Abstract
Previous research has not conclusively determined whether traditional femininity increases women’s risk for sexual assault. Certain femininity beliefs, such as valuing deference, may increase risk for sexual assault by discouraging sexual refusal assertiveness (SRA). Other femininity beliefs, such as valuing purity, may promote self-protective behaviors. College women (N = 254) provided self-report data on these beliefs, risk and protective behaviors, and sexual assault experiences. Traditional femininity was not directly associated with experiencing sexual assault. However, specific traditional beliefs were related to behaviors directly associated with risk. High SRA and sexual abstinence simultaneously reduced the odds for sexual assault.
Introduction
Sexual assault, defined by the Federal Bureau of Investigation as non-consensual sexual penetration (U.S. Department of Justice, 2012), is a common and harmful problem. Younger women appear to be at elevated risk, and specifically, many young women attending college experience sexual assault. In one survey of undergraduates from two large public universities, 14% of women reported a completed sexual assault during college (Krebs, Lindquist, Warner, Fisher, & Martin, 2007). Because many young women are negatively affected by sexual assault during their college years, research is needed to identify risk factors that can inform effective campus risk reduction efforts.
The present study investigated risk for sexual assault as related to college women’s investments in traditional femininity. Sexual assault was operationally defined as unwanted, non-consensual sexual vaginal or oral penetration due to coercive pressure, incapacitation related to substance use, or force (see also Franklin, 2010). Women’s investments in traditional femininity were defined in terms of their feminine ideology beliefs (Levant, Richmond, Cook, House, & Aupont, 2007). We explored whether and in what ways more traditional beliefs predict sexual assault during college, both directly and indirectly via different behavioral tendencies associated with risk.
Through the process of gender role socialization, girls learn about and develop personal values regarding their appropriate feminine attributes and behaviors. This process begins early in life (Murnen & Byrne, 1991). Girls and women typically learn that they are expected to maintain an attractive appearance, to care for others, to fulfill domestic roles, and to remain modest and soft-spoken. These expectations of traditional femininity as constructed by European American, middle and upper class heterosexuals in the U.S. should affect members of that group, as well as every other woman in U.S. society who is held up to those standards and who experiences acceptance or rejection from the majority culture, in part, based on adherence to the powerful group’s femininity norms. (Mahalik et al., 2005, p. 418)
As part of their gender socialization, women are also encouraged to value and maintain romantic relationships with men, an expectation referred to as compulsory heterosexuality (Rich, 1983). As such, traditional gender roles create a power imbalance between women and men. More generally, ideals about how women should act (feminine ideology) emphasize the importance of valuing a man’s wishes and desires over the woman’s own (Greene & Faulkner, 2005). Such power imbalances may affect women’s sexual interactions with men given traditional sexual scripts involving male pursuit and female response (Gavey, 2005). Not surprisingly, many women implicitly associate sex with submission (Sanchez, Kiefer, & Ybarra, 2006).
Some research suggests that women who explicitly endorse traditional feminine ideology beliefs are particularly disempowered during sexual interactions, which may increase their risk for sexual assault. For example, although they did not investigate women’s experiences of sexual assault, Curtin, Ward, Merriwether, and Caruthers (2011) found that college women’s endorsements of traditional feminine ideology were negatively correlated with sexual knowledge, sexual self-efficacy, sexual assertiveness, and comfort with their bodies. Likewise, Katz and Tirone (2009) found that dating women’s investment in ideal womanhood was positively related to consenting to unwanted sex. In other words, heterosexual women invested in ideal womanhood went along with their partner’s wishes instead of assertively refusing sex they did not want. Yet, women’s endorsements of traditional femininity have not consistently predicted sexual disempowerment. For example, Faulkner, Kolts, and Hicks (2008) found no direct relationship between heterosexual women’s feminine ideology beliefs and how long it took them to signal, in response to a vignette, that they would leave in response to a date’s unwanted sexually coercive advances.
