex used condoms 25% of the time or less, compared with 45.2% of the people who did not engage in anal sex. Clearly, a substantial percentage of young people are engaging in activities that put them at risk of contracting HIV, and the risk takers would be more likely than others to transmit HIV to a partner during vaginal intercourse because they are less likely to use condoms during vaginal intercourse. Overall, this and other studies indicate that health educators have not presented enough scientfic data to convince heterosexuals about the risks of anal intercourse and the need to abstain from—or use condoms during—this activity. Roughly 20% to 25% of sexually active college students report experience with anal intercourse, but most do not use condoms during this activity. Although some use condoms during vaginal intercourse, they use condoms less often during anal intercourse, which puts them at risk of contracting and transmitting STDs, including HIV. The traditional emphasis of sex education on teaching heterosexuals to abstain from sexual relations or use condoms during vaginal intercourse may not be adequate for communicating the need to use condoms for—or to abstain from—anal intercourse. Perhaps educators have not been as explicit as is needed about anal sex. If we feel awkward or ashamed to talk openly and candidly about anal sex, we must realize that approximately one-fifth to one-quarter of sexually active young adults are not ashamed to do it. This and other studies suggest that sex education must focus especially on the high sensation-seeking and risk-taking populations. If we fail to talk openly and honestly about anal sex and its serious health risks, we are leaving between one in four and one in five of young adults ill-informed about a behavior they may try. In the era of AIDS, education about anal intercourse is essential for all students, because even those who do not engage in anal interactions can be peer sex educators for others. Everyone needs to know the empirical facts relevant to all serious diseases. For heterosexuals, the risk of vaginal intercourse receives considerably more attention than do risks of anal intercourse, and this may give many students the false impression that anal intercourse is not problematic. We must confront and refute this false impression directly.
ACKNOWLEDGMENTS We thank Dr. Carol Geer, Director of Counseling and Career Services; Ms. D’Anne Kinkaid, research assistant; Ms. Kwen Kuhns, administrative analyst;
P1: FMN/FTT/FOP
P2: FQJ
Archives of Sexual Behavior [asb]
PL111-339
June 24, 2000
10:34
Heterosexual Anal Intercourse
Style file version Nov. 19th, 1999
371
and Dr. Michael Loewy, AIDS and Sexual Health Services Coordinator, for their invaluable assistance in conducting this study. REFERENCES Aggleton, P., O’Reilly, K., Slutkin, G., and Davies, P. (1994). Risking everything? Risk behavior, behavior change, and AIDS. Science 265: 341–345. Baker, S. A., Morrison, D. M., Gillmore, M. R., and Schock, M. D. (1995). Sexual behaviors, substance use, and condom use in a sexually transmitted disease clinic sample. J. Sex Res. 32: 37– 44. Baldwin, J. D., and Baldwin, J. I. (1988). Factors affecting AIDS-related risk taking behaviors among college students. J. Sex Res. 25: 181–196. Baldwin, J. D., and Baldwin, J. I. (2001). Behavior Principles in Everyday Life, 4th ed., Upper Saddle River, NJ: Prentice Hall. Baldwin, J. I., Whiteley, S., and Baldwin, J. D. (1990). Changing AIDS and fertility-related behavior: The effectiveness of sexual education. J. Sex Res. 27: 245–262. Baldwin, J. D., Whiteley, S., and Baldwin, J. I. (1992). The effect of ethnic group on sexual activities related to contraception and STDs. J. Sex Res. 29: 189–205. Belcastro, P. A. (1985). Sexual behavior differences between black and white students. J. Sex Res. 21: 56–67. Bogaert, A. F. (1996). Volunteer bias in human sexuality research: Evidence for both sexuality and personality differences in males. Arch. Sex Beh. 25: 125–140. Catania, J. A., Dolcini, M. M., Coates, T. J., Kegeles, S. M., Greenblatt, R. M., and Puckett, S. (1989). Predictors of condom use and multiple partnered sex among sexually-active adolescent women: Implications for AIDS-related health interventions. J. Sex Res. 26: 514–525. Catania, J. A., Gibson, D., Marin, B., Coates, T., and Greenblatt, R. (1990). Response bias in assessing sexual behaviors relevant to HIV transmission. Eval. Prog. Plann. 13: 19–29. Cl´ement, R., and Jonah, B. A. (1984). Field dependence, sensation seeking and driving behavior. Person. Individ. Diff. 5: 87–93. Cochran, S. D., Mays, V. M., and Leung, L. (1991). Sexual practices of heterosexual Asian-American young adults: Implications for risk of HIV infection. Arch. Sex. Beh. 20: 381–391. Daugherty, L., and Burger, J. (1984). The influence of parents, church, and peers on the sexual attitudes and behaviors of college students. Arch. Sex. Beh. 13: 351–359. Davidson, J. K., Darling, C. A., and Norton, L. (1995). Religiosty and the sexuality of women: Sexual behavior and sexual satisfaction revisited. J. Sex Res. 32: 235–243. DeBuono, B. A., Zinner, S. H., Daamen, M., and McCormack, W. M. (1990). Sexual behavior of college women in 1975, 1986, and 1989. N. Engl. J. Med. 322: 821–835. Earle, J. R., and Perricone, P. J. (1986). Premarital sexuality: A ten-year study of attitudes and behavior on a small university campus. J. Sex Res. 22: 304–310. Erickson, P. I., Bastani, R., Maxwell, A. E., Marcus, A. C., Capell, F. J., and Yan, K. X. (1995). Prevalence of anal sex among heterosexuals in California and its relationship to other AIDS risk behaviors. AIDS Educ. Prev. 7: 477–493. Fisher, W. A. (1988). The Sexual Opinion Survey. In Davis, C. M., Yarber, W. L., and. Davis, S. L (eds.), Sexuality-Related Measures, Graphic Publishing Co., Lake Mills, Iowa. Fisher, W. A., Byrne, D., and White, W. A. (1983). Emotional barriers to contraception. In Byrne, D., and Fisher, W. A. (eds.), Adolescents, Sex, and Contraception, Lawrence Erlbaum Associates, Hillsdale, New Jersey, pp. 207–239. Fleuridas, C., Creevy, K., and Vela, E. (1997). Sexual risk-taking in college students and functional families of origin. Fam. Sys. Health 15: 185–202. Gilbert, L., and Alexander, L. (1998). A profile of sexual health behaviors among college women. Psychol. Rep. 82: 107–116. Gold, R. S., Skinner, M. J., and Ross, M. W. (1994). Unprotected anal intercourse in HIV-infected and non-HIV-infected gay men. J. Sex Res. 31: 59–77.
P1: FMN/FTT/FOP
P2: FQJ
Archives of Sexual Behavior [asb]
372
PL111-339
June 24, 2000
10:34
Style file version Nov. 19th, 1999
Baldwin and Baldwin
Graves, K. L., and Leigh, B. C. (1995). The relationship of substance use to sexual activity among young adults in the United States. Fam. Plann. Perspect. 27: 18–22. Harvey, S. M., and Spigner, C. (1995). Factors associated with sexual behavior among adolescents: A multivariate analysis. Adolescence 30: 253–264. Heffernan, R., Chiasson, M. A., and Sackoff, J. E. (1996). HIV risk behaviors among adolescents at a sexually transmitted disease clinic in New York City. J. Adol. Health 18: 429–434. Hersch, J., and Viscusi, W. K. (1990). Cigarette smoking, seatbelt use, and differences in wage–risk tradeoffs. J. Hum. Res. 25: 202–227. Horvath, P., and Zuckerman, M. (1993). Sensation seeking, risk appraisal, and risky behavior. Person. Individ. Diff. 14: 41–52. Hovell, M. F., Hillman, E. R., Blumberg, E., Sipan, C., Atkins, C., Hofstetter, C. R., and Myers, C. A. (1994). A behavioral–ecological model of adolescent sexual development: A template for AIDS prevention. J. Sex Res. 31: 267–281. Hsu, B., Kling, A., Kessler, C., Knapke, K., Diefenbach, P., and Elias, J. E. (1994). Gender differences in sexual fantasy and behavior in a college population: A ten-year replication. J. Sex Mar. Ther. 20: 103–118. Huang, K., and Uba, L. (1992). Premarital sexual behavior among Chinese college students in the United States. Arch. Sex. Beh. 21: 227–240. Hunt, M. (1974). Sexual Behavior in the 1970’s, Playboy Press, Chicago. Hutchison, B., Jr. (1994). Social psychological aspects of AIDS-related sexual risk-taking behavior among college students. MA thesis, Sam Houston State University, Texas. Hutchison, B., Jr. (1996). Risky and safe sex: Heterosexual youth, gender differences, and sexual risk-taking. Soc. Rev. 6(2): 30–33. Johnson, E. H., Gilbert, D., and Lollis, C. (1994). Characteristics of African-American college students with HIV/AIDS. J. Natl. Med. Assoc. 86: 931–940. Keller, J. F., Elliott, S. S., and Gunberg, E. (1982). Premarital sexual intercourse among single college students: A discriminant analysis. Sex Roles 8: 21–32. Kotloff, K. L., Tacket, C. O., Wasserman, S. S., Bridwell, M. W., Cowan, J. E., Clemens, J. D., Brothers, T. A., O’Donnell, S. A., and Quinn, T. C. (1991). A voluntary serosurvey and behavioral risk assessment for human immunodeficiency virus infection among college students. Sex. Trans. Dis. Oct/Dec: 223–227. Ku, L., Sonenstein, F. L., and Pleck, J. H. (1993). Young men’s risk behaviors for HIV infection and sexually transmitted diseases, 1988 through 1991. Am. J. Pub. Health 83: 1609–1615. Langille, D. B., Beazley, R., Shoveller, J., and Johnston, G. (1994). Prevalence of high risk sexual behavior in adolescents attending school in a county in Nova Scotia. Can. J. Pub. Health 85: 227–230. Laumann, E. O., Gagnon, J. H., Michael, R. T., and Michaels, S. (1994). The Social Organization of Sexuality: Sexual Practices in the United States. University of Chicago Press, Chicago. Lazzarin, A., Saracco, A., Musicco, M., and Nicolosi, A. (1991). Man-to-woman sexual transmission of the human immunodeficiency virus: Risk factors related to sexual behavior, man’s infectiousness, and women’s susceptibility. Arch. Intern. Med. 151: 2411–2416. MacDonald, N. E., Wells, G. A., Fisher, W. A., Warren, W. K., King, M. A., Doherty, J.-A. A., and Bowie, W. R. (1990). High-risk STD/HIV behavior among college students. JAMA 263: 3155– 3159. Mahoney, E. R. (1980). Religiosity and sexual behavior among heterosexual college students. J. Sex Res. 16: 97–113. Mayer, K. H., and Anderson, D. J. (1995). Heterosexual HIV transmission. Inf. Ag. Dis. 4: 273–284. Mensch, E., and Kandel, D. B. (1988). Youth’s underreporting of substance use. Pub. Opin. Quart. 52: 100–124. Murstein, B. I., and Mercy, T. (1994). Sex, drugs, relationships, contraception, and fears of disease on a college campus over 17 years. Adolescence 9: 292–310. Norris, A. E., Ford, K., Shyr, Y., and Schork, M. A. (1996). Heterosexual experiences and partnerships of urban, low-income African-American and Hispanic youth. J. Acqu. Im. Def. Syn. Hum. Retro. 11: 288–300. Overby, K. J., and Kegeles, S. M. (1994). The impact of AIDS on an urban population of high-risk female minority adolescents: Implications for intervention. J. Adol. Health 15: 216–227. Prince, A., and Bernard, A. L. (1998). Sexual behaviors and safer sex practices of college students on a commuter campus. J. Am. Coll. Health 47: 11–21.
