Evaluation of a Sexual Assault Education/Prevention Program for Female U.S. Navy Personnel

Transcription

1 MILITARY MEDICINE, 176, 10:1178, 2011 Evaluation of a Sexual Assault Education/Prevention Program for Female U.S. Navy Personnel Terri J. Rau, PhD * ; Lex L. Merrill, PhD ; Stephanie K. McWhorter, MA ; Valerie A. Stander, PhD ; Cynthia J. Thomsen, PhD ; Christopher W. Dyslin, PhD ; Julie L. Crouch, PhD ; Mandy M. Rabenhorst, PhD ; Joel S. Milner, PhD ABSTRACT The U.S. Navy Sexual Assault Intervention Training (SAIT) program for women was evaluated in a randomized clinical trial. The SAIT uses multiple presentation modalities (lecture, slides, discussion, film) to provide information related to sexual assault, including risk factors, consequences, prevention, and relevant military regulations. Female personnel who had completed basic training ( N = 550) participated in the SAIT or a Comparison condition, and then completed measures of rape knowledge, empathy for rape victims, and acceptance of rape myths (false beliefs about rape justifying sexual violence). Results showed that the SAIT increased factual knowledge about rape. In addition, the SAIT increased empathy with rape victims in some groups of women. However, the program did not reduce women s rape myth acceptance. Given the enormity of the problem of sexual assault and these promising initial findings, additional research on the efficacy of the SAIT is clearly warranted. INTRODUCTION Experiences of sexual assault are associated with poor mental and physical health outcomes, as well as increased likelihood of future sexual assault. 1 3 Unfortunately, the sexual assault of women is also common. Based on surveys of general population and college student samples, more than one woman in four has experienced some form of sexual assault, and 15% to 20% have experienced completed rape. 2 5 Among female U.S. Navy (USN) recruits, 27% to 31% report premilitary rape, 6,7 suggesting higher rates of sexual assault in this population. The high prevalence of sexual assault among USN women has spurred prevention efforts. In 1994, the USN implemented the Sexual Assault Victim Intervention (SAVI) program. SAVI includes two components: education to increase awareness of sexual assault, and advocates helping sexual assault survivors and their families. 8 To complement the SAVI program, the Bureau of Naval Personnel sponsored efforts to develop separate evidence-based Sexual Assault Intervention Training (SAIT) programs for male and female USN personnel. A previous evaluation of the program for men showed that the SAIT increased men s knowledge about the problem of acquaintance rape and their empathy for women who have experienced rape, * Research, Development and Performance Management, Navy Fleet and Family Support Program, Navy Installations Command, Millington Detachment, 5720 Integrity Drive, Millington, TN Behavioral Science and Epidemiology Department, Naval Health Research Center, 140 Sylvester Road, San Diego, CA Center for the Study of Family Violence and Sexual Assault, Department of Psychology, Northern Illinois University, DeKalb, IL Division of Psychology and Counseling, Governors State University, One University Parkway, University Park, IL The views expressed in this article are those of the authors and do not reflect the official policy or position of the U.S. Department of the Navy, U.S. Department of Defense, or U.S. Government. This research has been conducted in compliance with all applicable federal regulations governing the protection of human subjects in research (Protocol NHRC ). and reduced their acceptance of rape myths. 9 Rape myths are commonly held false beliefs that might be used to justify male sexual aggression (e.g., A woman who was raped probably brought it on herself. ). 10,11 The present study evaluated the effectiveness of the SAIT program for women using a randomized, controlled clinical trial. Women were randomly assigned to participate in the SAIT program or a Comparison program. Furthermore, approximately half of the participants in each condition were randomly assigned to complete the outcome measures at pretest as well as post-test. This study parallels a previously published evaluation of the SAIT program for men, 9 facilitating a comparison of program effectiveness across the two genders. As in the evaluation of the SAIT program for men, it was expected that the SAIT program for women would effectively increase knowledge about rape, increase empathy for rape survivors, and reduce acceptance of rape myths. Beyond examining the overall effectiveness of the SAIT for women, this study addressed two additional issues. First, comparing participants who did or did not receive a pretest allowed us to determine whether pretesting influenced responses to the intervention (a pretest sensitization effect). Previous research on the effectiveness of mental health interventions suggest that treatment effects often are enhanced when pretests are administered. 12 Second, sexual assault risk reduction programs sometimes have been less effective for women who previously had been sexually assaulted. 13,14 Because of this, we examined whether the SAIT program was more effective for women without a history of sexual assault. METHODS Participants The participant pool consisted of female USN personnel at U.S. Naval Base, Great Lakes, Illinois, who had completed 1178 MILITARY MEDICINE, Vol. 176, October 2011