These inconsistent findings could be methodological; perhaps traditional feminine ideology affects women’s actual but not hypothetical sexual interactions with men. Alternatively, inconsistent findings might be explained by the fact that feminine ideology is multidimensional and, in some respects, contradictory. Divergent aspects of feminine ideology may be related to sexual assault, but for different reasons and in different ways. For example, a woman cannot be expected to (sexually) submit to men while remaining (sexually) pure, yet deference/dependency and purity are both dimensions of feminine ideology (Levant et al., 2007). A woman who values both aspects of femininity equally will have difficulty responding to unwanted sexual advances because non-consensual surrender is incompatible with chastity. However, individual women may vary in terms of how much they value these different aspects of traditional femininity. Furthermore, adherence to different types of traditional femininity beliefs may promote different behaviors that affect sexual assault risk. Because many college women experience sexual assault, and because traditional feminine beliefs have been inconsistently related to sexual assault, research is needed to clarify the potential link between feminine beliefs and women’s risk for being assaulted. The current study investigated two specific types of traditional feminine beliefs that may simultaneously increase and decrease women’s risk for sexual assault: deference/dependency and purity.
Traditional feminine deference/dependency (hereafter referred to as deference) reflects the idea that women should depend on a male partner and maintain this relationship, in part, by acquiescing to his goals and desires (Levant et al., 2007). Valuing deference could affect women’s behavior generally as well as during sexual encounters specifically. Compared with their less deferent counterparts, more deferent women may capitulate to unwanted sexual advances by men given their greater comfort with submission than assertiveness. This may be especially true given that many sexually coercive men persist with unwelcome sexual advances even after being told “no” (Struckman-Johnson, Struckman-Johnson, & Anderson, 2003). Although direct sexual refusals do not always prevent sexual assault, strenuous, repeated refusals may deter some perpetrators in some situations (Turchik, Probst, Chau, Nigoff, & Gidycz, 2007). For women who value traditional deference, however, strenuous, repeated refusals violate internalized ideals for self-sacrificial, cooperative behavior.
For these reasons, women’s endorsement of feminine deference was expected to be negatively related to their sexual refusal assertiveness (SRA). Low SRA, in turn, was expected to increase women’s risk for sexual assault. Livingston, Testa, and VanZile-Tamsen (2007) found that women’s low SRA predicted both initial sexual victimization and subsequent revictimization. Specifically, in a college student sample Katz, May, Sörensen, and DelTosta (2010) found that lower SRA predicted revictimization specifically among women in their first year. Although these two studies defined sexual victimization as involving a range of unwanted sexual contact experiences including coerced or forced sexual touching, some of the reported incidents likely involved non-consensual penetration. Overall, previous research is consistent with the hypotheses that beliefs about traditional feminine deference inhibit SRA and, in turn, lower SRA increases risk for sexual assault.
In contrast to deference, women’s endorsement of traditional feminine purity may protect against sexual assault. Purity beliefs converge with fundamental religious traditions that encourage modesty, conservative dress and behaviors, and sexual abstinence until heterosexual marriage (Levant et al., 2007). Feminine purity beliefs also include the notion that women should neither experience primal urges such as the desire for sex nor act on such urges. As such, women invested in feminine purity may avoid situations or encounters that could lead to sexual desire or to sex, both of which might compromise their values. Young women and men alike presume that men in heterosexual relationships will push for sexual activity, including penetrative sex (e.g., Gavey, 2005; Holland, Ramazanoglu, Sharpe, & Thomson, 2004; Tolman, Spencer, Rosen-Reynoso, & Porche, 2003). Because girls expect that they will be pressured for sex, those invested in remaining abstinent (i.e., “sexually pure”) say that they attempt to limit possible opportunities for this to occur; at times, this means avoidance of dating (Tolman, 2002) or ending dating relationships (Tolman et al., 2003). These findings support the hypothesis that an investment in feminine purity may encourage sexual abstinence. In fact, presumably related to an investment in feminine purity, many young women participate in social ceremonies known as “purity balls” in which they collectively promise to remain abstinent before marriage (Fahs, 2010).