P1: FMN/FTT/FOP
P2: FQJ
Archives of Sexual Behavior [asb]
PL111-339
Heterosexual Anal Intercourse
June 24, 2000
10:34
Style file version Nov. 19th, 1999
373
Reinisch, J. M., Sanders, S. A., Hill, C. A., and Ziemba-Davis, M. (1992). High-risk sexual behavior among heterosexual undergraduates at a midwestern university. Fam. Plann. Perspect. 24: 116– 121. Robinson, I., Ziss, K., Ganza, B., Katz, S., and Robinson, E. (1991). Twenty years of the sexual revolution, 1965–1985: An update. J. Marr. Fam. 53: 216–220. Roche, J. P., and Ramsbey, T. W. (1993). Premarital sexuality: A five-year follow-up study of attitudes and behavior by dating stage. Adolescence 28: 67–80. Rotheram-Borus, M. J., Rosario, M., Meyer-Bahlburg, H. F. L., Koopman, C., Dopkins, S. C., and Davies, M. (1994). Sexual and substance use acts of gay and bisexual male adolescents in New York City. J. Sex Res. 31: 47–57. Seidman, S. N., Mosher, W. D., and Aral, S. O. (1994). Predictors of high-risk behavior in unmarried American women: Adolescent environment as risk factor. J. Adol. Health 15: 126–132. Senf, J. H., and Price, C. Q. (1994). Young adults, alcohol and condom use: What is the connection? J. Adol. Health 15: 238–244. Shapiro, J., Radecki, S., Charchian, A. S., and Josephson, V. (1999). Sexual behavior and AIDS-related knowledge among community college students in Orange County, California. J. Comm. Health 24: 29–43. Silverman, B. G., and Gross, T. P. (1997). Use and effectiveness of condoms during anal intercourse: A review. Sex. Trans. Dis. 21: 11–17. Small, S. A., and Luster, T. (1994). Adolescent sexual activity: An ecological, risk-factor approach. J. Marr. Fam. 56: 181–192. Stanton, B., Xiaoming, L., Black, M., Ricardo, I., and Galbraith, J. (1994). Anal intercourse among preadolescent and early adolescent low-income urban African-Americans. Arch. Pediat. Adol. Med. 148: 1201–1204. Taylor, S. E., Diloria, C., Stephens, T. T., and Soet, J. E. (1997). A comparison of AIDS-related sexual risk behaviors among African-American college students. J. Natl. Med. Assoc. 89: 397–403. Thompson, S. C., Anderson, K., Freedman, D., and Swan, J. (1996). Illusions of safety in a risk world: A study of college students’ condom use. J. Appl. Soc. Psychol. 26: 189–210. Thornton, A., and Camburn, D. (1989). Religious participation and adolescent sexual behavior and attitudes. J. Marr. Fam. 51: 641–653. Turner, J. C., Korpita, E., Mohn, L. A., and Hill, W. B. (1993). Reduction in sexual risk behaviors among college students following a comprehensive health education intervention. J. Am. Coll. Health 41: 187–193. Valois, R. F., Kammermann, S. K., and Drane, J. W. (1997). Number of sexual intercourse partners and associated risk behaviors among public high school adolescents. J. Sex. Educ. Ther. 22: 13–22. Voeller, B. (1991). AIDS and heterosexual anal intercourse. Arch. Sex. Beh. 20: 233–276. Weinberg, M. S., Lottes, H. L., and Aveline, D. (1998). AIDS risk reduction strategies among United States and Swedish heterosexual university students. Arch. Sex Behav. 27: 385–401. White, H. R., and Johnson, V. (1988). Risk taking as a predictor of adolecent sexual activity and use of contraception. J. Adol. Res. 3: 317–331. Zuckerman, M. (1974). The sensation seeking motive. In Maher, B. A. (ed.), Progress in Experimental Personality Research, Vol. 7, Academic Press, New York, pp. 79–148. Zuckerman, M. (1976). Sensation seeking and anxiety, traits and states, as determinants of behavior in novel situations. In Sarason, I. G., and Spielberger, C. D. (eds.), Stress and Anxiety, Vol. 3, Hemisphere, Washington, DC, pp. 141–170. Zuckerman, M. (1979). Sensation seeking and risk taking. In Izard, C. E. (ed.), Emotions in Personality and Psychopathology, Plenum, New York, pp. 163–197. Zuckerman, M. (1994). Behavioral Expressions and Biosocial Bases of Sensation Seeking, Cambridge University Press, New York. Zuckerman, M., Bone, R. N., Neary, R., Mangelsdorff, D., and Brustman, D. (1972). What is the sensation seeker? Personality trait and experience correlates of the Sensation Seeking Scales. J. Consult. Clin. Psychol. 39: 308–321. Zuckerman, M., Tushup, R., and Finner, S. (1976). Sexual attitudes and experience: Attitude and personality correlates and changes produced by a course in sexuality. J. Consult. Clin. Psychol. 44: 7–19.
P1: FNN/FTT
P2: FLF
Archives of Sexual Behavior [asb]
PL111-340
June 17, 2000
13:46
Style file version Nov. 19th, 1999
Archives of Sexual Behavior, Vol. 29, No. 4, 2000
Attitudes Toward Transsexualism in a Swedish National Survey Mikael Land´en, M.D., Ph.D.,1,2 and Sune Innala, Ph.D.1
A general inventory of the views on sex reassignment and attitudes toward transsexuals in Sweden was attemped. Whether the view on these matters differ between people embracing biological theories in explanation of transsexualism and those embracing psychological theories was tested. Third, whether men and the older age groups hold a different view on transsexualism than women and younger age groups was investigated. For these purposes, in October–December 1998, a questionnaire was mailed to a randomly selected national sample of 992 Swedish residents; 668 persons returned the questionnaire, giving a 67% response rate. Results showed that a majority supports the possibility for transsexuals to undergo sex reassignment; however, 63% thought that the individual should bear the expenses for it. In addition, a majority supported transsexuals’ right to get married in their new sex and their right to work with children. Transsexuals’ right to adopt and raise children was supported by 43%, whereas 41% were opposed. Results indicated that those who believed that transsexualism is caused by biological factors had a less restrictive view on transsexualism than people who held a psychological view. Men and the older age group were found to hold a more restrictive view on these issues than women and the younger age group. Future studies should address these questions to elucidate differences between cultures and the process of change in attitudes over time. KEY WORDS: gender identity disorder; Sweden; attitudes; national sample; ethics.
INTRODUCTION The treatment of choice for a selected group of transsexuals is sex reassignment (for a recent review, see Cohen-Kettenis and Gooren, 1999). This includes 1 Section
of Psychiatry, Institute of Clinical Neuroscience, G¨oteborg University, Sweden. whom correspondence should be addressed at Institute of Clinical Neuroscience, Section of Psychiatry, Sahlgrenska University Hospital/M¨olndal, SE-431 80 M¨olndal, Sweden; e-mail:
[email protected].
2 To
375 C 2000 Plenum Publishing Corporation 0004-0002/00/0800-0375$18.00/0 °
P1: FNN/FTT
P2: FLF
Archives of Sexual Behavior [asb]
PL111-340
June 17, 2000
13:46
Style file version Nov. 19th, 1999
376
Land´en and Innala
medical (hormonal) and surgical measures to change the physical appearance of the body, and legal measures to change the person’s legal status of being a woman or a man. In Sweden, a law regulating sex reassignment came into force in 1972. Ever since, sex reassignment costs are covered by the public health care system and the sex-reassigned person receives the rights and duties of the new sex, including the right to get married and adopt and raise children. Yet, sex reassignment is a controversial issue and raises a number of ethical questions among many people (e.g., matrimonial issues, child custody issues, whether the treatment involved should be publicly funded). People’s attitudes toward transsexuals and whether transsexuals encounter prejudices and discrimination in society are of importance for the transsexuals’ quality of life. Aims with this study were threefold. First, to make a general inventory of the ethical views on sex reassignment and attitudes towards transsexuals in Sweden. Second, to test our main hypothesis, derived from studies of attitudes about homosexuality (Ernulf et al., 1989), that the view on transsexuals might differ between people depending on whether they embrace biological or psychological theories in explanation of transsexualism. Third, to test if the attitudes towards transsexuals differ between men and women, and between younger and older age groups. The present study is to our knowledge the first to address these questions in a national sample survey of lay people. MATERIALS AND METHODS Study Sample and Methods of Selection The study group comprised individuals 18–70 years of age in Sweden. A sample of 992 persons was randomly selected from the national registration. Information about the person’s civil status, age, sex, and income was also retrieved from the national registration. Questionnaire A questionnaire comprising 13 questions was distributed by mail together with a cover letter explaining the aim of the study. On the cover letter, transsexualism was briefly defined and the demarcation against transvestitism was explained. The wording of this definition in translation read as follows: Transsexualism occurs in both men and women, and is characterized by a gender identity of the opposite sex. A transsexual person is often said to be trapped in the body of the wrong sex, and have a strong desire to live and be accepted as a member of the opposite sex and to “change sex.” A sex change implies a new name, treatment with the hormones of the opposite sex, and surgery of the genitals to make his or her body as congruent as possible with the preferred sex.
P1: FNN/FTT
P2: FLF
Archives of Sexual Behavior [asb]
PL111-340
June 17, 2000
13:46
Style file version Nov. 19th, 1999
Attitudes Toward Transsexuals
377
Transsexualism is not the same as transvestism, which refer to men who occasionally dress in women’s clothes. A transvestite does not wish to change sex.
It was emphasized that participation was voluntary and that the data were to be analyzed anonymously throughout the study. Telephone numbers to persons responsible for the project were specified. No pecuniary reward was offered to the participants.
Data Collection The questionnaire was sent out on October 16, 1998. On October 29, a reminder was sent out by mail. Thereafter, two additional reminders were sent, along with a new questionnaire. The data collection closed on December 16, 1998. Two male respondents expressed their indignation of the form and chose not to answer the questions. One person was mentally retarded and the questionnaire was therefore returned unanswered.
Statistics The Pearson’s chi-square test was used for determining the significance of the relationship between categorical variables. The multiple comparisons made in this study imply a risk for Type I error, which can be adjusted for by, for example, the Bonferroni method. However, the Bonferroni method is too conservative for a large number of comparisons (Altman, 1991). We therefore chose a reasonably conservative correction in employing p < .005 instead of p < .05 as the level of significance for rejecting the null hypothesis. RESULTS A total of 668 questionnaires was returned, resulting in an overall response rate of 67%. Results concerning demographics are displayed in Table I. Significantly more women (72%) than men (63%) returned the questionnaire (chi-square, df = 1, p < .01). The other demographic characteristics of the respondents returning the questionnaires and the ones failing to do so were fairly evenly dispersed among the two groups. All respondents’ answers to the questionnaire items are shown in Table II, in absolute numbers as well as in percentages. Group comparisons between people who believed in a biological explanation of transsexualism versus people who believed in a psychological explanation were also conducted. The items “You are born that way” formed the biological explanation group (n = 351), whereas the items “You choose to be
P1: FNN/FTT
P2: FLF
Archives of Sexual Behavior [asb]
PL111-340
June 17, 2000
13:46
Style file version Nov. 19th, 1999
378
Land´en and Innala Table I. Demographic Characteristics of People to Whom the Questionnaire Was Sent People who returned the questionnaire Males Characteristic Age (years) –19 20–29 30–39 40–49 50–59 60–69 70–79 Civil status Unmarried Married Divorced Widow/widower Income 0 1–84,999 85,000–159,999 160,000–234,999 235,000–309,999 310,000+ Total
People who did not return the questionnaire
Females
Males
Females
n
%
n
%
n
%
n
%
9 55 68 68 61 51 4
2.8 17.4 21.5 21.5 19.3 16.1 1.3
14 55 76 79 58 63 7
4.0 15.6 21.6 22.4 16.5 17.9 2.0
5 27 44 39 42 30 1
2.7 14.4 23.4 20.7 22.3 16.0 0.5
4 25 20 30 34 21 2
2.9 18.4 14.7 22.1 25 15.4 1.5
118 167 26 5
37.3 52.8 8.2 1.6
112 176 50 14
31.8 50 14.2 4.0
84 80 22 2
44.7 42.6 11.7 1.1
40 72 20 4
29.4 52.9 14.7 2.9
11 34 58 105 63 45
3.5 10.8 18.4 33.2 19.9 14.2
14 69 120 110 29 10
4.0 19.6 34.1 31.2 8.2 2.8
12 33 41 57 27 18
6.4 17.6 21.8 30.3 14.4 9.6
14 23 54 36 4 5
10.3 16.9 39.7 26.5 2.9 3.7
316
100.0
352
100.0
188
100.0
136
100.0
that way,” “You learn to be that way,” and “It is due to different experiences during childhood” formed the psychological explanation group (n = 191). The results, shown in Table III, indicate that those who believed that transsexualism is caused by biological factors had a less restrictive view on sex reassignment measures than people who held a psychological view. The biological group was also more prone to suggest publicly funded sex reassignment, to allow marriage in the new sex, and to allow transsexuals to adopt and work with children. The comparison between men and women is also displayed in Table III. Men expressed a more restrictive view than did women on the sex reassignment measures. Moreover, men held a more restrictive view on sex-reassigned persons’ marriage and on the questions dealing with the respondents’ potential personal relations with a transsexual. The median age in the study sample was 44 years. The older age group (>44 years, n = 330) was compared with the younger (≤44 years, n = 338). The results are shown in Table III. The older group was significantly more restrictive towards sex reassignment per se, whereas to a higher degree they reckoned that sex reassignment should be publicly funded than did the younger group. The two groups did not differ on the rest of the items.