2 boot camp and were awaiting further training. Of 556 women who arrived at a session, 550 (99%) completed the study. The average participant was single (89%), 20 years old ( M = 19.9, SD = 3.0, range = 17 35), and had 12.6 years of education (SD = 1.1). Over half (58%) of the sample was White; the remainder was African American (19%), Hispanic (11%), Asian (4%), Native American (3%), or other (5%). Annual family of origin income was less than $25,000 for 34%, between $25,000 and $50,000 for 42%, and more than $50,000 for 24%. Preliminary analyses comparing the demographic characteristics of women by treatment Condition and Pretesting revealed that race/ethnicity varied across Condition, c 2 (3, N = 546) = 8.42, p < Although Whites were evenly distributed across the SAIT and Comparison conditions (49% vs. 51%), a greater percentage of women of other racial/ethnic groups had been assigned to the SAIT condition (African Americans, 65%; Hispanics, 58%; other, 56%). To control for this initial difference, analyses included race/ethnicity (White/ non-white) as a factor. No other characteristics differed by Condition or Pretesting. Instruments and Educational Materials Rape knowledge was assessed using a 14-item Rape Knowledge Scale (RKS) developed for the present study. Item content was developed based on material presented in the SAIT, and included rape prevalence, military definitions of rape, effects of sexual assault on women, ways to avoid sexual assault, and strategies for coping with sexual assault. The number of correct responses could range from 0 to 14. Empathy for rape survivors was assessed with the 19-item Rape Empathy Scale (RES).15 Each item consists of a pair of statements about rape, one representing the psychological perspective of a rape survivor and the other representing the perspective of the rapist. For example, survivor (rapist) sympathy is indicated by agreeing that I would find it easier to imagine how a rape victim (rapist) might feel during an actual rape than how a rapist (rape victim) might feel. Respondents were asked to choose the statement from each pair that best reflected their own beliefs and to rate their preference for the chosen statement (1 = no preference, 7 = strong preference). Responses to each item were recoded to a 14-point scale on which 1 = strong preference for the rapist perspective and 14 = strong preference for the survivor perspective. Total RES scores thus have a theoretical range of 19 to 266, with higher scores indicating greater empathy for rape survivors (Cronbach s a = 0.80). Acceptance of rape myths was assessed with two measures: the 19-item Rape Myth Scale (RMS) 10 and the 11-item Rape Myth Acceptance Scale (RMAS). 11 Both employ 7-point Likert-type response scales, and item responses are summed to compute total scale scores, with higher scores reflecting greater rape myth acceptance (Cronbach s a s = 0.90 and 0.70, respectively). Sample items include Many so-called rape victims are actually women who had sex and changed their minds afterwards (RMS ) and When women go around braless or wearing short skirts and tight tops, they are just asking for trouble (RMAS). Finally, women s prior rape histories were assessed using the Sexual Experiences Survey (SES). 16 The SES has been widely used in research, facilitating comparisons with previous studies. It describes 10 different forms of unwanted sexual experiences with men. Respondents who reported an experience meeting legal definitions of rape since the age of 14 were classified as having experienced rape. SAIT for Women The SAIT program for women focuses on preventing sexual assault by male acquaintances. It covers military definitions and consequences of sexual assault perpetration. The program discusses the epidemiology of sexual assault, the physical and psychological consequences, differences in women s and men s perceptions of sexual situations, and specific risk reduction strategies. It also debunks common rape myths, reviews women s options in the event of sexual assault, and provides suggestions for assisting other women who have