In turn, women who remain sexually abstinent may experience a decreased likelihood of encountering or being targeted by a potential perpetrator. This hypothesis is supported by research showing that college women with more sexually conservative attitudes had fewer consensual partners and were less likely to be victims of unwanted sexual contact or penetration (Himelein, 1995). Likewise, this hypothesis also is supported by research documenting that college students with more sexual partners are at greater risk for unwanted, possibly non-consensual sex (Flack et al., 2007). Even outside of college, Testa, VanZile-Tamsen, and Livingston (2007) found that a greater number of sexual partners prospectively predicted women’s subsequent sexual victimization by a non-intimate partner (e.g., stranger/acquaintance/friend). No woman should be blamed for sexual assault, regardless of whether she is sexually active or abstinent. Overall, however, the empirical literature is consistent with the hypothesis that sexual abstinence may protect women in college from sexual assault.
Women who value traditional feminine purity also may be less available as targets for sexual assault because they may avoid other “impure” behaviors associated with risk, such as alcohol consumption. Women’s liberal use of alcohol seems to violate expectations for traditional feminine morality or purity. Peralta (2010) found that participants viewed women who drank heavily as “bad women” who did not deserve or receive the same amount of respect as “good women.” In fact, Peralta concluded that “women perceived to drink to excess were not only considered ‘unlady-like’ but were perceived to be likely targets for sexual assault by the majority of respondents” (p. 399). It may be hypothesized that women invested in feminine purity have internalized these beliefs and thus avoid heavy drinking, although no research to date has examined this possibility. Among college students, most drinking occurs in public during social interactions with others at parties and bars (e.g., Hartford, Weschler, & Sebring, 2002). Given that women’s drinking behaviors may be judged disparagingly, women who drink more frequently may be targeted and thus be at elevated risk for sexual assault. Based on the available research, we expected that women’s frequency of alcohol use will be negatively related to women’s endorsement of traditional feminine purity and positively associated with risk for sexual assault.
Alcohol-related sexual assault is prevalent in college. For example, more than 54% of college women in one sample reported some form of sexual victimization, some of which likely included non-consensual penetration, and 42% were drinking at the time (Ullman, Karabatsos, & Koss, 1999). Conversely, college women who avoid alcohol are at decreased risk for sexual victimization (Howard, Griffin, & Boekeloo, 2008). Research has also shown that women who have experienced attempted rape or rape in college consumed more alcohol than women who reported either no assault or unwanted sexual contact (Benson, Gohm, & Gross, 2007), although it’s unclear whether alcohol use preceded the assault or was a consequence of being assaulted.
Overall, past studies of women’s endorsement of traditional feminine ideology have shown inconsistent results regarding women’s disempowerment and risk for sexual assault. The present research examined different, potentially conflicting aspects of feminine ideology that may affect risk via different behavioral tendencies. More specifically, college women’s sexual assault experiences were investigated as related to their investment in specific traditional feminine ideals of deference and purity. Endorsement of these beliefs was expected to affect different behaviors (SRA, sexual abstinence, and alcohol use) that simultaneously increase and decrease risk for sexual assault. Because the relationships between differing femininity beliefs and sexual assault have not been established in past research, the purpose of this study is to identify risk and protective factors that may inform research and prevention efforts. Women who reported experiencing sexual assault were expected to report greater deference (Hypothesis 1) but less purity (Hypothesis 2) than non-assaulted women. In addition, women’s endorsement of beliefs about traditional feminine deference was expected to be negatively associated with SRA (Hypothesis 3), and SRA was expected to be lower among sexually assaulted women than other women (Hypothesis 4). Furthermore, women’s endorsement of beliefs about traditional feminine purity was expected to be positively related to sexual abstinence (Hypothesis 5), and we expected lower rates of sexual assault among sexually abstinent than non-abstinent women (Hypothesis 6). Similarly, feminine purity beliefs were expected to be negatively related to women’s alcohol use frequency (Hypothesis 7), and more frequent alcohol use was expected among sexually assaulted women compared with other women (Hypothesis 8). These hypothesized relationships are depicted in Figure 1.