P1: FNN/FTT
P2: FLF
Archives of Sexual Behavior [asb]
PL111-340
June 17, 2000
13:46
Style file version Nov. 19th, 1999
Attitudes Toward Transsexuals
379
Table II. Replies to a Questionnaire about Transsexualism, All Respondents (n = 668) 1. Do you consider transsexualism to be a disease that can be treated? 2. Are you of the opinion that transsexual person should have the opportunity to: change name change identity be treated with the sex hormones of the opposite sex undergo surgical operation of the genitals 3. Who should bear the expenses for a sex change? 4. Are you of the opinion that persons who have undergone a sex change should have the right to get married in their new sex? 5a. Are you of the opinion that persons who have undergone a sex change and are single should have the right to adopt and raise children on equal terms with other single people? 5b. Are you of the opinion that persons who have undergone a sex change and live together with a partner as a husband or wife should have the right to adopt and raise children on equal terms with other married people? 6a. Are you of the opinion that a person who have undergone a sex change from female to male should be allowed to work with children, e.g., be a teacher or youth worker? 6b. Are you of the opinion that a person who has undergone a sex change from male to female should be allowed to work with children, e.g., be a teacher, or youth worker? 7. Would it be possible for you to have an openly transsexual person as a fellow worker?
Yes No Have no opinion/Have not thought about it
22% (148)a 52% (345) 26% (175)
Yes No Have no opinion/Have not thought about it Yes No Have no opinion/Have not thought about it Yes No Have no opinion/Have not thought about it Yes No Have no opinion/Have not thought about it Public funds The individual Have no opinion/Have not thought about it Yes No Have no opinion/Have not thought about it
64% (427) 14% (94) 22% (147) 52% (349) 22% (144) 26% (175) 53% (357) 19% (127) 27% (183) 56% (377) 18% (122) 25% (169) 15% (100) 63% (420) 21% (139) 56% (377) 23% (154) 21% (137)
Yes No Have no opinion/Have not thought about it
29% (194) 52% (346) 19% (128)
Yes No Have no opinion/Have not thought about it
43% (287) 41% (273) 16% (108)
Yes No Have no opinion/Have not thought about it
61% (410) 20% (133) 19% (125)
Yes No Have no opinion/Have not thought about it
61% (409) 20% (135) 19% (124)
Yes No Have no opinion/Have not thought about it
71% (479) 11% (73) 18% (116) (Continued )
P1: FNN/FTT
P2: FLF
Archives of Sexual Behavior [asb]
PL111-340
June 17, 2000
13:46
380
Style file version Nov. 19th, 1999
Land´en and Innala Table II. (Continued )
8. Would it be possible for you to have an openly transsexual person as a friend? 9. Would it be possible for you to have an openly transsexual person as a partner? 10. Are you of the opinion that society and the media pay too much attention to transsexualism? 11. Do you know anyone who is transsexual? 12. Do you think the incidence of transsexualism has increased in Sweden in the last 20 years?
Yes 60% (402) No 18% (121) Have no opinion/Have not thought about it 22% (145) Yes 2% (13) No 84% (563) Have no opinion/Have not thought about it 14% (92) Too much attention Just enough attention Too little attention Have no opinion/Have not thought about it Yes No Yes No Have no opinion/Have not thought about it
13. What do you think it is that makes a person transsexual? (Choose one alternative)
a Absolute
17% (114) 30% (200) 12% (81) 41% (272) 8% (53) 91% (609) 38% (256) 23% (151) 39% (261)
You choose to be that way 9% (60) You learn to be that way 1% (6) You are born that way 53% (351) It is due to different experiences 19% (125) during childhood It is a disease that may affect you 4% (27) Other 11% (73)
numbers are shown in parentheses.
DISCUSSION We studied attitudes about transsexualism and transsexuals in a random national sample by means of mailing out a questionnaire to 992 Swedish citizens. The main finding was positive and tolerant attitudes toward transsexuals among lay people in Sweden. Subjects who believed that transsexualism is caused by biological factors held significantly less restrictive attitudes than subjects who believed in psychological explanations. The transsexual person’s right to undergo sex reassignment was supported by a majority of the respondents. In Sweden, a sex-reassigned person obtains all the rights and duties of the new sex, including the right to get married in the new sex and adopt and raise children. Whereas transsexuals’ right to marry in their new sex was supported by 56% of the respondents, the right of transsexuals to adopt and raise children seems to be a more controversial issue that was accepted by 43% and opposed by 41% of the respondents. This can be compared with a study of 318 psychology students, in which 51% of the females and 39% of the males were in favor of transsexuals being allowed to adopt a child (Leitenberg et al., 1983). It is known, however, that employment discrimination against transsexuals occurs (see, for example, Green, 1992). However, more than 70% of the respondents could accept having an openly transsexual person as a fellow worker. Furthermore, an overwhelming majority did not think transsexualism should prevent a person from working with children. This is compared with the study by Leitenberg
Yes No Have no opinion/Have not thought about it
undergo surgical operation of the genitals
64 10 19 68 9 17 24 56 18
41d 31d 23d 43d 30d 23d 3d 78d 19d
13 67 17
50b 24b 19b
47c 25c 20c
45a 26a 21a
16 59 23
62 13 20
60 14 22
59 18 19
69 11 17
13a 70a 15a
64d 20d 14d
60d 21d 16d
54c 25c 18c
72d 14d 13d
20 57 22
17 55 25 (Continued )
49 17 25
46 17 26
50 18 22
55 15 25
24 46 28
13:46
Public funds The individual Have no opinion/Have not thought about it
Yes No Have no opinion/Have not thought about it
be treated with the sex hormones of the opposite sex
63 15 17
42d 32d 21d
58 17 22
12 62 24
June 17, 2000
3. Who should bear the expenses for a sex change?
Yes No Have no opinion/Have not thought about it
74 9 14
54d 20d 23d
33d 40d 26d
PL111-340
change identity
15 60 24
24 49 27
Archives of Sexual Behavior [asb]
1. Do you consider Yes transsexualism to be a disease No that can be treated? Have no opinion/Have not thought about it 2. Are you of the opinion that a transsexual person should have the opportunity to: change name Yes No Have no opinion/Have not thought about it
Psychological Biological Below Above group, group, Males, Females, median age, median age, n = 191 n = 351 n = 316 n = 352 n = 338 n = 330
Table III. Group Comparisons
P1: FNN/FTT P2: FLF Style file version Nov. 19th, 1999
Attitudes Toward Transsexuals 381
34 46 20
52 32 15
72 14 14
23a 63a 14a
31d 56d 14d
50d 30d 20d
Yes No Have no opinion/Have not thought about it
Yes No Have no opinion/Have not thought about it
Yes No Have no opinion/Have not thought about it
50 33 17
66 15 18
d? 56a 26a 18a
31 48 21
66 15 19
382
60 19 20
42 39 19
26 52 22
53 22 25
13:46
63 20 16
44 43 12
32 52 15
59 25 16
June 17, 2000
35 50 15
27 56 16
46 32 21
PL111-340
5b. Are you of the opinion that persons who have undergone a sex change and live with a partner as a husband or wife should have the right to adopt and raise children on equal terms with other married people? 6a. Are you of the opinion that a person who has undergone a sex change from female to male should be allowed to work with children, e.g. be a teacher or youth worker?
68 15 17
44d 32d 24d
Yes No Have no opinion/Have not thought about it
Archives of Sexual Behavior [asb]
4. Are you of the opinion that persons who have undergone a sex change should have the right to get married in their new sex? 5a. Are you of the opinion that persons who have undergone a sex change and are single should have the right to adopt and raise children on equal terms with other single people?
Psychological Biological Below Above group, group, Males, Females, median age, median age, n = 191 n = 351 n = 316 n = 352 n = 338 n = 330
Table III. (Continued )
P1: FNN/FTT P2: FLF Style file version Nov. 19th, 1999
Land´en and Innala
0.5 90 9 21 34 9 36
Yes No Have no opinion/Have not thought about it Too much attention Just enough attention Too little attention Have no opinion/Have not thought about it
12 32 14 42
14 30 13 43
2 84 13
(Continued )
21 30 12 43
2 85 13
58 18 24
68 12 19
61 19 19
13:46
22 28 10 39
2 81 17
2a 88a 9a
63 18 18
75 10 15
62 21 16
June 17, 2000
14 33 15 38
68 11 22
80 5 15
66 15 18
52d 27d 20d
63d 17d 19d
55a 26a 18a
PL111-340
3 83 15
71 10 19
48d 27d 25d
Yes No Have no opinion/Have not thought about it
82 4 14
63d 17d 19d
8. Would it be possible for you to have an openly transsexual person as a friend? 9. Would it be possible for you to have an openly transsexual person as a partner? 10. Are you of the opinion that society and the media pay too much attention to transsexualism?
72 14 14
49d 31d 19d
Yes No Have no opinion/Have not thought about it
Yes No Have no opinion/Have not thought about it
Archives of Sexual Behavior [asb]
7. Would it be possible for you to have an openly transsexual person as a fellow worker?
6b. Are you of the opinion that a person who has undergone a sex change from male to female should be allowed to work with children, e.g., be a teacher, or youth worker?
P1: FNN/FTT P2: FLF Style file version Nov. 19th, 1999
Attitudes Toward Transsexuals 383
Yes No Have no opinion/Have not thought about it
12. Do you think the incidence of transsexualism has increased in Sweden in the last 20 years? 13. What do you think it is that makes a person transsexual? (Choose one alternative)
35 29 36 — — 100 — — —
50d 13d 37d — 31 3 — 65 — —
3 10
10 1 59 13
8c 1c 45c 25
3 10
10 1 49 22
40 24 36
8 91
5 9
8 1 56 16
37 22 41
8 91
13:46
6 9
36 24 40
9 91
41 21 37
7 91
June 17, 2000
384
All figures are percentages. The statistical significances shown in 1) the psychological group column refer to a comparison of the psychological group vs. the biological group; 2) the male group column refer to a comparison of the male group vs. the female group; 3) the below median age column refer to a comparison of the group below median age vs. the group above the median age. a p < .005. b p < .001. c p < .0005. d p < .0001.
10 89
4 96
PL111-340
You choose to be that way You learn to be that way You are born that way It is due to different experiences during childhood It is a disease that may affect you Other