been sexually assaulted. The program consisted of a lecture supplemented by 57 slides, two brief discussion opportunities, three audio dramatizations, and 25 minutes of the film When a Kiss Is Not Just a Kiss: Sex Without Consent. Comparison Education Program Women in the Comparison condition viewed an educational film, entitled Conditions of Secrecy: A Video Drama about AIDS and Youth. The film portrayed a male college student who contracted HIV through a sexual encounter. This topic was chosen because it focused on sexual content, but did not address issues related to sexual assault. The film was edited to be approximately equal in length to the SAIT program. Procedure Study sessions were held during prearranged times each week. Female USN personnel were referred by their commanders. Unknown to the referral source, each session had been randomly assigned to one of four conditions (SAIT with pretest, SAIT without pretest, Comparison with pretest, and Comparison without pretest). Across an 18-month period, 74 sessions were conducted (38 SAIT and 36 Comparison). The number of participants in each session ranged from 1 to 24 ( M = 7.4, Mdn = 5) and did not vary by Condition, Pretesting, or their interaction ( p > 0.5). Overall, 297 women participated in the SAIT program (153 with pretesting and 144 without) and 253 participated in the Comparison program (121 with pretesting and 132 without). Two civilian female instructors, who conducted both SAIT and Comparison sessions, read a military Privacy Act Statement and informed consent information to women at each session. Participants were then given an opportunity to ask questions before signing consent forms and completing personal information forms. To ensure anonymity, these forms were collected and placed in envelopes before the study session began. Personnel who elected not to participate (6/556; 1%) returned to their commands for alternative assignment. MILITARY MEDICINE, Vol. 176, October

3 Analytic Strategy The efficacy of the SAIT program was tested in two ways. First, we examined changes in the outcome measures from pretest to post-test, using 2 (Condition) 2 (Time: pretest/ post-test) mixed effects analyses of variance (ANOVAs). By design, these analyses included only participants who completed the pretest. It was expected that participants in the SAIT condition, compared to those in the Comparison condition, would show greater increases in knowledge and rape survivor empathy, as well as greater reductions in rape myth acceptance (i.e., an interaction of Condition Time). Second, post-test scores on the outcome measures were compared for all groups using a 2 (Condition) 2 (Pretesting) betweensubjects ANOVAs. It was expected that scores would be more favorable following exposure to the SAIT program vs. the Comparison program (i.e., a main effect of Condition). In addition, we expected that differences between participants in the SAIT and Comparison programs would be greater among those who completed a pretest. Both sets of analyses also included two additional between-subjects factors: race/ ethnicity (White/non-White) and prior rape (yes/no). RESULTS Descriptive statistics and correlations between the outcome measures at pretest are in Table I. Respondents evidenced low rape myth acceptance and high rape survivor empathy. Correlations among all measures were statistically significant ( p < 0.001). Preliminary 2 (Condition) 2 (Race/Ethnicity) ANOVAs revealed that pretest levels of rape knowledge, rape myth acceptance, and rape empathy did not differ by Condition. However, there was a significant Race/Ethnicity effect for three of the four outcome measures. As can be seen in Table II, TABLE I. Means and Correlations Between Outcome Measures at Pretest Outcome Measure RKS RMAS RMS RES RMAS 0.28 * RMS 0.25 * 0.69 * RES 0.25 * 0.49 * 0.59 * M SD n * p < White women, relative to non-white women, exhibited significantly higher levels of rape knowledge and rape survivor empathy and lower levels of rape myth acceptance (on the RMAS but not the RMS). The interaction of Condition Race/Ethnicity was not significant for any of the four outcome measures. More than half of the participants (52%) reported some form of adult sexual assault, and more than one in four (26%) reported completed rape. Women s Race/Ethnicity was unrelated to the likelihood that they reported previous rape ( p > 0.6). Rape Knowledge As expected, among women in the SAIT condition who completed a pretest, rape knowledge scores increased significantly from pretest to post-test ( M = ; p < 0.001). In contrast, among women in the comparison condition who completed a pretest, knowledge scores decreased significantly from pretest to post-test ( M = ; p < 0.001), F Time Condition (1, 254) = , p < Consistent