Summary of hypothesized associations between specific feminine ideology beliefs and campus sexual assault via assertiveness, abstinence, and alcohol use.
Method
Participants
Data were collected from 254 heterosexual undergraduate women at a small residential public college in Western New York. On average, participants were 19.17 years old (SD = 1.07, range 17-23). Most self-identified as White (85.4%), although participants also identified as Hispanic (5.9%), Asian (4.3%), Black (2.4%), and Other (2.0%). Most participants reported being in their first or second year of college (34.6% and 37.8%, respectively), although participants also reported being in their third (17.7%) and fourth year (9.4%).
Measures
The Femininity Ideology Scale (FIS; Levant et al., 2007) is a 45-item self-report measure of adherence to traditional beliefs about femininity. There are five subscales: Stereotypic Images and Activities (e.g., “Women should have large breasts”), Caretaking (e.g., “An appropriate female occupation is nursing”), Emotionality (e.g., “It is expected that a woman will be viewed as overly emotional”), Dependency/Deference (e.g., “Women should act helpless to attract a man.”), and Purity (e.g., “A woman should not tell dirty jokes.”). Participants indicated their agreement with each item on a 7-point scale (1 = strongly disagree, 7 = strongly agree). A composite score is derived by averaging across all items; higher scores reflect greater adherence to traditional femininity. In addition, the 10-item Dependency/Deference subscale and 9-item Purity subscale were scored separately. In a sample of women in college, the authors reported Cronbach’s alphas of .93 for the composite scale, .83 for dependence/deference, and .85 for purity. In the present study, the internal consistency estimates (Cronbach’s α) were .90 for the composite scale, .66 for deference, and .64 for purity. The authors of the scale provided evidence for reliability and validity.
The Sexual Self-Efficacy Scale (SSE; Rosenthal, Moore, & Flynn, 1991) is a 20-item self-report measure reflecting various sexual activities. Participants were asked, “How certain are you that you would be able to engage in each of the following activities?” The 8-item “Say No” subscale of the SSE was used to assess SRA. A representative item is, “You refuse to do something sexually which you don’t feel comfortable about.” Participants indicated their level of certainty with each item on a 5-point scale (1 = not at all certain, 5 = extremely certain). Across this subscale, items were averaged such that higher scores reflected greater SRA. The authors of the scale reported a Cronbach’s alpha of .75 for the SRA subscale in a sample of women in college. In the current study, Cronbach’s alpha was .71.
Sexual abstinence with male partners was assessed by a single item adapted from Livingston, Buddie, Testa, and VanZile-Tamsen (2004): “At any time prior to today have you consented to sex with a person of the opposite sex? Sex is defined here as any penetration by a male’s penis into a female, regardless of whether he ejaculated.” Participants who responded “no” were identified as sexually abstinent.
Frequency of alcohol use was assessed by a single item adapted from Paul, McManus, and Hayes (2000). The item asked, “In general, since you began college, approximately how often have you consumed alcohol?” Participants rated how often they drink alcohol on a 6-point ordinal scale (1 = never, 2 = once a month or less, 3 = once every two weeks or less, 4 = once a week, 5 = more than once a week, 6 = every day).