Yes No
11. Do you know anyone who is transsexual?
Archives of Sexual Behavior [asb]
Psychological Biological Below Above group, group, Males, Females, median age, median age, n = 191 n = 351 n = 316 n = 352 n = 338 n = 330
Table III. (Continued )
P1: FNN/FTT P2: FLF Style file version Nov. 19th, 1999
Land´en and Innala
P1: FNN/FTT
P2: FLF
Archives of Sexual Behavior [asb]
PL111-340
Attitudes Toward Transsexuals
June 17, 2000
13:46
Style file version Nov. 19th, 1999
385
and Slavin (1983) in which 65% of females and 37.5% of males thought that a transsexual individual should be allowed to work as a teacher. That this is not always the actual case, however, is illustrated by several cases in the United States, where male-to-female transsexual teachers have been dismissed because of an expected negative impact on children. The outcomes of the court rulings in these cases have been different (Green, 1992; Hucker, 1985). Only 2% of the respondents thought that they could have a transsexual as a partner, which would implicate that the prospect of future marriage is poor. At odds with this, however, several outcome studies have found that half or more than half of the patients establish lasting relationships after sex reassignment (Eicher, 1992; Eldh et al., 1997; Pf¨afflin et al., 1990; W˚alinder et al., 1975). The group who believed that transsexualism is caused by biological factors represented more than 50% of the sample. The basis for the notion that beliefs concerning the origin of atypical sexual behavior may influence attitudes is a study of attitudes toward homosexuals (Ernulf et al., 1989). The authors found that subjects who believed that homosexuals are “born that way” held significantly more positive attitudes toward homosexuals than subjects who believed that homosexuals “choose to be that way” and/or “learn to be that way.” In the present study, the groups differed similarly concerning attitudes toward transsexuals. Taken together, these findings support the notion that a biological view on atypical sexual behavior is linked to a more tolerant attitude toward the people with these sexual variants. Of the respondents, 8% answered that they knew a transsexual. Given that there—at the time of the survey—were approximately 400 known transsexuals (Land´en et al., 1996a) and 8,860,000 inhabitants in Sweden (Official Statistics of Sweden, 1998), each transsexual would be known to approximately 1,772 people. This high figure could mean that the respondents interpret the question as if they have heard of any transsexual in their vicinity. In a small city for instance, a transsexual is so unusual that most people have heard of him/her. The figure may also—although it is less likely—reflect a hidden population of transsexualism (i.e., that the prevalence of transsexualism is higher than is reflected in the official figures). The question whether the incidence of transsexualism has increased since the early 1980s yielded a result that is not in accordance with current knowledge. Transsexualism is a rare phenomenon in Sweden. The annual incidence of sex reassignment in Sweden being in the interval of 0.14–0.17/100,000 inhabitants over 15 years of age (Land´en et al., 1996a). This estimate has remained stable since the early 1970s (Land´en et al., 1996b). Only 23% of the respondents suggested no increase of the incidence. The answers probably reflect the increased visibility of this hidden population over the years. Men held more restrictive views toward transsexualism and transsexuals than did women. This is in accordance with the results of Leitenberg and Slavin (1983). Few previous studies addressed the question of attitudes toward transsexualism and transsexuals. Although it is commonly believed that society has adopted
P1: FNN/FTT
P2: FLF
Archives of Sexual Behavior [asb]
386
PL111-340
June 17, 2000
13:46
Style file version Nov. 19th, 1999
Land´en and Innala
more tolerant and reconciled attitudes toward people with an atypical sexual identity, the literature on this subject is scarce. Green and coworkers studied attitudes toward sex reassignment among medical professionals and sexual minorities in 1966. Although that study, due to methodologic differences, is difficult to compare with the current survey, the responses as well as the questionnaire used by Green et al. reflect a more restrictive view on sex reassignment than we faced. Researchers that have investigated health-care professionals over time have found striking changes toward more liberal views (Franzini et al., 1986). In the present study, the younger group was more positive to sex reassignment than the older group, which may reflect a change in attitudes over time. However, transsexuals are still met with suspicion and hostility by some people (Rehman et al., 1999), but probably to a lesser degree than before. Attitudes toward people with an atypical sexual identity differ between cultures. In Burma, for example, transsexuals are sometimes respected in roles as shamans and seers (Coleman et al., 1992), whereas the condition in Western societies warrants a medical diagnosis (American Psychiatric Association, 1994; World Health Organization, 1993). The current public financing of sex reassignment in Sweden is a result of the medical establishment’s view of transsexualism as a disease. In contrast, however, a majority of respondants in our survey did not regard transsexualism as a disease. This may be one reason why they did not agree with the current manner of public funding and thought that the individual him-/herself should bear the expenses for sex reassignment. In addition, many people may regard sex reassignment as a cosmetic treatment, and the results may thus reflect the current debate regarding whether such treatment should be funded publicly. Legislation influences moral values in a society and contributes to more positive attitudes (Monteith, 1993). Thus, the relatively positive view on transsexualism and transsexuals reflected in this study might partly be explained by the legislation in Sweden. Since 1972, the sex-reassigned person is granted the rights and obligations of the new sex (W˚alinder et al., 1976). Several other countries have since followed and legislated on this issue (Petersen et al., 1995; Weitze et al., 1996). There is, however, reason to believe that the significant import of these laws is unknown to the majority of the population. Like all questionnaire surveys, we faced the problem of nonrespondents. However, in this study the overall response rate was 67%, which is higher than in previous studies of sexual behavior in Britain and some other countries (Kupek, 1998). A recent interview study of sexual behavior in Sweden 1996 had a response rate of 59% (Lewin, 1996). Although somewhat more women than men returned the questionnaire, the remaining demographic characteristics of the respondents and the dropouts were similar (Table I). The results can thus be expected to be reasonably representative for the population as a whole. It is possible, however, that people who do not want to discuss atypical behavior in any context are overrepresented in the nonrespondent group. In that case, the survey has probably yielded a less restrictive view than is the actual case.
P1: FNN/FTT
P2: FLF
Archives of Sexual Behavior [asb]
PL111-340
June 17, 2000
13:46
Attitudes Toward Transsexuals
Style file version Nov. 19th, 1999
387
In conclusion, this study found positive and tolerant attitudes toward transsexuals among lay people in Sweden. Subjects who believed that transsexualism is caused by biological factors, women, and the younger age group held significantly less restrictive attitudes than subjects who believed in psychological explanations, men, and the older age group. Future studies should address these questions to elucidate differences in attitudes between cultures and the process of change over time.
ACKNOWLEDGMENTS The authors are indebted to Drs. Bengt Lundstr¨om and Rolf Persson for valuable comments on the questionnaire. This study was supported by grants from the Swedish Medical Research Council (Grant No. 8668), Stiftelsen Torsten Amundsens Fond, Svenska Lundbeckstiftelsen, and the G¨oteborg Medical Society.
REFERENCES Altman, D. G. (1991). Practical Statistics for Medical Research, Chapman & Hall, London. American Psychiatric Association (1994). Diagnostic and Statistical Manual of Mental Disorders, APA, Washington, DC. Cohen-Kettenis, P. T., and Gooren, L. J. G. (1999). Transsexualism: A review of etiology, diagnosis and treatment. J. Psychosom. Res. 46: 315–333. Coleman, E., Colgan, P., and Gooren, L. (1992). Male cross-gender behavior in Myanmar (Burma): A description of the acault. Arch. Sex. Behav. 21: 313–321. Eicher, W. (1992). Transsexualismus: M¨oglichkeiten und Grenzen der Geschlechtsumwandlung, G. fischer Verlag, Stuttgart, Jena, New York. Eldh, J., Berg, A., and Gustafsson, M. (1997). Long-term follow up after sex reassignment surgery. Scand. J. Plast. Reconstr. Surg. Hand Surg. 31: 39–45. Ernulf, K. E., Innala, S. M., and Whitam, F. L. (1989). Biological explanation, psychological explanation, and tolerance of homosexuals: A cross-national analysis of beliefs and attitudes. Psychol. Rep. 65: 1003–1010. Franzini, L. R., and Casinelli, D. L. (1986). Health professionals’ factual knowledge and changing attitudes toward transsexuals. Soc. Sci. Med. 22: 535–539. Green, R. (1992). Sexual Science and the Law, Harvard University Press, Cambridge, MA. Green, R., Stoller, R., J., and MacAndrew, C. (1966). Attitudes toward sex transformation procedures. Arch. Gen. Psych. 15: 178–182. Hucker, J. S. (1985), Medical–legal issues, In Steiner, B. W. (ed.), Gender Dysphoria, Plenum Press, New York. Kupek, E. (1998). Determinants of item nonresponse in a large national sex survey. Arch. Sex. Behav. 27: 581–94. Land´en, M., W˚alinder, J., and Lundstr¨om, B. (1996a). Incidence and sex ratio of transsexualism in Sweden. Acta Psych. Scand. 93: 261–263. Land´en, M., W˚alinder, J., and Lundstr¨om, B. (1996b). Prevalence, incidence and sex ratio of transsexualism. Acta Psych. Scand. 93: 221–223. Leitenberg, H., and Slavin, L. (1983). Comparison of attitudes toward transsexuality and homosexuality. Arch. Sex. Behav. 12: 337–346. Lewin, B. (1996). Sex i Sverige [Sex in Sweden], Folkh¨alsoinstitutet, Stockholm, Sweden. Monteith, M. J. (1993). Self-regulation of prejudiced responses: Implications for progress in prejudicereduction efforts. J. Pers. Soc. Psychol. 65: 469–485.
P1: FNN/FTT
P2: FLF
Archives of Sexual Behavior [asb]
388
PL111-340
June 17, 2000
13:46
Style file version Nov. 19th, 1999
Land´en and Innala
Official Statistics of Sweden (1998). Statistical Yearbook of Sweden 1999, Statistics Sweden, Stockholm. Petersen, M. E., and Dickey, R. (1995). Surgical sex reassignment: A comparative survey of international centers. Arch. Sex. Behav. 24: 135–156. Pf¨afflin, F., and Junge, A. (1990). Nachuntersuchung von 85 operierten Transsexuellen. Z. Sexualforsch. 3: 331–348. Rehman, J., Lazer, S., Benet, A. E., Schaefer, L. C., and Melman, A. (1999). The reported sex and surgery satisfactions of 28 postoperative male-to-female transsexual patients. Arch. Sex. Behav. 28: 71–89. Weitze, C., and Osburg, S. (1996). Transsexualism in Germany: Empirical data on epidemiology and application of the German Transsexuals’ Act during its first ten years. Arch. Sex. Behav. 25: 409–425. World Health Organization (1993). The ICD-10 Classification of Mental and Behavioural Disorders. Diagnostic Criteria for Research, WHO, Geneva. W˚alinder, J., and Thuwe, I. (1975). A Social–Psychiatric Follow-up Study of 24 Sex-Reassigned Transsexuals, Scandinavian University Books, G¨oteborg, Sweden. W˚alinder, J., and Thuwe, I. (1976). A law concerning sex reassignment of transsexuals in Sweden. Arch. Sex. Behav. 5: 255–258.
P1: FVK Archives of Sexual Behavior [asb]
PL111-341
June 29, 2000
11:39
Style file version Nov. 19th, 1999
Archives of Sexual Behavior, Vol. 29, No. 4, 2000
The Field of Sex Research: Responsibility to Ourselves and to Society Milton Diamond, Ph.D.1
I would like to address some ideas in general which I think are of importance to all of us interested in sex research. These ideas are not new nor thought up just for the occasion. I’ve thought about them for some time and personally tried to put some of them into action. These ideas are not unique to me, but I believe worth emphasis. Academy presidents in the past have voiced ideas somewhat similar in nature (e.g., Tiefer, 1994). The first and most important point I would like to emphasize is the need to establish sexology firmly and unequivocally as a fully fledged field of study. This is not to be a branch of biology or medicine or sociology or psychology or any other discipline. Yes, professionals of those areas may have a professional interest in sex, but those trained and training in sexology will be individuals aware and knowledgeable of the relevant factors in each of those other fields so that they can integrate, evaluate, produce, and otherwise further the overall subject. My own graduate studies in anatomy had me learn histology, physiology, embryology, endocrinology, biochemistry, and some electron microscopy, as well as neuroscience, statistics, and more. But it still provided a limited view of sex. I also did a graduate minor in psychology to augment my knowledge from that perspective and took medical training as well and passed basic boards in that discipline. Even that was not enough, and over the years I’ve tried to add knowledge of sociology, anthropology, and other fields. I don’t think that was the best way to approach our discipline. It would have been better to get all the needed courses in one department. I see a sexology department doing that. Subjects in a sexology department can and should range from clinical to philosophical, from ethics to religion, literature, and history, from psychology to sociology and anthropology, from art to pornography, prostitution, sexually transmitted diseases and public health matters, to medical conditions, to women’s 1 International Academy of Sex Research, Presidential Address, 26 June 1999, State University of New
York at Stony Brook. 389 C 2000 Plenum Publishing Corporation 0004-0002/00/0800-0389$18.00/0 °
P1: FVK Archives of Sexual Behavior [asb]
390
PL111-341
June 29, 2000
11:39
Style file version Nov. 19th, 1999
Diamond
and men’s studies. The potential scope is perhaps greater than that for most existing departments and the talents needed to support such a program are diverse. In the present case, I’m calling for those of you in each of your disciplines, particularly the younger members with your careers still ahead of you, to come together for the benefit of yourselves and future generations. I am not simply speaking of holding a joint course of Sexology 101 where contributions come from different departments. I’m calling for a full establishment of departments of Sexology to offer a set of courses which, after Sexology 101 and 102, can carry the student into any of 10 or so directions for specialization up to the Ph.D. degree. I’m asking that those who come together in such a department do not do so just in token affiliation, but for true collaboration, mental enrichment, and support in research, teaching, and other directions; for your own professional and intellectual growth. There are universities in the U.S. which offer many courses, but not yet a single department in any fully accredited university where an individual can pursue a doctorate in one of a dozen or more aspects of sexology. It can be anyone of you to take the lead in your own institution to bring the relevant parties together. I understand this is presently a goal at the Kinsey Institute at the University of Indiana, and with programs at the University of California at Northridge and the University of California at San Francisco. I truly hope these programs succeed. These and additional ones are needed. There is little doubt that the students enrolling in such courses will benefit personally in immeasurably more way than they might from other courses they take. I’m further asking that those of us here, for certain, and those who would join or form such a department, to proudly identify as sexologists even while remaining in your present departments. One can be the sexology expert in the psychology department, or a sociology department, or at a medical school, and so on. Colleagues will be infected by the enthusiasm in which you embrace the calling and realize that studies of all aspects of sex are legitimate, and a professional interest in sex is not prurient but as worthy as any other academic, scientific, or clinical interest. But more importantly, it will establish in the minds of the administration, students, fellow faculty, and eventually the community at large that the study of sex—its teaching, research and other endeavors—are as fully legitimate as any other discipline listed in the institution’s catalogue. It will foster the recognition that “academic freedom” in research and teaching is not only for orthodoxy, but for ideas that perhaps challenge the status quo as well. Parenthetically, I would add that calling sexology sexual science, to me, is a pretentious attempt at upgrading the field. It is a testimony to insecurity. We don’t need terms like psychological science or sociological science. Psychology and sociology do nicely, and so does sexology. Yes, continue to attend psychology or sociology meetings if your area of interest is such, or neurobiology meetings, or history society conferences or medical