with this, an analysis of post-test scores for all participants (not just those who had taken the pretest) revealed that women in the SAIT condition, compared to those in the Comparison condition, had greater rape knowledge at post-test ( M s = and 9.73, respectively), F (1, 519) = , p < Although this held true regardless of pretesting, as predicted the effect was stronger for women who had completed a pretest ( M s = vs. 9.44) than for those who had not ( M s = vs. 9.99), F Condition Pretesting (1, 519) = 21.04, p < The effectiveness of the SAIT in increasing women s rape knowledge did not differ depending on their ethnicity or history of prior rape. Rape Empathy Comparisons of changes in RES scores from pretest to posttest revealed significant increases in rape empathy over time, F (1, 222) = 29.71, p < Although significant in both treatment and control conditions, the increase in rape empathy was greater for women in the SAIT program ( M s = ) than for those in the Comparison condition ( Ms = ), F Time Condition (1, 222) = 4.71, p < These effects did not vary by ethnicity or rape history. Comparing post-test RES scores of all women revealed significant main effects of Condition and Previous Rape. As predicted, empathy scores were higher in the SAIT condition ( M = ) than in the Comparison condition ( M = ), F (1, 473) = 9.07, TABLE II. Results of ANOVAs on Pretest Scores Outcome Measure White Non-White df error F Race/Ethnicity F Condition F Race Condition RKS * RES * RMAS ** RMS n s for White participants ranged from 155 to 166; n s for non-white participants ranged from 90 to 107. * p < **p < MILITARY MEDICINE, Vol. 176, October 2011

4 p < In addition, women who had experienced rape ( M = ) had greater rape empathy than women who had not been raped ( M = ), F (1, 473) = 4.10, p < A significant interaction effect indicated that differences in RES scores between women in the SAIT and Comparison conditions varied depending on both whether women had been pretested and whether they had previously been raped, F Condition Pretesting Prior Rape (1, 473) = 5.16, p < Relevant means are provided in Table III. In contrast to some previous research in which sexual assault prevention programs were ineffective for women who had been sexually assaulted in the past, 13,14 the present results indicate that the effects of the program on rape empathy were greatest for women who had both experienced prior rape and completed the pretest (see Table III ). Rape Myth Acceptance There was no evidence of significant effects of the SAIT on rape myth acceptance. Although there were significant decreases in rape myth acceptance on both measures from pretest to post-test (for the RMS, M = ; for the RMAS, M = ), these effects did not vary by Condition (for the RMS, F [1, 243] = 2.40, p > 0.1; for the RMAS, F [1, 252] = 0.46, p > 0.4). Similarly, analyses of posttest rape myth scores did not reveal the predicted difference at post-test between the SAIT and Comparison conditions (for the RMS, F [1, 511] = 0.06, p > 0.8; for the RMAS, F [1, 518] = 1.60, p > 0.2). Further, significant Pretesting main effects were found for the RMS and the RMAS, F (1, 243) = 9.93, p < 0.01 and F (1, 252) = 59.26, p < 0.01, respectively, indicating significant decreases in rape myth acceptance from pretest to post-test. DISCUSSION Recent meta-analyses have highlighted the paucity of research on the efficacy of rape education/prevention programs designed specifically for women. 17,18 Unarguably, it is essential to change the rape-supportive attitudes of men, who are typically the perpetrators of sexual aggression. At the same time, however, educating women about situations and behaviors associated with increased risk of sexual assault (e.g., going to a secluded place with a man one doesn t know well) may enable them to take precautions that may reduce the likelihood of TABLE III. Mean Scores on Rape Empathy at Post-Test by Condition and Previous Rape Condition Previous Rape Pretesting SAIT Comparison df error t No previous rape Pretest No pretest Previous rape Pretest * No pretest * p < sexual assault. This is the aim of the USN SAIT program for women. Of the three types of outcomes examined, the SAIT program was most effective at increasing rape knowledge. The impact of the SAIT program on rape knowledge was moderate in size, whereas the effects on other outcomes were uniformly small. This is consistent with meta-analytic results showing that sexual assault education programs generally have larger effects on knowledge than on other outcomes, 17 as well as the results of the previously published evaluation of the SAIT for men. 9 Rape knowledge increased for all women in the SAIT condition, regardless of their race/ethnicity or rape history. However, as in previous research on other types of mental health programs, 12 the SAIT increased knowledge more for women who took the pretest than for those who did not, suggesting a pretest sensitization effect. Completing the knowledge scale at pretest may have helped women in the SAIT condition to focus on the most relevant information during the SAIT. Importantly, however, the SAIT was associated with increased knowledge even among women who did not complete a pretest. Consistent with some previous research, at baseline women who had been raped, compared to those who had not, exhibited greater empathy with rape survivors. 