Sexual assault was assessed with behaviorally specific items adapted from the Sexual Experiences Survey (SES; Koss & Oros, 1982), similar to Franklin’s (2010) adaptation. Participants responded to three questions that began, “How many times have you had vaginal or oral sex that you did not agree or consent to because . . . ” The items differed in how the statements ended: (a) “the other person used physical force or somehow made you afraid to say no?” (b) “you were incapable of giving consent or resisting due to intoxication?” and (c) “you were overwhelmed by a person’s continual arguments and pressure?” The tactics of sexual assault in these three questions correspond to those assessed by the SES Force, Intoxication, and Sexual Coercion subscales. Behaviorally specific items reflecting completed sexual penetration are both reliable and valid (Testa, VanZile-Tamsen, Livingston, & Koss, 2004) and represent clear sexual violations recognized by the federal authorities (U.S. Department of Justice, Federal Bureau of Investigation, 2012). Sexual assault was scored as a dichotomous variable reflecting either the presence or absence of non-consensual vaginal or oral sexual penetration during college.
Procedure
Undergraduates were recruited online from a voluntary psychology department subject pool for a study of “College Students’ Sexual Experiences.” Data collection sessions were held in campus classrooms. All participants provided informed consent and anonymously responded to self-report surveys administered by 1 to 2 female undergraduate researchers. To ensure privacy, participants were seated in alternating rows. Completed survey materials were submitted to a slotted box. Data collection sessions lasted no longer than 60 min. For compensation, participants were given course credit. A full debriefing was provided. All procedures were approved by the Institutional Review Board.
Results
Based on overall feminine ideology composite scores, women reported low investment in traditional femininity (M = 2.38, SD = 0.59, range 1.00-4.33). In terms of subscale scores, overall, women reported very low levels of feminine deference beliefs (M = 1.37, SD = 0.44; range 1.00-3.50) and somewhat low levels of beliefs in traditional feminine purity (M = 2.69, SD = 0.79; range 1.00-5.33). The mean level of SRA was 4.26 (SD = 0.58; range 2-5), which reflects moderately high assertiveness. About 22% (n = 55) reported sexual abstinence. The average level of alcohol consumption was 3.50 (SD = 1.29; range 1.00-5.50), which corresponds to between “once every two weeks or less” and “once a week,” indicating a moderate frequency. About 15.7% (n = 40) of the sample reported one or more experiences of sexual assault during college.
Two-tailed univariate tests were conducted to compare the demographic characteristics of women as a function of their experiences of college sexual assault. The groups were equivalent in terms of race/ethnicity and dating status, but not age. Specifically, the sexually assaulted group was significantly older (M = 19.57, SD = 1.08) than the non-sexually assaulted group (M = 19.09, SD = 1.06), t(252) = 2.66, p < .01. However, because there were no significant associations between age and any other study variables (see Table 1), age was not controlled for in subsequent analyses.
Zero-Order Correlations Among Study Variables.
Note. Race/ethnicity was self-identified and coded as 1 = White, 0 = racial/ethnic minority. SRA = sexual refusal assertiveness.
p < .05. **p < .01. ***p < .001.
Two-tailed zero-order correlations (Pearson’s r for associations between continuous variables or Spearman’s rho for associations between one or both categorical variables) among all study variables are shown in Table 1. As can be seen, women’s overall investment in feminine ideology was unrelated to sexual assault. Furthermore, overall investment in feminine ideology was negatively associated with SRA, but unrelated to either sexual abstinence or alcohol use. Race/ethnicity was unrelated to study variables except for sexual abstinence.
It was expected that more specific dimensions of feminine ideology (deference and purity) would predict women’s risk for sexual assault. Specifically, we expected that sexually assaulted women would report greater investments in feminine deference (Hypothesis 1) and weaker investments in feminine purity (Hypothesis 2). As shown in Table 1, however, neither of these hypotheses was supported. There was no evidence for a direct relationship between women’s investment in either type of traditional femininity and sexual assault experiences.