P1: FVK Archives of Sexual Behavior [asb]
PL111-341
The Field of Sex Research
June 29, 2000
11:39
Style file version Nov. 19th, 1999
391
clinical symposia or women’s study programs if they are to your taste. But bring the information and expertise to bear within the broader discipline of sexology so that the challenges, even debates and arguments of colleagues that see the world differently, can be put to the strongest tests. Do not avoid interdisciplinary challenges—seek them. Discussion with those with whom you intellectually disagree will provide the best way to perfect your own thinking and theory construction. It would make whatever evolves that much stronger. I’ve always told my students, “Don’t aim at straw men or weak arguments to disprove, but find the strongest argument of others and strive to better them.” Don’t do only what is easy; do what is difficult. It’s probably not been done before. When the HIV/AIDS epidemic hit the world, there should have been a clamor for sexologists to provide answers or directions of attack to the problem just as there was a call on physicians and public health professionals. Yes, those professionals were and are needed in the fight. But, as it turns out still today, the best medicine for this catastrophic disease is prevention and sex education, not a pill nor a potion nor a form to fill out. It is sexologists drawn from all the fields mentioned above that must, until a vaccine is available, integrate the behavioral–social–cultural factors which might eventually penetrate the individual and collective consciousness to slow down the spread of this dread disease with appropriate sex education. If pure water were the antidote for HIV infection, most of those with the disease could not afford it, and for those in risk of the disease, any vaccine’s cost would keep it out of reach. The formation of our discipline into a frontline one will not be easy. Aside from the institutional arguments of budget and administrative red tape, there will be fights over territory, course content, and negative publicity. There will be scrutiny, which few other fields have. And there might be ridicule. But there can also be support for dynamic teaching ideas or research efforts. There will certainly be help to students in helping them understand some important aspects of their lives. There can be a legitimization for providing input to administrative and social matters, which would otherwise not seem appropriate both on and off campus. The second point, which goes along with the first and in many way is supported by it, is to, as individual professionals, proudly evolve ourselves in public discourse and even the political process. The research findings of many of use are useful, and in many ways, crucial in helping matters of everyday life. Unfortunately, publication in Archives of Sexual Behavior, Science, or even the New England Journal of Medicine doesn’t easily get translated into public thinking or action. What is at least needed is calling relevant findings to public awareness. This can be done via institutional public relations offices or directly to the media. This can be by letter writing or offering public testimony when appropriate. It is certainly of use to do so in organizational advising and such. Most of us are relatively humble and conservative in this regard. I think to hide our lights under a bushel is selfdefeating and of less value to the greater social good than it might be. Moreover,
P1: FVK Archives of Sexual Behavior [asb]
392
PL111-341
June 29, 2000
11:39
Style file version Nov. 19th, 1999
Diamond
the improvement of society is one justification for research. Brian Gladue wrote of the American Psychological Association and some of its roles and the push to initiate a Division of Sexology within that organization (Gladue, 1977). I strongly support his efforts and the idea that a Division of Sexology be formed within the American Psychological Association (APA) and in many other organizations as well; say the American Medical Association. While we are all super busy, our results might be important for civil rights, say for gays, bisexuals, transgendered or intersexed individuals; for those who might be called to task for displays of dress, images, or behavior considered “inappropriate.” Our work can be useful in the fight for the right to control one’s reproductive potential and to have children by choice. Our work can advise on matters of pornography, rape, the treatment of sex offenders or sex abuse victims, and on the structure of sex-related laws. Our work can be instrumental in increasing the public’s understanding and dealing with questions of prostitution, individual sexual differences, relationships, love, marriage, child rearing, censorship, and more. In short, the topics we investigate every day have wide-ranging importance to society. I think we need to have society come to see our value as sexologists, and that is done by ourselves first demonstrating our own pride in what we do and by our willingness to present our findings for public consumption. Do I think that all of us would agree with the solutions we might offer for societies’ sexual concerns? Or do I think we would all agree on how to approach socially sensitive and important sex-related issues? Of course not. But I think our open discourse and debate over the issues, with cogent critique and analysis, would be better than gut feelings and recourse to tradition or religion. And the final result will be instructive to the general public. I don’t think we ought to surrender the discussion or sound bites to those whose motives are sexophobic and whose knowledge of sex is minimal. The third matter I would like to stress is in regard to our organization within itself, our meetings and functioning and our relationship to the communities in which we exist. One of the precipitating factors in this introspection was the formation of a new organization spawned by one of our members, Gil Herdt. This new society is called the “International Association for the Study of Sexuality, Culture and Society.” This organization has its own journal—Culture, Health and Sexuality. While I wish it luck, it is obvious that its focus seems diametrically opposed to what we in the academy hope to achieve and be part of. One of our strongest draws to our organization for me is that we strive to integrate rather than separate; we want to encourage the meeting of minds, not mentally separate “Us” from “Them.” For me, research is somewhat like intellectual solitaire, where the game is enhanced by difficulty, not by “cheating” by using only easy card sorts. It is na¨ıve of any society to think it can understand sexual culture and society without considering how biology, medical conditions, or evolution might be influential. In
P1: FVK Archives of Sexual Behavior [asb]
PL111-341
The Field of Sex Research
June 29, 2000
11:39
Style file version Nov. 19th, 1999
393
so doing, it throws out the baby with the bathwater. Just to give a personal example, I recently spent 6 months trying to better my understanding of sexual development by researching intersexuality in Gaza, Israel, where, due to intrafamilial marriage and genetic inheritance, there is a large intersex population (see, e.g., Rosler and Kohn, 1983). There is no way the intersex condition can be looked at and comprehended without fully considering both the biological and sociocultural situations in which these individuals are raised and live. And this microculture is only understood by recognizing some of the biological factors with which these people have to deal and those which shape their lives. I fear that the new organization will suffer for its narrow focus and am thankful that we are more cosmopolitan and broader in thinking. What would be regrettable from the point of view of our own organization is if it splinters off some of our current members so they do not again participate in the cross-fertilization of ideas, comments, and cogent critiques here and receive the views of others so all mutually benefit. We are now celebrating our 25th anniversary, and that concept of multidisciplinary representation has been one of our academy’s founding principles. I think the cross-fertilization and interactions offered here is one of the organization’s strengths. Biological ideas are leavened by those of sociologists, and vice versa; the clinical approaches of psychologists are augmented by those of medical clinicians, and vice versa; the concerns of philosophers and ethicists are placed in juxtaposition with reality as seen by behaviorists, zoologists, or lawyers. In short, our meetings and organization provides a venue, not where we preach to the choir, but where we discuss, dispute, and even argue with peers of different expertise; sexology is not a religion where everything is settled by dictum or holy writ. It is a spirited field where I see challenge and debate as positive as is the combining of our brain power to achieve broader understanding. That said, how can we improve as an organization? How can we better offer our members what it is they want or need? I offer some of my insights and preferences for consideration. The first set of ideas is in regard to our meetings, the second is in regard to efforts aside from the meetings. Members: One simple suggestion is to invite those whose work you respect or often quote to become members of the society. They do not have to be individuals you know personally. All you have to know is that they have a serious interest in sex research, have demonstrated so with publication, and you think they have something to offer. Meetings: I recommend that symposia be structured so that different views are offered to cover any topic. A presentation of homosexuality, for instance, can include a talk on how atypical orientation effects society, and another talk can discuss ethical or philosophical issues related to such studies, while a third offers a review of what is know of the genetic and neural components to sexual partner choice.
P1: FVK Archives of Sexual Behavior [asb]
394
PL111-341
June 29, 2000
11:39
Style file version Nov. 19th, 1999
Diamond
Posters: I suggest that our meetings display any poster any member cares to present. Membership in our society, by itself, testifies to the caliber of the member so that any idea he or she wishes to present deserves consideration and viewing, acceptance, debate, or rejections by other members. And time and place must be given for full and honorable attention to these posters so that they are considered a valuable academic presentation on a par with oral presentation. Another suggestion is to have a “Hot Seat.” This is where someone volunteers to present a controversial thesis or some findings for challenge or to ask for help on attacking a research problem. It might take the form of a 5-minute presentation followed by 10–15 minutes of exchange with the audience. For instance, someone may have a controversial theory or set of results to be explored in depth. These can be philosophical or data oriented (e.g., why are men more often fetishists than women? Is religion good or bad for sexual activities? Does the “G-spot” and female ejaculation really exist?) The volunteer, of course, can be invited. The Rest of the Year: With Mike Bailey’s agreement, as keeper of the listserve, I would encourage and welcome all IASR members and other sex researchers to participate or at least lurk on Sexnet
[email protected] to offer whatever input is felt worthy, to contribute so we can learn from the discussions. A more difficult suggestion: We must find a way to deal with the scurrilous and ridiculous remarks of individuals like Judith Reisman and the ranting of Dr. Laura. Their comments smear not only the practices they find offensive, but the individuals as well. The pro and con actions of the American Psychological Association has also made the news. I think it is our responsibility as a field and surely as a society to defend our colleagues’ right to academic freedom, and, after having passed peer review, to protect their integrity and their work and to expose to the media and elsewhere, those whose thinking on the matter is unschooled or wrong. Doing so as a professional society has weight that a critique by only one or several of us would not have. And I think we could have voiced our opinions about the firing of Journal of American Medical Association editor George Lundberg for publication of sex survey results and at so many other times when a rational voice would be of value to counter all the misinformation the public hears that colors their views of what is right or correct or even true or false about sex. I think we, as a society, should defend academic freedom for those who present unpopular but research- or theory-defensible ideas, and write of child–adult sexual activity or other issues. We need, as an organization, to defend the Kinsey Institute, the person Kinsey, and all serious researchers. The answer to charges of bad research and thinking is a call for better research and better thinking. We ought to defend academic freedom as an organization. For surely, if we don’t hang together, we will hang separately. I would like to see the academy offer itself for amicus briefs or position papers, either directly or via consultation with appropriate legal firms or organizations, for many sex-related matters. In short, I think we need to venture outside of our ivory
P1: FVK Archives of Sexual Behavior [asb]
PL111-341
June 29, 2000
11:39
The Field of Sex Research
Style file version Nov. 19th, 1999
395
towers, not to make money, but to better society and sexology as a field. We must make other professional organizations and the media know we exist and have intellectual weight. The mechanisms for how to do this can be worked out. For the presidential talk, I’ve tried to go beyond the usual recitation of some research efforts of my past and talk of my vision for the organization’s present and our field’s future. Thank you all for allowing me the opportunity to do so. The past 25 years of the academy is filled with prideful years. I trust that much before the International Academy of Sex Research comes together for its 50th anniversary, we can look back on many of my suggestions as having been long accomplished, and look forward to ever greater achievements for the field of sexology.
REFERENCES Gladue, B. (1997). Update on Planning Process. 18 July 1997, 06:39:35—1000; gladueba@email. uc.edu Rosler, A., and Kohn, G. (1983). Male pseudohermaphroditism due to 17B-hydroxysteroid dehydrogenase deficiency: Studies on the natural history of the defect and effect of androgens on gender role. J. Ster. Biochem. 19: 663–674. Tiefer, L. (1994). Three crises facing sexology. Arch. Sex. Behav. 23: 361–374.