19 Furthermore, the SAIT program significantly increased empathy with rape survivors specifically among women who had been raped and who completed the pretest. The direction of this finding is in contrast to previous research, where education interventions have sometimes been less effective with women who have experienced sexual assault than with those who have not. 13,14 Pending replication, this finding suggests that the SAIT intervention, in contrast to some previous interventions, may be particularly effective for women with a history of sexual assault. Although the present research showed improvements in rape empathy only for a subset of the women exposed to the SAIT, this finding is encouraging for two reasons: first, because effects of previous educational interventions on rape empathy have typically been small and nonsignificant 17 ; and second, because the SAIT was most effective in increasing rape empathy among women with a history of sexual assault, who are at increased risk of future sexual assault. 1 In contrast to the positive findings for rape knowledge and empathy, our results provide no evidence that the SAIT program was effective in reducing women s rape myth acceptance. It is not clear why this was the case. It is tempting to argue that scores on the two rape myth scales were, at the outset, quite low and positively skewed, so that it would be difficult to reduce them further (a floor effect ). However, rape empathy scores were equally skewed, and yet significant effects were observed for rape empathy, at least in some groups of women. Overall, rape myth scores were lower at post-test than pretest (regardless of condition), suggesting the operation of pretest sensitization. That is, completing the rape myth scales at the outset of the study session may have sensitized women to MILITARY MEDICINE, Vol. 176, October

5 relevant issues, thereby reducing their post-test scores on those measures. Previous reviews of the effectiveness of sexual assault education programs have not examined the impact of pretesting. Because we used a research design in which half of the women in each condition were pretested and half were not, we were able to examine this issue. Some evidence of pretest sensitization was found for all outcome measures examined. Although pretesting effects are often considered artifacts, they could be interpreted substantively, as indicating that women derive greater benefit from sexual assault education programs when the program is preceded by initial assessments that heighten the salience of relevant issues. That is, completing a pretest may help women to focus on the highlighted material, thereby improving their comprehension and retention of that material. In general, this suggests that in conducting mental health programs, a renewed focus on baseline assessment as an important active ingredient in treatment, rather than simply a means to establish a metric for success, is likely to improve outcomes.20 Moreover, the finding of pervasive pretest sensitization effects suggests that it is important to consider research design when interpreting the results of sexual assault prevention evaluations. Specifically, because of pretest sensitization, studies employing pretests may appear to show greater effects of interventions than do studies without pretests. 12 The effects of the SAIT for women observed in the present study were less pervasive than the effects of the previously evaluated SAIT for men, which produced significant changes in men s rape myth acceptance, as well as in their rape knowledge and empathy. 9 It may be more difficult to effect changes in women s than in men s attitudes regarding sexual assault because women have less room for improvement in terms of their knowledge and attitudes about rape. Furthermore, although rape knowledge, rape empathy, and rape myth acceptance are very commonly used outcomes in evaluations of sexual assault education programs, 17 they may be more relevant indicators of change for men than for women. In future work, it will be important to examine outcomes that might be more closely associated with women s risk of future sexual assault, such as dating behavior or sexual communication. 