In contrast, as shown in Table 1, support was obtained for both Hypotheses 3 and 4. We expected and found a negative association between deference and SRA; more deferential women reported less assertiveness in responding to unwanted sexual advances. In turn, we also expected and found a significant negative association between SRA and sexual assault (see Table 1). Furthermore, an independent-samples t test revealed that sexually assaulted women reported lower levels of SRA (M = 3.95, SD = 0.65) than non-assaulted women (M = 4.31, SD = 0.56), t(252) = −3.69, p < .001. Low SRA increased women’s risk for sexual assault.
We also expected that women’s traditional agreement with feminine purity would be positively related to sexual abstinence (Hypothesis 5), and abstinent women, in turn, would be less likely to report sexual assault than non-abstinent women (Hypothesis 6). Unexpectedly, Hypothesis 5 was not supported. As shown in Table 1, women’s investment in feminine purity was not significantly related to sexual abstinence. Also unexpectedly, sexual abstinence was positively associated with feminine deference; abstinent women reported greater deference than other women. In support of Hypothesis 6, a chi-square analysis showed that only one of the 55 sexually abstinent women (2.5%) reported sexual assault, whereas almost one fifth of non-abstinent women (19.6%, n = 39) reported experiencing assault, χ2(1) = 10.27, p < .01. Sexual abstinence seemed to protect women from sexual assault.
Finally, we expected that feminine purity would be negatively related to women’s frequency of alcohol use (Hypothesis 7), and women who reported experiencing sexual assault would report higher levels of alcohol use than other women (Hypothesis 8). Both hypotheses were supported. As shown in Table 1, feminine purity was negatively associated with women’s alcohol use. Furthermore, an independent-samples t test showed that sexually assaulted women reported greater alcohol use (M = 3.91, SD = 1.12) than non-assaulted women (M = 3.42, SD = 1.30), t(252) = 2.48, p < .05. More frequent alcohol use seems to be a risk factor for sexual assault.
Figure 2 depicts the observed pattern of associations between specific feminine ideology beliefs and risk and protective behaviors related to women’s experiences of sexual assault. Feminine deference was not directly related to experiencing sexual assault. However, deference was negatively associated with SRA and positively associated with sexual abstinence. In turn, both SRA and abstinence were related to experiencing sexual assault, albeit in different directions. Likewise, feminine purity was not directly related to experiencing sexual assault. However, purity was related to less frequent drinking, which, in turn, was negatively related to experiencing sexual assault.
Summary of observed associations between specific feminine ideology beliefs and campus sexual assault via assertiveness, abstinence, and alcohol use.
Finally, we explored which of the different behaviors associated with feminine deference (SRA and sexual abstinence) and feminine purity (alcohol use frequency) uniquely predicted risk for sexual assault. A multiple logistic regression analysis was conducted in which level of SRA, sexual abstinence (1 = yes, 0 = no), and level of alcohol use were entered simultaneously as predictors of sexual assault. Greater SRA (odds ratio [OR] = 0.42, p = .003) and sexual abstinence (OR = 0.09, p = .022) each independently predicted lower odds for sexual assault, χ2(3) = 26.15, p < .001, Nagelkerke R2 = .17. In contrast, alcohol use was not significantly associated with sexual assault independent of SRA and abstinence.
Discussion
The present study investigated college women’s investment in traditional feminine ideology, behavioral risk factors, and experiences of sexual assault. Sexual assault was specifically defined in terms of non-consensual vaginal or oral penetration. We found no direct relationship between women’s endorsement of traditional feminine ideology and sexual assault. Overall, more traditional women were not at increased risk for being assaulted. Past research (e.g., Faulkner et al., 2008) also reported no relationship between women’s general feminine ideology and response to hypothetical risk, and the current study replicated and extended this null result in a study of women’s actual experiences.