P1: FNN Archives of Sexual Behavior [asb]
PL111-342a,b,c,d
June 24, 2000
10:36
Style file version Nov. 19th, 1999
Archives of Sexual Behavior, Vol. 29, No. 4, 2000
BOOK REVIEWS Second Skins: The Body Narratives of Transsexuality. By Jay Prosser. Columbia University Press, New York, 1998, 270 pp., $21.95. Reviewed by Patricia Elliot, Ph.D.1
I have been interested in transsexuals not as “exotic” but as persons who suffer for both personal and political reasons. Following Phillips (1995), I would say that the importance of psychoanalysis lies in its capacity to provide a language “for what matters most to us; for what we suffer from and for, for how and why we take our pleasures” (p. xvi). Prosser’s analysis of transsexuality, defined by the activity of “entering into a lengthy, formalized, and normally substantive transition: a correlated set of corporeal, psychic, and social changes” (p. 4), provides considerable insight into a specific process of suffering and transformation. Drawing on transsexual autobiographies as well as his own experience as a FTM (female-to-male transsexual), Prosser produces a psychoanalytically inspired analysis of processes that are part of his own history. Moreover, in daring to address concepts of fantasy, yearning, and loss, Prosser offers us a psychological account of unusual experiences that are rarely presented in purely descriptive terms. I believe that Prosser’s study offers the most sophisticated theoretical analysis of at least one form of transsexuality to date, even if, as I suggest, he tends to underestimate the role of the subject in the process of embodiment. Prosser argues that transsexuals have a different experience of embodiment than nontranssexuals, and a particular response to that experience that distinguishes them, for the most part, from other transgender persons. Indeed, he urges that new “paradigms for writing bodily subjects” (p. 12) be developed from the analysis of transsexual narratives, paradigms that would address the question of the body’s materiality. Making this argument leads Prosser to a series of debates with queer and transgender theory on the one hand, and to a detailed, psychoanalytically informed theorization of transsexual experience on the other. Taking issue with the tendency of some queer and transgender theorists, such as Butler (1993) and Stone 1 Department
of Sociology and Anthropology, Wilfrid Laurier University, Waterloo, Ontario N2L 3C5
Canada. 397 C 2000 Plenum Publishing Corporation 0004-0002/00/0800-0397$18.00/0 °
P1: FNN Archives of Sexual Behavior [asb]
398
PL111-342a,b,c,d
June 24, 2000
10:36
Style file version Nov. 19th, 1999
Book Reviews
(1991), to read transsexuality as either a “bad literalization” of gender or a “good deliteralization” of gender, Prosser argues that these readings inevitably eclipse the specificity of transsexual experience. And contrary to queer theory’s attempts to champion the transsexual as revealing a shift “away from the embodiment of sexual difference” (through blurring boundaries, revealing that “sex” is immaterial or really gender all along), he argues that transsexual transition shifts the subject “more fully into it” (p. 6). Prosser’s description of transsexual transition as motivated by “an initial absence of and striving for” the feeling of embodiment (p. 7), and as “the transformation of an unlivable shattered body into a livable whole” (p. 92), prompts him to turn to psychoanalysis in order to deepen his understanding of these processes. Although Prosser’s engagement with queer theory and with psychoanalytic theory both constitute valuable contributions to the larger field of transsexual and transgender research, Prosser’s engagement with psychoanalytic theory will be discussed in more detail here. For Prosser, the importance of transsexual narrative is to emphasize the necessity of (sexed) embodiment to the subject, where “embodiment is as much about feeling one inhabits material flesh as the flesh itself” (p. 7). In fact, the claim that “owning one’s skin” is necessary for any subject’s well-being is the first assumption in Prosser’s argument. The second is that some transsexuals, whose body image contrasts with their visible body, are unable to “own” their material bodies. The third is that, faced with what for them is an unlivable conflict, these transsexuals locate inappropriateness in the material body, not the body image. The logical conclusion is that they seek to alter the flesh (not the body image) as a solution to the human necessity to feel embodied. A few more points suffice to complete this very brief outline of his theory. Read through French psychoanalyst Anzieu, Prosser argues Freud’s point that the body, as an image of self in which we invest, is not synonymous with the material body or organism. Moreover, he argues that “body image can feel sufficiently substantial as to persuade the transsexual to alter his or her body to conform to it. The image of wrong embodiment describes most effectively the experience of pretransition (dis)embodiment: the feeling of a sexed body dysphoria profoundly and subjectively experienced” (p. 69). In this description, it is the body image that “has a material force, not the physical body itself” (p. 69). According to Anzieu (1989), the skin ego is an embodied entity, not just an image; it is “an organ enabling and illustrating the psychic/corporeal interchange of subjectivity” (p. 72). As Prosser explains it, the skin, “as the point of contact between material body and body image, between visible and felt matter” (p. 72), becomes a major site of conflict for the transsexual. If “the skin is the locale for the physical experience of body image and the surface upon which is projected the psychic representation of the body,” then one can only feel “at home in one’s skin” (p. 72) if material body and body image correspond. Finally, it is important to understand Prosser’s view of surgery and its effects as both creating a “feeling of bodily integrity” (p. 79) and rectifying specific “body
P1: FNN Archives of Sexual Behavior [asb]
Book Reviews
PL111-342a,b,c,d
June 24, 2000
10:36
Style file version Nov. 19th, 1999
399
image distortions” (p. 85). Surgery is said to be “a fantasy of restoring the body to the self” (p. 82) and a “drive to get the body back to what should have been” (p. 83). Like some forms of cosmetic surgery, Prosser explains, in which a body part that has been disowned is reconstructed, there is a beneficial or healing effect. Prosser refers to the work of Hans Sacks to describe two body-image distortions that also accompany transsexual experience: (1) body agnosia, or the forgetting of specific body parts, and (2) phantomization, or the ability to imagine body parts that have been “lost” and that should have been there: “In the case of the transsexual the body constructed through sex reassignment surgery is not one that actually existed in the past, one that is literally re-membered, but one that should have existed; sex reassignment surgery is a recovery of what was not” (p. 84). My critical engagement with this text begins with an appreciation of Prosser’s powerful and insightful reading of this experience and is intended as a respectful tribute to it. Prosser provides us with a coherent narrative and a detailed description of at least one form of transsexual experience. I suggest, however, that more careful attention to the analysis of “feelings” and to the psychoanalytic emphasis on the subject of the unconscious could enhance an already fruitful discussion. Prosser borrows Freud’s concept of the body ego in order “to materialize the psyche, to argue its corporeal dependence” (p. 42). The body image has a material reality that is said both to produce the image and to be produced by it: sex is always invested in the flesh. But here Prosser seems to conflate the feeling of one’s sex with the flesh itself. For example, he argues that the key to transsexual narrative is to see “how the material flesh may resist its cultural inscription” (p. 7). But to argue that feelings about one’s embodiment are synonymous with the flesh is to lose the concept of psychic reality and of the subject. Processes of representation, the idiosyncratic creation of meaning, and the specific dynamics by which a relationship to an (internalized) “Other” is forged are also lost to Prosser’s analysis. Given that for him a relationship seems to exist between the subject’s sense of gender identity and their “feelings of the body,” it is difficult to understand why he would reduce that relationship to bodily feelings alone. We do not have to leap to the opposite conclusion that bodies (or subjects, for that matter) are merely discursive effects in order to entertain the idea that our psychic investments in the body have a relationship to what is physically experienced. For Freud (1984), the ego may be “first and foremost a body ego” (p. 364), but it is also an agency that strives to create a life for the subject between the demands of the social world, the demands of the internalized Other, and its own demand for love. Prosser’s point that transsexuals demonstrate the inadequacy of theories of social determinism, in which the body is written on by the social, is well taken. However, this critique comes as no surprise to psychoanalytic theorists for whom the body is reducible neither to biology nor to the social. For the psychoanalytic theorist, however, it is not the material flesh that resists cultural inscription but the subject who resists. For Shepherdson (1994), this is because the subject is not a socially determined thing; for Rose (1986), it is because “there is a resistance
P1: FNN Archives of Sexual Behavior [asb]
PL111-342a,b,c,d
June 24, 2000
400
10:36
Style file version Nov. 19th, 1999
Book Reviews
to identity at the very heart of psychic life” (p. 91) so that the subject never fits neatly into socially prescribed identities; for Copjec (1990), it is because there is no subject without conflict; and, for Freud, it is because there is such a thing as psychic reality. I am suggesting that psychoanalysis offers additional contributions, three of which I describe, that could be used to enhance Prosser’s theorization of transsexual embodiment. First, rather than seeing conflict as a result of bodily resistance or a function of simple human need, as Prosser does, we could use psychoanalytic distinctions among need, demand, and desire to sort out different kinds of conflict and examine how those conflicts locate themselves in specific bodily zones and functions. Second, to avoid reducing the complexity of the body to either a biological given or a cultural product, it is useful to keep in mind the psychoanalytic emphasis on the corporeal satisfaction of the drive and the symbolic nature of the symptom (where memory is written on the flesh). Third, Freud taught us to read in the body the effects of unconscious investments made as a result of our inevitable dependence on and construction of the (internalized) Other. Far from reducing the body to “someone else’s idea” or rendering it unknowable, as Prosser fears, psychoanalysis offers us a way to make sense of what we have become both subject of and subject to. Obviously, I find Prosser’s description of transsexual embodiment compelling. It both confirms other autobiographical descriptions I’ve read and raises useful questions for contemporary queer and transgender theory. Moreover, Prosser’s use of psychoanalysis to theorize bodies that, like all bodies, are never simply given by nature or by culture, is a promising way to pursue the complexity he describes. When Prosser writes about the transsexual “body narrative,” the “skin ego,” and the “recalcitrance of the flesh,” however, it seems as though the subject gets left out. But when he describes the disembodied subject who fantasizes a return to “what should have been”; who talks about yearning, loss, and even desire, then the subject reappears. A more detailed psychoanalytic account of these unconscious aspects of the transsexual subject’s relationship to embodiment would enrich our understanding of the specific experience Prosser has done so much to clarify. REFERENCES Anzieu, D. (1989). The Skin Ego (Trans., C. Turner), Yale University Press, New Haven, CT. Butler, J. (1993). Bodies that Matter: On the Discursive Limits of “Sex,” Routledge, New York. Copjec, J. (1990). m/f or not reconciled. In P. Adams and E. Cowie (eds.), The Woman in Question, MIT Press, Cambridge, MA, pp. 10–18. Freud, S. (1984). The Ego and the Id (ed., A. Richards), Penguin Books, Harmondsworth, England. Phillips, A. (1995). Terrors and Experts, Harvard University Press, Cambridge, MA. Rose, J. (1986). Sexuality in the Field of Vision, Verso, London. Shepherdson, C. (1994). The role of gender and the imperative of sex. In J. Copjec (ed.), Supposing the Subject, Verso, London, pp. 158–184. Stone, S. (1991). The “empire” strikes back: A posttranssexual manifesto. In J. Epstein and K. Straub (eds.), Body Guards: The Cultural Politics of Gender Ambiguity, Routledge, New York, pp. 280– 304.
P1: FNN Archives of Sexual Behavior [asb]
PL111-342a,b,c,d
Book Reviews
June 24, 2000
10:36
Style file version Nov. 19th, 1999
401
Pedophiles and Priests: Anatomy of a Contemporary Crisis. By Philip Jenkins. Oxford University Press, New York, 1996, 214 pp., $36.95 (Canadian). Reviewed by Jay R. Feierman, M.D.2
Starting in the mid-1980s, The National Catholic Reporter, a liberal Catholic newspaper, began publishing articles about “pedophile priests” in North America, culminating in reporter Jason Berry’s (1992) influential book, Lead Us Not Into Temptation: Catholic Priests and the Sexual Abuse of Children. In Andrew M. Greeley’s foreword, he describes its content as revealing “what may be the greatest scandal in the history of religion in America and perhaps the most serious crisis Catholicism has faced since the Reformation” (p. xiii). Berry reports that in the decade after 1982, about 400 Catholic clergy in North America were accused of sexual misconduct with minors, part of what he saw as a larger issue—the homosexual subversion of the Roman Catholic Church—with estimates that up to half of all Catholic priests in North America may be homosexual. Berry’s book ends with an accusatory indictment of the church hierarchy, which he says has allowed the erosion of seminary formation and gay behavior patterns in rectories and religious communities—all in the name of celibacy. Berry describes mandatory celibacy, which has been in effect since 1139, as a political and theological model that has failed and whose continuity is propagated by an ecclesiastic, pedophile-harboring power structure that rules the church out of fear of loss of their own power. With the above as background, enter the volume under review, in which Berry’s book is cited 39 times in endnotes. Jenkins claims that, “In reality, Catholic clergy are not necessarily represented in the sexual abuse phenomenon at a rate higher than or equal to their numbers in the clerical profession as a whole” (p. 8). Both Berry and Jenkins acknowledge that almost all of the reported cases involve priests and pubescent boys, which technically is called either ephebophilia or pederasty, rather than pedophilia. Jenkins estimates that between 1960 and 1992, approximately 150,000 Roman Catholic clergy served in North America. Only a handful of Catholic priests were publicly accused of sexual impropriety with minors in the period prior to 1982. Of the approximately 400 that were publicized after 1982, many involved behaviors that had occurred decades earlier. Jenkins argues that 400 priests accused out of 150,000 priests served represents only a fraction of a percent, hardly enough to qualify as the greatest scandal in the history of religion in America and the greatest crisis for Roman Catholicism since the Protestant Reformation. Jenkins argues that the pedophile priest crisis was “framed” by the already existing women’s movement that began in the 1970s and the growing awareness of child sexual abuse in the 1980s. He argues that in reality the pedophile priest crisis is a “social construction” that fits best into what sociologists since the 1970s 2 Department
of Psychiatry, University of New Mexico, P.O. Box 57088, Albuquerque, New Mexico 87187-7088.
