THROUGH HER EYES: Factors Affecting Women's Perception of and Resestance to Acquaintance Sexual Aggression Threat

Abstract

A major component of a woman's ability to resist assaults by strangers versus acquaintances lies in the social and cognitive context in which she is engaged with the perpetrator and within which she must recognize potential threat before engaging in a behavioral response. This paper presents questionnaire and focus group findings of heterosexual college sorority women's social contexts, perceived risks, responses, and psychological barriers to protecting themselves from sexual aggression threat by fraternity acquaintances. Several social and cognitive factors, including alcohol consumption and psychological barriers, were related to projected responses to sexual aggression. Participants in general held a high sense of invulnerability to victimization and an optimistic belief in their ability to resist sexual aggression. Several differences between previously victimized and nonvictimized women also emerged.

Greater awareness of the high prevalence (Kilpatrick et al., 1985; Koss, Gidycz & Wisniewski, 1987; Muehlenhard & Linton, 1987) and negative impact (Koss, 1993; Koss & Burkhart, 1989) of sexual aggression perpetrated by acquaintances points to the importance of rape prevention efforts. Key components of rape prevention include early detection of risk and an effective defense. The purpose of the study reported in this paper was to identify factors that can affect a woman's ability to perceive potential danger in a social situation and to respond effectively to it. The responsibility for sexual assault lies completely with the assailant. However, we recognize that for the foreseeable future women will encounter sexually aggressive men. Our goal in understanding how women perceive risk is to support women's ability to anticipate, detect, and take self-protective measures against risk of sexual aggression, particularly within the context of social situations and relationships that would incline them to anticipate safety.

FACTORS ASSOCIATED WITH DETECTING AND RESPONDING TO RISK

Alcohol consumption by a victim, an assailant, or both is known to increase the risk for sexual assault (Koss et al., 1987; Muehlenhard & Linton, 1987). An important component of this relationship is the way that women view how alcohol affects their risk. This view can affect the measures that women take to protect themselves. For instance, if a woman perceives that her drinking alcohol increases risk primarily through its physically debilitating effects, then she might be most receptive to prevention efforts that promote moderate drinking. Of course, alcohol can increase risk for sexual assault both through its physical effects and through psychological mechanisms (see Abbey, Ross, & McDuffie, 1994; Norris, 1994 for reviews). Physical effects occur when cognitive and motor abilities are impaired as blood alcohol level increases, and a woman becomes physically less able either to perceive risk or to respond to it.

Psychologically, alcohol might operate as a barrier to perceiving and responding to risk. Because of the social stigma attached to a drunken woman (Fillmore, 1984; Van Amberg, 1943), as well as the cultural view, ascribed to by both men and women, that she is sexually available (George, Gournic, & McAfee, 1988), a woman drinking socially might be less likely than a nondrinking woman to call attention to a man's unwanted sexual advances. That is, she might be embarrassed by her own drunkenness and blame herself for inviting these advances. Such feelings and self-perceptions can serve to delay or mitigate assertive selfprotection. ¹

Besides alcohol, other psychological barriers to resisting sexual assault have not received a great deal of attention. Although it is known that physical resistance decreases the likelihood of a completed acquaintance rape (Levine-MacCombie & Koss, 1986; Murnen, Perot, & Byrne, 1989), not all women feel comfortable being physically, or even verbally, assertive. Relating socially to a man, especially if the woman regards him as a friend or potential mate, could place her in conflict about responding assertively to his sexual advances. For instance, if she were to misinterpret his actions, she might become embarrassed or fear his rejection of her. The goal in this study was to explore the extent to which these possible psychological barriers—perceived effects of alcohol consumption, embarrassment, and fear of rejection by the man—were associated with the use of verbal and physical assertiveness, as well as with projected use of indirect methods of resistance, such as joking with or distracting the man, in college sorority women. We hypothesized that these psychological barriers would be negatively correlated with projected verbal assertiveness and physical resistance and positively correlated with projected use of indirect forms of resistance.

In addition, several studies have found that a substantial number of rape victims were previously victimized (see Browne & Finkelhor, 1986 for a revie). However, a history of victimization does not directlvpredict victimization as an adult (Atkeson, Calhoun, & Morris, 1989; Mandoki & Burkhart, 1989). For example, earlier victimization was found to be related to an increased number of sex partners, which in turn predicted adult sexual victimization (Mandoki & Burkhart, 1989). This suggests that effects of prior victimization on victims' responses to sexual assault may be indirect (e.g., through engagement in behaviors and situations associated with high risk). Understanding differences between women who have and have not previously experienced sexual victimization can lead to developing interventions to lower the risk of future victimization. These interventions are of particular relevance to those who have previously encountered sexual aggression. We cautiously hypothesized, following Mandoki and Burkhart (1989), that prior victimization would be related to a higher number of sex partners and greater alcohol consumption, as well as greater endorsement of psychological barriers and greater projected use of indirect resistance, but lower projected use of verbal assertiveness and physical resistance. However, we recognized that itwould not be possible to disentangle whether these effects would have occurred prior to the victimization or resulted from it.