The present study also extended previous research by separately investigating women’s investments in specific types of traditional femininity. Deference and purity, two potentially conflicting aspects of feminine ideology, were hypothesized to affect women’s risk for sexual assault in different ways. Although neither feminine deference nor feminine purity directly predicted sexual assault, we found evidence that both were indirectly associated with sexual assault via associations with different behavioral tendencies. Specifically, beliefs about deference were associated with decreased SRA as well as sexual abstinence, whereas beliefs about purity were associated with less frequent alcohol use. In contrast, women’s overall feminine ideology scores were associated only with SRA. Compared with overall feminine ideology, specific dimensions were more strongly linked to behavioral tendencies that predicted sexual assault experiences.
As expected, results showed that women invested in feminine deference were less assertive in refusing unwanted advances. Being less assertive, in turn, appears to increase risk; sexually assaulted women reported less SRA than other women. These findings are consistent with past research (e.g., Katz et al., 2010; Livingston et al., 2007) showing that SRA is specifically related to sexual victimization defined broadly as including both unwanted touching and penetration. The overall pattern of findings showed that investment in traditional feminine deference appears to indirectly increase women’s risk for sexual assault by discouraging SRA.
Results also showed that women invested in feminine purity consumed alcohol less often. The observed negative relationship between purity and alcohol use extends Peralta’s (2010) findings that others negatively evaluate women as immoral for heavy or frequent drinking; women themselves who are invested in feminine purity also drink less, probably to avoid similarly negative self-evaluations. Reduced drinking, in turn, was a protective factor; we found that sexually assaulted women drank more frequently than other women. These findings are consistent with previous research. Because sexual assaults among college students commonly involve alcohol use by both the victim and perpetrator (e.g., Benson et al., 2007; Howard et al., 2008; Ullman et al., 1999), drinking less decreases risk. Overall, we found that an investment in feminine purity is indirectly linked to sexual assault through frequency of alcohol consumption.
Unexpectedly, in the present study, women’s investment in purity was unrelated to sexual abstinence. Although abstinence has not been studied in relation to feminine purity in past research, feminine purity was associated with women’s avoidance of non-traditional behaviors such as masturbation and casual sex (Levant, Rankin, Hall, Smalley, & Williams, 2012). Taken together, these results suggest that some traditional forms of partnered sex may be compatible with valuing feminine purity. For example, women invested in traditional purity may tolerate or even embrace opportunities for vaginal sex with a monogamous male partner in response to his expressions of desire to express her love for him. This possibility is consistent with Carpenter’s (2002) finding that some traditionally feminine women described their first experience of sex in terms of offering that partner “a gift.” Because null results are ambiguous, however, additional research is needed before concluding that women’s sexual abstinence is unrelated to an investment in feminine purity.
Also unexpectedly, we found that women’s sexual abstinence was positively associated with valuing feminine deference. Because abstinence was related to deference, not purity, it may be speculated that the decision to remain sexually abstinent may be more strongly related to adhering to the wishes of authority figures than to internalized standards for purity. For example, some deferent women may have participated in “purity balls” or made “chastity pledges” in which they publicly promise to avoid sex before marriage (Fahs, 2010). Remaining abstinent may represent women’s deference to the values of parents, clergy, or other authority figures, although future research is needed before definitive conclusions can be drawn.
Women’s sexual abstinence also was associated with greater SRA. These unexpected results suggest several intriguing possibilities that warrant further study. First, SRA may be expressed differently by sexually active versus abstinent women. Sexually active women may be more deferential specifically regarding penetrative sex than abstinent women, whose deference may be mitigated by low sexual permissiveness, fears of pregnancy or sexually transmitted infections (STIs), or other factors. Second, sexually abstinent women’s refusals may be afforded greater respect. Some men feel entitled to ignore a woman’s sexual refusals or exert coercive pressure for sex if the woman has previously consented to sex with them (Seal, O’Sullivan, & Ehrhardt, 2007) or with someone else (Bay-Cheng & Eliseo-Arras, 2008). A third, related possibility is that sexually abstinent women may avoid socializing with others who do not value abstinence, enabling them to feel assertive without needing to enact assertiveness. For example, sexually abstinent women may avoid alcohol-saturated college parties that serve as preludes to casual sex (Flack et al., 2007); this hypothesis is consistent with the observed negative association between abstinence and frequency of alcohol use. At any rate, as expected, abstinence was a protective factor; we found lower rates of sexual assault among sexually abstinent than active women.