P1: FNN Archives of Sexual Behavior [asb]
402
PL111-342a,b,c,d
June 24, 2000
10:36
Style file version Nov. 19th, 1999
Book Reviews
have called a moral panic, which he defines as “a sudden manifestation of exaggerated public fear and concern over an apparently novel threat” (p. 169). Jenkins adds: “Panics are important because they reflect deep underlying social tensions over matters as diverse as ethnicity, social change, and a crisis in values and social attitudes” (p. 170). He continues: As in other historical periods in which there has been intense hostility to the church and its clergy, it is not necessary to suppose that the frequency or gravity of clerical misdeeds has increased significantly in recent years. What has changed is the moral perception of the public and, crucially, of the Catholic laity themselves, who now provide a hungrily receptive audience for claims of priestly atrocities. (p. 170)
Jenkins argues that “the anticlerical insurgency has historically portended periods of sweeping and often painful internal reform for the church, marked by growing intrusions from the secular state and its legal apparatus” (p. 170). With such different perspectives, the real question is, Who is correct—Berry or Jenkins? Berry is a journalist, not an academician. His methods are those of descriptive, investigative journalism, which, by selectively directing the reader’s awareness and attention, help to create the crisis about which he writes, similar to how a measuring device can effect what is being measured in physics. Berry does not have the training, nor is he expected to, to determine the causes of social phenomena. Jenkins, however, is an academic social scientist who potentially has this capacity. If social science is judged by the same criteria by which physical and natural sciences are judged, Jenkins’ arguments, albeit rhetorically persuasive, are scientifically weak. All of the data that Jenkins used is public information, available in various databases, most of which are accessible on the Internet. However, Jenkins treats the data anecdotically rather than systematically without using the power of quantitative, statistical methods. An example of the persuasive power of quantitative, statistical social science would be Daly and Wilson (1988), who used previously published information on human homicide as evidence for the predictive potency of a particular behavioral paradigm about the biosocial causes of homicide. In contrast to homicide, however, where the public record is very likely to be reflective of the actual prevalence of the behavior in a society, sexual behavior between adult men and pubescent boys has been a very private affair in Western societies since the early days of the Roman Empire. It is legendary that in preRoman Greek society such behavior was socially acceptable and normative (Dover, 1978; cf. Thornton, 1999), as it still is in a somewhat different form in certain Highland New Guinea societies, as reviewed by Schiefenhovel (1990). In the Orient, open pederasty between Buddhist priests and their pubescent male acolytes persisted well into the second millennium (Schalow, 1990). Often, over various time frames, for biosocial reasons that are not well understood, the prevalence of a particular behavior in a particular human society varies. For reasons that are beyond the scope of this review, this variation almost always manifests itself by contracting or expanding the contexts in which the behavior
P1: FNN Archives of Sexual Behavior [asb]
Book Reviews
PL111-342a,b,c,d
June 24, 2000
10:36
Style file version Nov. 19th, 1999
403
is expressed. There are many potential examples, of which I cite three that are sexologically relevant: (1) adult male/adult male, physical contact aggression in Western, industrialized societies has decreased over the last millennium and is now contextually channeled into highly ritualized, contact sports; (2) adult male/adult male, full body contact, physical, affiliative behavior (e.g., hugging) in Western industrialized societies has increased over the last few generations, but, except between homosexual males, is contextually limited to greetings, departures, and brief moments of joy and grief; (3) publicly exposing the previously covered female breast in at least some Western industrialized societies (e.g., in Europe) has increased over the last few generations, but is contextually limited to the beach. Following the above examples, pederasty, which was openly lauded and expressed in general pedagogical contexts in ancient Greece (Plato, Symposium 209C), is now narrowly contextualized as a nonopenly discussed and expressed behavior in what appears to be a functionally–lineal vestige of the ancient Greek, mentor–pupil, role-type, adult male/pubescent boy relationship with a very notable exception that is addressed later. Within this narrowly contracted context of the 20th century, all we know about the current prevalence and expression of pederasty is what has been publicly spoken. It is presumptuous to assume that what has been publicly spoken, often in the context of criminal prosecution followed by million-dollar-a-touch civil litigation is a representative sample of the actual expression of the behavior in our society. As pointed out by Boswell (1980): “Since Christianity was the official religion of the Roman Empire from the fourth century on and was the only organized force to survive the final disintegration of Roman institutions in the West after the barbarian invasions of the fifth century, it became the conduit through which the narrower morality of the later Empire reached Europe” (pp. 127–128). In Western civilizations, information that was culturally transmitted through this conduit appears to have been selectively triaged into verbal and nonverbal behavioral components by the standard bearers, most of whom over the last two millennia have been Roman Catholic priests. Some of the most anciently practiced behavior, pederasty to be specific, appears to have been transmitted in the Western, industrialized world in nonverbal behavior only, by the very persons who publicly, at least since mandatory celibacy in the 12th century, were speaking against it (Boswell, 1980). The above type of discordance between speech and behavior is both common and understandable in terms of the function of human language, which, according to evolutionary theory, had to have evolved to get listeners to do what is in the best interest of speakers, not to convey “truth.” Because natural selection often configures our perceptual ability so as to block our awareness of the motivation of our own social behavior (i.e., self-deception), it is often easier to see the discrepancies between our own speech and behavior by analogy by observing similar discrepancies in non-Western societies. The pioneering, early studies in Polynesian and Melanesian anthropology (e.g., Mead and Malinowski) have embarrassingly taught us that in simple societies, in reference to sexual behavior in particular,
P1: FNN Archives of Sexual Behavior [asb]
404
PL111-342a,b,c,d
June 24, 2000
10:36
Style file version Nov. 19th, 1999
Book Reviews
there is almost always a marked difference between what people say should be done, what they actually do, and what they say they do. Even today, in some of the highly sex-stratified societies in the Highlands of New Guinea, the older males still publicly warn the younger adult males of the dangers of women, especially of their vaginal secretions, while at the same time they covertly continue to father the children (Schiefenhovel, 1990). In light of the above, both Berry and Jenkins are partially correct, although both of their books reflect a narrow anthropo-, ethno-, and chronocentrism that precludes any real understanding of pederasty with anything more than the blindsightedness of our times. Berry did publicly verbalize (an “outing” in the vernacular) what for two millennia had been expressed privately in behavior only (e.g., Quinn, 1989), although in the process of doing this he unfairly blurred the important distinctions between homosexuality and pederasty. Jenkins correctly pointed out that the media magnification of the process can be conceptualized within the paradigm of social constructionism. The area in which they disagree—the extent to which mandatory celibacy in Roman Catholic priests contributes to the continuing propagation of pederasty in Western societies—is still an open question. In defense of the Roman Catholic priesthood, pederasty has also been propagated by adult men in other mentor-like relationships with pubescent boys in institutions like the Episcopalian priesthood, secular schools, and the Boy Scouts (Boyle, 1994). However, common sense does suggest that in a sexually reproducing species, a celibate class of adult men should attract “sexual outliers,” which it does. Second, based on unpublished data that I have gathered over the last 20 years in a nowclosed, residential treatment center for Roman Catholic priests in New Mexico, and extrapolating from who could least be celibate with whom, adult women are a very underrepresented category. Bullough’s (1976) historical description of Greek pederasty that existed more than two millennia ago can be extrapolated with a notable exception to the types of relationships that have continued to covertly exist between some priests and some pubescent boys over the centuries: Adding to the acceptance of homosexuality was the institutionalization of pederasty within the educational system. The relationship between the adult and the adolescent boy was maintained by daily association, personal contact and example, intimate conversations, a sharing in more or less common life, and the gradual initiation of the younger into the social activities of the older men. . . . In this education, the family was more or less ignored. . . . Fathers paid little attention to their own male children but instead left their upbringing to an adult male whose relationship as lover of the son was a “union far closer” than what bound parents to children. Public opinion, and in some cases even the law, held the lover morally responsible for the development of his beloved. Pederasty came to be considered the most beautiful, the most perfect form of education. (pp. 108–109)
And this is the notable exception. In ancient Greece, parents sought out adult male mentors for their young sons. The relationships occurred at least with parental consent, although the issue of the capacity of consent of the younger interactant is as much a morally troublesome issue then as it is now. Nevertheless, even with this notable exception, without the aid of historians, human memory is conveniently short.
P1: FNN Archives of Sexual Behavior [asb]
PL111-342a,b,c,d
June 24, 2000
Book Reviews
10:36
Style file version Nov. 19th, 1999
405
Who should buy this book? Jenkins’ book is useful to persons interested in the application of methods of social psychology and the theory of social constructionism to certain social phenomena (“moral panics”) that appear to be magnified by the mass media. In spite of my previous reservations about its qualitative rather than quantitative methodology, as well as its narrow centrisms, Jenkins’ book is a well-done contribution to social science. Compared to some of the more notorious books on the subject of pederasty (e.g., Brongersma, 1986), it is at least relatively objective. However, in spite of its title, the book has only minor usefulness to sexologists interested in furthering their understanding of pedophilia and priests. REFERENCES Berry, J. (1992). Lead Us Not Into Temptation: Catholic Priests and the Sexual Abuse of Children, Doubleday, New York. Boswell, J. (1980). Christianity, Social Tolerance and Homosexuality: Gay People in Western Europe from the Beginning of the Christian Era to the Fourteenth Century, University of Chicago Press, Chicago. Boyle, P. (1994). Scout’s Honor: Sexual Abuse in America’s Most Trusted Institution, Prima Publishing, Rocklin, CA. Brongersma, E. (1986). Loving Boys (Vol. 1): A Multidisciplinary Study of Sexual Relations Between Adult and Minor Males, Global Academic Publishers, Elmhurst, NY. Bullough, V. L. (1976). Sexual Variance in Society and History, University of Chicago Press, Chicago. Daly, M., and Wilson, M. (1988). Homicide, Aldine de Gruyter, New York. Dover, K. J. (1978). Greek Homosexuality, Harvard University Press, Cambridge, MA. Quinn, P. A. (1989). Better than the Sons of Kings: Boys and Monks in the Early Middle Ages, Peter Lang, New York. Schalow, P. G. (1990). Introduction. In Saikaku, I. (ed.), The Great Mirror of Male Love, Stanford University Press, Stanford, CA. (Originally published 1687.) Schiefenhovel, W. (1990). Ritualized adult-male/adolescent-male sexual behavior in Melanesia: An anthropological and ethological perspective. In Feierman, J. R. (ed.), Pedophilia: Biosocial Dimensions, Springer-Verlag, New York, pp. 394–421. Thornton, B. S. (1999). Review of Bisexuality in the Ancient World. Arch. Sex. Behav. 28: 400–404.
Sex and Sexuality in Latin America. Edited by Daniel Balderston and Donna J. Guy. New York University Press, New York, 1997, 304 pp., $20.00 (paperback). Reviewed by Richard G. Parker, Ph.D.3
There has been a veritable boom in research on sexuality in Latin America over the course of the 1990s. The reasons for this boom are diverse, ranging from the emergence and development of feminist and gay/lesbian rights movements to Brasileira Interdisciplinar de AIDS (ABIA), Avenida Rio Branco 42, 22◦ andar, CEP 20090-003, Centro Rio de Janeiro, RJ, Brasil.
3 Associacao
P1: FNN Archives of Sexual Behavior [asb]
406
PL111-342a,b,c,d
June 24, 2000
10:36
Style file version Nov. 19th, 1999
Book Reviews
concern with a range of health issues, such as population, reproductive health and rights, and the HIV/AIDS epidemic. Whatever the reasons, however, sexuality has become a focus for research in a growing range of disciplines, and an important window for understanding life in Latin American cultures. This volume brings together some of the best of this work, particularly in the disciplines of anthropology, history, and literary and cultural studies. Following a brief introduction by the editors that effectively maps some of the major areas of research on sex and sexuality in Latin America, the volume has been divided into four major sections. The first of these sections, “Questioning Identities,” focuses above all on the fluidity of gender and sexuality in a range of Latin American societies, drawing attention in particular to transgender movements and the playful transgression of sexual boundaries. The second section, “Policing Sexuality,” on the contrary, looks at the diverse ways in which sexual (and gendered) transgressions have been repressed and controlled by diverse social institutions, such as the military, the police, the criminal justice system, and even governmental decree in different Latin American societies. The third section, “Family Values,” revisits the importance of family in structuring the sexual field in Latin American societies, but without the facile assumptions about traditionalism and patriarchal values that have typically taken shape as stereotypes about Latin American culture—offering, on the contrary, a far more nuanced vision of the remarkably complex family systems found throughout the region, and their multidimensional configuration of gender and sexuality among their members. The fourth section, “Redefinitions,” concentrates analysis on a theme that in fact runs throughout the volume: the complex processes of change that can be found taking place today in the social construction of gender and sexuality in Latin American societies, in a region in which tradition, modernity, and postmodernity so often seem to coexist and interact in the course of daily life. Although all of the essays collected in this volume are generally of high quality, a number stand out in particular as especially characteristic of some of the best research in the field. The lead essay by Lancaster on playful gender crossing in the daily interaction of a Nicaraguan family offers a particularly compelling application of recent work by Butler and others on gender as performance (as well as on the phenomenologic conditions in which anthropologic field research on gender and sexuality in fact gets done). Caulfield’s historical study of prostitution in Rio de Janeiro between 1850 and 1942 provides vivid evidence of the extent to which sexuality has served as a complex field of power in which race and ethnicity, gender and sexuality, and local and national politics have intersected and interacted in the construction or constitution of nation building and national identity. Guy’s historical study of Argentine concepts of mothering in the late 19th and early 20th centuries offers equally compelling insight into the complex tensions and interplay between religion and social hygiene as competing forces shaping multiple, contested, and constantly changing understandings of the nature of mothering and motherhood. And Balderston’s analysis of the films of Mexican
P1: FNN Archives of Sexual Behavior [asb]
PL111-342a,b,c,d
Book Reviews
June 24, 2000
10:36
Style file version Nov. 19th, 1999
407
director Jaime Humberto Hermisillo offers a sense of the fundamental importance and complexity of the cultural construction of bisexuality as well as of the complex sexual politics that is evolving today in Latin American cultural production. Indeed, it is perhaps this tension between an emphasis on the cultural construction of sexuality and the sexual construction of culture that most clearly runs throughout the essays in this volume, in a sense creating a thread that brings them together conceptually as well as topically. Although it is unfair to expect everything of a single volume, the limitations of Balderston and Guy’s collection make it more useful for some readers than for others. Perhaps most obviously, all of the essays emerge from the growing field of cultural studies, primarily in the United States, and the volume thus provides little or no insight into the rapidly growing body of literature on sexual behavior and practice that has emerged particularly in the wake of HIV/AIDS (usually with a more sociologic or even epidemiologic and/or sexologic perspective). Perhaps even more troubling is that although many of the authors are themselves Latin American or Latinos, all are based in North American (or, in one case, European) universities, so that the collected essays offer no sense of the important body of research that is currently emerging in Latin America itself. In this regard, however, the editors have usefully chosen to close off the volume with an extensive bibliography of gender and sexuality studies in Latin America that provides one of the best available overviews of the current research in this area, and that helps to direct interested readers to at least some of the other lines of research currently being carried out on and in Latin America.