The social context is recognized as an important component in making judgments about risk and appropriate responses to it. To the extent that a woman feels comfortable in her social surroundings she may feel highly indisposed toward anticipating or interpreting a man's actions toward her as threatening. The present study was conducted among members of sororities. The Greek system consitutes a historically stable social system with many aspects that increase feelings of comfort and conformity among its members: established charters and bylaws, longstanding traditions involving highly scripted events and family-like referents, for example, "brothers" and "sisters," degrees of relatedness among specific fraternity and sorority houses, and social and economic similarity among members (Larimer, 1992; Moffatt, 1989; O'Sullivan, 1991). However, alcohol consumption has been found to be high within the Greek system and to be part of a tradition of socializing. Findings indicate that a strong predictor of heavy drinking among this population is the belief that alcohol makes it easier to act out sexually (Baer, Kivlahan, & Marlatt, 1995). Thus, the risk for sexual assault may be commensurately higher, although research findings are mixed on this point (Copenhaver & Grauerholz, 1991; Kalof, 1993; Kirkpatrick & Kanin, 1957; Rivera & Regoli, 1987). In any case, sorority members are at no lower risk for sexual victimization than other female college students. Because of the familial bondings provided by their social situation, they may feel at lower risk than other college women. Focusing on this group provided an important opportunity to examine women's perception of risk and responses to sexual aggression.

METHOD

Research Participants and Recruitment

Research participants consisted of 66 women, predominantly Caucasian (84.6%), between the ages of 18 and 21 (M = 19.2 years, SD = .91), drawn from 10 sorority chapters at a large West coast university. Fifty percent of the participants were in their first year of college, 36 % were sophomores, and 11% were seniors. For the purpose of this study, only currently "single" women were included; the majority (67%) indicated that they had experienced consensual sexual intercourse with men.

This project began after the investigators were approached by student leaders from an acquaintance rape education and prevention committee of the Panhellenic Association. In collaboration with them, we developed the aims and design for the study, in addition to a strategy that would encourage participation by other sorority members. Before participant recruitment, formal approval for the study was obtained from the Panhellenic Council, the umbrella organization that provides leadership and unity to the campus' 18 separate sororities, as well as from its advisor. Ten sorority houses were randomly selected from the 17 chapters on campus. Each sorority chapter president was then individually approached and given an overview of the proposed study for purposes of gaining permission to allow recruitment of research participants from her chapter. Only one chapter declined permission because of the sorority's national policies.

Participant recruitment was undertaken through informational announcements at the sororities' weekly business meeting by either an investigator or the house's rape prevention student leader. These announcements included a description of the study's purpose, procedures to be used, and assurances of confidentiality to protect both participants and the chapter as a whole. Participation was voluntary, and participants received $10.

Procedures and Measures

Overview

Each of eight groups was comprised of between four and seven women representing between two and four separate sororities. We reasoned that this combination would minimize the likelihood of participants only providing the chapter's "party line" by having different points of view offered by women from different houses and allowing some degree of personal comfort by allowing each participant to attend with a friend. All sessions were conducted by one of the three female investigators. Participants first completed a self-report questionnaire that took approximately 20 minutes and assessed demographic characteristics, alcohol consumption patterns, dating and sexual experiences, and attitudes and experiences with sexual aggression. The subsequent focus groups lasted 90 minutes each and consisted of a standard series of nine questions that provided a general structure. During debriefing, participants were encouraged to contact the investigators with future questions and concerns, and a list of sexual assault resources was provided in case the discussion triggered any later distress or interest in further information or assistance.

Questionnaire

Five sets of measures were used in the self-report questionnaire. Prior victimization was based on yes answers to at least one of the following items derived from the Sexual Experiences Survey (Koss & Gidycz, 1985) during the past year: (a) been in a situation where a man became so sexually aroused that you felt it was useless to stop him even though you did not want to have sexual intercourse? (b) had sexual intercourse with a man when you didn't really want to because you felt pressured by his continual arguments? (c) been in a situation where a man used some degree of physical force to get you to have sexual intercourse with him when you didn't want to, whether or not it actually occurred? (d) had a man attempt sexual intercourse with you by giving you alcohol or other drugs, but intercourse did not occur? (e) had sexual intercourse when you didn't want to because a man gave you alcohol or other drugs?