14 In summary, although the SAIT for women did not decrease rape myth acceptance, it increased rape knowledge, and it increased empathy for rape survivors in some subgroups of women (i.e., pretested women with a history of sexual assault). Importantly, the program was equally effective for White and non-white women. Given the large sample and low attrition rate, the conclusions of the present research are likely applicable to the population of female USN recruits who have completed basic training; our results suggest that this population would likely gain both knowledge and empathy for rape survivors as a result of participating in the SAIT, particularly if they first completed pretest measures sensitizing them to relevant issues. Although the results of this evaluation are encouraging regarding the effectiveness of the SAIT for women, it is important to replicate the present findings. In addition, future evaluations of the SAIT program for women should evaluate the persistence over time of the observed changes in attitudes and knowledge, and measure women s intentions to engage in high-risk behaviors (e.g., heavy drinking while alone at a bar). Ultimately, the best evidence of the SAIT program s effectiveness would be documentation that the program reduces the likelihood of women s future sexual assault. Given the enormity of the problem of sexual assault and these promising initial findings, additional research on the efficacy of the SAIT among USN women is clearly warranted. ACKNOWLEDGMENTS The Naval Health Research Center Report No was supported by the Fleet and Family Support Programs, Personnel Support Department (N2), Commander, Navy Installations, under Work Unit No REFERENCES 1. Classen CC, Palesh OG, Aggarwal A : Sexual revictimization: a review of the empirical literature. Trauma Violence Abuse 2005 ; 6: Elliott DM, Mok DS, Briere J : Adult sexual assault: prevalence, symptomatology, and sex differences in the general population. J Trauma Stress 2004 ; 17 (3) : Tjaden P, Thoennes N : Full Report of the Prevalence, Incidence, and Consequences of Violence Against Women: Findings from the National Violence Against Women Survey. NCJ Washington, DC, U.S. Department of Justice, Koss MP, Gidycz CA, Wisniewski N : The scope of rape: incidence and prevalence of sexual aggression and victimization in a national sample of higher education students. J Consult Clin Psychol 1987 ; 55 (2) : Testa M, VanZile-Tamsen C, Livingston JA, Koss MP : Assessing women s experiences of sexual aggression using the Sexual Experiences Survey: evidence for validity and implications for research. Psychol Women Q 2004 ; 28: Merrill LL, Newell CE, Milner JS, et al : Prevalence of premilitary adult sexual victimization and aggression in a Navy recruit sample. Mil Med 1998 ; 163 (4) : Stander VA, Merrill LL, Thomsen CJ, Crouch JL, Milner JS : Pre-military adult sexual assault victimization and perpetration. J Interpers Violence 2008 ; 23 (11) : Kelley ML, Schwerin MJ, Farrar KL, Lane ME : An evaluation of a sexual assault prevention and advocacy program for U.S. Navy personnel. Mil Med 2005 ; 170 (4) : Rau TJ, Merrill LL, McWhorter SK, et al : Evaluation of a sexual assault education/prevention program for male U.S. Navy personnel. Mil Med 2010 ; 175 (6) : Burt MR : Cultural myths and supports for rape. J Pers Soc Psychol 1980 ; 38: Lonsway KA, Fitzgerald LF : Attitudinal antecedents of rape myth acceptance: a theoretical and empirical reexamination. J Pers Soc Psychol 1995 ; 68 (4) : Willson VL, Putnam RR : A meta-analysis of pretest sensitization effects in experimental design. Am Educ Res J 1982 ; 19 (2) : Hanson Breitenbecher K, Scarce M : A longitudinal evaluation of the effectiveness of a sexual assault education program. J Interpers Violence 1999 ; 14 (5) : Hanson KA, Gidycz CA : An evaluation of a sexual assault prevention program. J Consult Clin Psychol 1993 ; 61: MILITARY MEDICINE, Vol. 176, October 2011

6 15. Dietz SR, Blackwell KT, Daley PC, Bentley BJ : Measurement of empathy toward rape victims. J Pers Soc Psychol 1982 ; 43: Koss MP, Gidycz CA : Sexual experiences survey: reliability and validity. J Consult Clin Psychol 1985 ; 53 (3) : Anderson LA, Whiston SC : Sexual assault education programs: a metaanalytic examination of their effectiveness. Psychol Women Q 2005 ; 29: Brecklin LR, Forde DR : A meta-analysis of rape education programs. Violence and Victims 2001 ; 16 (3) : Smith CA, Frieze IH : Examining rape empathy from the perspective of the victim and the assailant. J Appl Psychol 2003 ; 33: Poston JM, Hanson WE : Meta-analysis of psychological assessment as a therapeutic intervention. Psychol Assessment 2010 ; 22 (2) : MILITARY MEDICINE, Vol. 176, October