Role of Sexual Abuse in the Etiology of Borderline Personality Disorder. Edited by Mary C. Zanarini. American Psychiatric Press, Washington, DC, 1997, 236 pp., $40.00. Reviewed by Elsa Marziali, Ph.D.4
In the first chapter of this edited volume, Zanarini sets the tone for the research reports to follow by cautioning the reader that although childhood sexual abuse may be present in the histories of some disorders common to women, as, for example, borderline personality disorder (BPD), not all women with this diagnosis have been sexually traumatized during childhood. Rather, the etiology of BPD is best represented by a multidimensional model, which includes issues of genetic/biological predisposition, adverse environmental/familial factors, possible neurologic impairments, and early childhood trauma (sexual abuse being one example). 4 Faculty of Social Work, University of Toronto, 246 Bloor St. West, Toronto, Ontario M5S 1A1 Canada.
P1: FNN Archives of Sexual Behavior [asb]
408
PL111-342a,b,c,d
June 24, 2000
10:36
Style file version Nov. 19th, 1999
Book Reviews
Several chapters (2, 3, 8, and 10) focus directly on examining associations between the experience of childhood sexual abuse and subsequent diagnosis of BPD in adult women. There is considerable consensus across these studies regarding the prevalence of childhood sexual abuse in BPD patients, with rates ranging between 20% and 80%. The higher rates overrepresent single incidents of sexual abuse and include, as well, all categories of perpetrators (Paris and Zweig-Frank, Chapter 2), whereas the lower rates are more apt to signify caretaker perpetrators (father, mother, stepfather, stepmother). Of importance for the BPD group who have experienced childhood sexual abuse are the distinctions made as to (1) age of first experience; (2) whether penetration occurred; (3) frequency and duration of episodes; and (4) number of perpetrators. Equally important in determining the impact of early sexual abuse on the severity of symptoms of BPD are the coexisting factors of physical abuse, emotional and physical neglect, and repeated separations from or loss of a parent. The reported studies challenge the hypothesis suggested by Herman and van der Kolk (1987) that BPD is related to posttraumatic stress disorder (PTSD) due to the experience of early childhood trauma, including sexual abuse. Not surprisingly, this single-factor hypothesis as to the etiology of BPD is not supported. Rather, several of the chapter authors conclude that distinctions must be made as to which etiologic factors apply to which subgroups of borderline patients. For example, certain features of temperament and/or neurologic dysfunction combined with early experiences of abuse (physical, sexual, and neglect) may predict to more severe, long-term pathologic outcomes. Whereas when these early experiences do not occur in combination, the diagnosis of BPD in adults may result, either through a pathway of abuse and neglect or through a pathway of neurologic injury (Rising Kimble et al., Chapter 9). These are important distinctions because of their obvious relevance for understanding (1) symptom range and severity; (2) types of comorbidity; (3) response to different treatment orientations, structures, and duration; (4) lifetime course; and (5) ultimate outcomes. For example, the possibility that a spectrum of severity exists within the BPD diagnostic group has been supported (Marziali et al., 1994). Three chapters (4, 6, and 7) examine the prevalence and meanings of specific acting out and self-destructive behaviors in women with BPD. Jordan et al. (Chapter 4) conducted a study of the psychosocial status of women prison inmates. In particular, they explored the current and life history experiences of abuse and “extreme events” in three disorders identified in the study population: BPD, PTSD, and antisocial personality disorder (ASPD). Of the original sample of 805 female inmates, about 50% met criteria for one or more of the three disorders. A total of 28% met criteria for BPD and, of these, 50% also met criteria for PTSD. The analyses of early traumatic experiences for inmates with a diagnosis of BPD showed that 32% had been sexually assaulted before age 11. Also, women inmates with BPD tended to grow up in homes where violence against some family members was common; however, few of the inmates (either with or without BPD) had
P1: FNN Archives of Sexual Behavior [asb]
Book Reviews
PL111-342a,b,c,d
June 24, 2000
10:36
Style file version Nov. 19th, 1999
409
experienced serious physical abuse in early childhood. Of note is the fact that in a model-fitting regression analysis, only 17% of the variance was accounted for by the study variables (e.g., sexual abuse, family violence, early loss, and/or separations). Thus, the contribution of factors not included in the study (e.g., biologic disposition, neurologic impairment, other family dynamics) are unknown. Zweig-Frank and Paris (Chapter 6) and Dubo et al. (Chapter 7) present data on the possible associations between childhood trauma and self-mutilation and suicidal behaviors in BPD. Zweig-Frank and Paris’ interesting study looked at childhood sexual abuse as a predictor of dissociation and self-mutilation in adult subjects with BPD. Neither dissociation nor childhood sexual abuse accounted for self-mutilation, and the authors review a number of possible explanations for these negative findings, including biologic and social factors. As expected, Dubo et al. found that 78% of the BPD subjects reported histories of self-mutilation and 80% reported histories of suicide attempts. Of the multiple childhood abusive factors studied, only caretaker sexual abuse was associated with lifetime number of self-mutilation episodes and duration of self-mutilation behavior. There were no associations between any of the abuse variables and number of suicide attempts, but duration of suicidal behaviors across the life span was strongly associated with caretaker sexual abuse, caretaker emotional denial, and lack of a real relationship with a caretaker. In discussing their findings, this group of investigators raise some interesting questions about the etiology of BPD. Does early physical and/or sexual abuse, neglect, loss or separations predict the onset of BPD? Or is a more complex explanatory model needed? They speculate that affect dysregulation witnessed in patients with BPD is associated with self-harming behaviors, and that the inability to process emotions effectively is due to early disruptions of attachment to primary caregivers due to separations, abuse, and neglect. Furthermore, these negative experiences may interfere with normal development of biologic systems of affective self-regulation (Cicchetti, 1989; van der Kolk and Greenberg, 1987). These findings have been further supported by studies that show that early childhood experiences of abuse may affect the laterality of the developing brain, possibly leading to increased right hemispheric activity (Teicher et al., 1994). The resulting diminished hemispheric intercommunication may contribute to the development of affective instability and impulsivity. Other recent work also suggests that biologic factors may be associated with self-destructive behavior; for example, Simeon et al. (1992) suggest that underlying serotonergic dysfunction may facilitate suicidal behavior and self-mutilation. The final two chapters of the book deal with the treatment implications of the study findings that link early childhood trauma to adult BPD symptoms and self-destructive behaviors. Wagner and Linehan (Chapter 11) provide an overview of their model of treatment, Dialectical Behavior Therapy (DBT), and place special emphasis on the importance of understanding the BPD patient’s early childhood experiences of an “invalidating environment.” The therapeutic strategies of DBT
P1: FNN Archives of Sexual Behavior [asb]
410
PL111-342a,b,c,d
June 24, 2000
10:36
Style file version Nov. 19th, 1999
Book Reviews
reflect acceptance of the patient’s current capacities balanced with acknowledgment of the patient’s potential for change. Linehan (1993) outlines three stages of DBT: (1) building skills for managing daily living more effectively; (2) emotionally processing past traumatic events; and (3) developing the ability to trust the self. The importance of developing a positive therapeutic relationship, necessary for sustaining the work of therapy, is emphasized. In the final chapter, Gunderson argues that the treatment needs of BPD patients are long term and best met by having one therapist–manager oversee the multiple inputs and course of the therapy. The initial phase of therapy requires the therapist to help the patient to cope with real life stresses and a cognitive behavioral approach plus the involvement of significant others are thought to be important. To sustain this preliminary therapeutic work, the patient and therapist must develop a stable therapeutic alliance that involves the therapist’s validation of the patient’s painful early life trauma while being sensitive to the transference significance of what is disclosed. Gunderson believes that dynamic psychotherapy is effective only after stabilization of disruptive behaviors and alliance formation. In summary, Zanarini has collected a series of important studies that provide a balanced view of the contribution of early childhood trauma to the development of BPD in adults. Both her own work and that of the investigators whose research reports comprise the book chapters support a multifactorial model of the etiology of BPD. The implication for treatment are equally complex. Etiologic models that include neurologic factors would suggest that pharmacologic treatments might be more efficacious but there is as yet little evidence to support this approach. Psychotherapeutic models that include components of cognitive behavioral interventions have been shown to be especially effective in reducing self-destructive behaviors (Linehan, 1993). Individual psychodynamic psychotherapy has been the most frequently used model of treatment for BPD. Patients who are able to engage in, and sustain, this long-term, exploratory model of treatment do improve. Of concern has been the high dropout rate (i.e., for a subgroup of patients with BPD the frequency and severity of their acting out behaviors preclude commitment to a self-reflective, insight-oriented approach to changing pathologic behavior). Interpersonal Group Psychotherapy (IGP) has proved effective in dealing with early alliance building, management of self-destructive behaviors, and changes in self concept that generalize to situations outside of treatment (Marziali and MunroeBlum, 1994). In contrast to other models of treatment that have been tested with BPD, IGP places special emphasis on helping the patient mourn the loss of a wished for, nurturing, sensitive primary caregiver. In actuality, the primary caregiver was psychologically lost because early childhood experiences of abuse and neglect disrupted the development of an optimal attachment experience. It is this painful reality that the patient must acknowledge while drawing on personal resources needed to mourn the loss and to affirm self-respect and control over one’s life. A group therapeutic environment is well suited to achieving these goals.
P1: FNN Archives of Sexual Behavior [asb]
PL111-342a,b,c,d
June 24, 2000
Book Reviews
10:36
Style file version Nov. 19th, 1999
411
REFERENCES Cicchetti, D. (1989). How research on child maltreatment has informed the study of child development: Perspectives from developmental psychology. In D. Cicchetti and V. Carlson (eds.), Child Maltreatment, Cambridge University Press, Cambridge, England, pp. 377–431. Herman, J. L., and van der Kolk, B. A. (1987). Traumatic antecedents of borderline personality disorder. In B. A. van der Kolk (ed.), Psychological Trauma, American Psychiatric Press, Washington, DC, pp. 11–26. Linehan, M. M. (1993). Cognitive–Behavioral Treatment for Borderline Personality Disorder, Guilford Press, New York. Marziali, E., and Munroe-Blum, H. (1994). Interpersonal Group Psychotherapy for Borderline Personality Disorder, Basic Books, New York. Marziali, E., Munroe-Blum, H., and Links, P. (1994). Severity as a diagnostic dimension of borderline personality disorder. Can. J. Psychiat. 39: 540–544. Simeon, D., Stanley, B., Frances, A., Mann, J. J., Winchel, R., and Stanley, M. (1992). Self-mutilation in personality disorders: Psychological and biological correlates. Am. J. Psychiat. 149: 221–226. Teicher, M. L., Ito, Y., Glod, C. A., et al. (1994). Early abuse, limbic system dysfunction, and BPD. In Silk, K. R. (ed.), Biological and Neurological Studies of Borderline Personality Disorder, American Psychiatric Press, Washington, DC, pp. 177–207. van der Kolk, B. A., and Greenberg, M. S. (1987). The psychobiology of the trauma response: Developmental issues in the psychobiology of attachment and separation. In B. A. van der Kolk (ed.), Psychological Trauma, American Psychiatric Press, Washington, DC, pp. 63–87.