The overall pattern of results suggests that feminine deference specifically is associated with behaviors that, in turn, are directly related to experiencing sexual assault. At the same time that feminine deference discourages SRA (a risk factor), it encourages sexual abstinence (a protective factor). Despite the observed association between SRA and abstinence, in multivariate analyses, SRA and sexual abstinence both uniquely and independently predicted lower odds for sexual assault. In other words, these two behavioral tendencies associated with deference simultaneously predicted risk, but in opposing directions. In contrast, alcohol use was not related to risk for sexual assault beyond the effects of SRA and abstinence.
Limitations of the present study should be acknowledged and addressed in future research. Because data were collected from a primarily White convenience sample of heterosexual college women, the current findings may not generalize to more diverse groups. Women reported extremely low levels of feminine deference and purity, and the subscale estimates of internal consistency were low. It is notable that despite this restricted range of scores and relatively low internal consistency, significant associations with risk behaviors were observed. Future research is needed before the ambiguous null relationships observed here can be confidently interpreted. For example, in geographic regions other than the Northeastern United States, college women might report greater investment in purity or deference and different patterns of results would emerge. Studies employing more comprehensive measures of alcohol use might also yield more robust results. Another limitation of the current study was the relatively restrictive definition of sexual assault. We did not assess assault due to incapacitation other than that created by substance use. Furthermore, we did not assess experiences of non-consensual anal penetration. It is possible that some women may have been misidentified as non-victims if they experienced sexual assault due to medical or psychological incapacitation or non-consensual anal penetration but not oral or vaginal penetration. It is encouraging that rates of completed sexual assault in the present study are similar to rates in studies that assess completed anal as well as vaginal or oral penetration (Krebs et al., 2007). Nonetheless, the current results should be interpreted with these limitations in mind.
Future studies should evaluate the social and relational contexts of SRA, sexual abstinence, and alcohol use with more comprehensive assessment tools. Future studies should also investigate different types of sexual assault (e.g., verbally coerced, intoxicated, or forcible assault, see Brown, Testa, & Messman-Moore, 2009) enacted by different perpetrators (e.g., intimate vs. non-intimate partners). It may be hypothesized that more deferential women are at particular risk for intimate partner verbal coercion because investment in an intimate relationship may heighten pre-existing values in feminine deference. Finally, by assessing women’s beliefs and experiences cross-sectionally, SRA and alcohol use might precede or follow sexual assault experiences. In fact, past research suggests reciprocal relationships between women’s sexual victimization and both SRA (Livingston et al., 2007) and alcohol use (Gidycz et al., 2007). In contrast, sexual abstinence seems clearly protective, although mechanisms for this effect should be evaluated. Longitudinal studies are needed to investigate prospective risk for sexual assault as a function of feminine deference and associated behavioral tendencies.
Despite these limitations, this present study suggests that only some traditionally feminine women are at elevated risk for sexual assault in college: non-abstinent women who exhibit lower SRA. More generally, our results help reconcile inconsistent associations between general feminine ideology and women’s sexual disempowerment in previous research. Stronger associations with risk are obtained when specific types of feminine ideology, namely, feminine deference, are investigated. By addressing risk and protective factors associated with traditional feminine deference, sexual assault prevention efforts may be more successful in reducing women’s risk for sexual assault on college campuses.
Footnotes
Acknowledgements
The authors thank Catherine Herman and Melanie Schukrafft for collecting and entering the data discussed in this article.
Authors’ Note
Portions of these data were presented at the 46th annual meeting of the Association for Behavioral and Cognitive Therapies.
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.