Perceived risk of sexual victimization was assessed through five items, derived from the Sexual Experiences Survey (Koss & Gidycz, 1985), concerning respondents' perceived likelihood of encountering unwanted sexual intercourse: (a) you feel it would be useless to stop him because he was so sexually aroused; (b) you would feel pressured by continual arguments; (c) he would threaten to use physical force; (d) he would use physical force; ( e) he would give you alcohol or other drugs. These items were answered on 1-7 scales, ranging from "extremely unlikely" to "extremely likely" and created a mean-based scale (α = .90).

A modified version of the Daily Drinking Questionnaire (Collins, Parks, & Marlatt, 1985) was used to assess the typical pattern of alcohol consumption in the past 30 days. Participants reported how many of five types of social events they attend in a typical month; how many drinks they consume and over what time period. They also reported the most they drank on any particular occasion, and the average number of drinks typically consumed on a daily basis in a given week. Average and peak blood alcohol level in a typical week in the past 30 days was calculated using a mathematical formula based on the participant's weight, sex, the total number of drinks and hours spent drinking for each day of the week (Mathews & Miller, 1979).

Resistance to unwanted sexual advances was based on participant ratings of 18 possible defensive strategies, by means of 5-point scales (0-4; "not at all" to "highly likely"), indicating how likely they would be to employ each in a situation where they were being pressured by an acquaintance to engage in sexual activities when they did not want to. Respondents were asked to think about a typical social situation such as a fraternity party or exchange. As an example, a hypothetical situation was provided in which a man takes her on "a house tour that ends in an empty bedroom." After engaging in consensual kissing, he pressures her to engage in sexual intercourse, and she wants to stop. Defense strategies were organized into three mean-based subscales: (a) gentle or indirect messages (8 items, e.g., "Jokingly try to tell him that he is coming on too strong"; α = .57); (b) verbal assertiveness (5 items, e.g., "Raise your voice and use stronger language, e.g., Hey. LISTEN! I really mean it"; α = .87); and (c) physical resistance (4 items, e.g., hitting, kicking, scratching, or openly seeking help from others; α = .83). One item was dropped because of its low correlation with other items.

Psychological barriers to resisting unwanted sexual advances were based on 21 items providing reasons that would make it difficult for respondents to remove themselves from a situation in which they were being pressured to engage in sexual activities that they did not want. As in the previous section of the questionnaire pertaining to resistance responses, participants were asked to think about a typical social situation, such as a fraternity party, and were presented with the same scenario described above concerning being taken on a house tour. Responses were rated on 1-5 scales from "not significant" to "extremely significant" and organized into three mean-based subscales: (a) concern about embarrassment (6 itmes, e.g., "I wouldn't want to embarrass myself by screaming out loud"; α = .82); (b) fear of rejection by the man (5 items, e.g., "I might like him and wouldn't want to ruin things for the future"; α = .77); and (c) disabling effects of alcohol consumption (3 items, e.g., "I might be too intoxicated to think through a plan to get out of the situation"; α = .81). Seven items were dropped because of their low correlations with other items.

Focus group

Because little is empirically known about how sorority women perceive the threat of acquaintance sexual aggression and how they guard against it, we elected to use focus groups as a way of both understanding these respondents' experiences and for developing hypotheses that could be tested in future studies. Focus groups have been used frequently as a method of gathering qualitative data to gain valuable insights into how people think about and understand health behaviors, or to research relatively underinvestigated fields for hypothesis development and questionnaire construction (Krueger, 1994; Morgan, 1993; O'Brien, 1993). Within feminist research, focus groups have been used to provide a "voice" to the research participant by giving her an opportunity to define what is relevant and important to understand her experience (Jayaratne & Stewart, 1991). Development of the focus groups and analysis of the transcripts were informed by methodological literature (Chenitz & Swanson, 1986; Krueger, 1994; Morgan, 1993; O'Brien, 1993; Strauss & Corbin, 1990). Quotes from focus groups included here are used to illustrate common themes that emerged from discussions, and to reinforce findings from the questionnaire data. In order to ensure that quotes selected from focus group discussions were truly representative, careful analysis of focus group material was undertaken.

Participants were asked to honor an agreement of privacy and confidentiality by not discussing the contents of the group discussion or the participants. Because of the sensitivity of the topic and our awareness that some of our participants had probably been victims of acquaintance sexual aggression and might not want to make this disclosure in the group, we gave participants an explanation of what we were seeking that acknowledged varying experiences and feelings.

The focus groups began with an "ice-breaker" (e.g., "What did you expect Greek life would be like before you joined?"). This allowed the interviewer and group members to become acquainted and more comfortable with each other. It also provided the interviewer an opportunity to gauge the group dynamic (e.g., who talks most, least; who is willing to express socially undesirable remarks about the Greek system, etc.) to increase the probability that all members would have an equal opportunity to speak (Krueger, 1988). The remaining questions served to guide the groups through a "typical" social occasion where sexual aggression might occur. More specifically, the questions began by asking participants to describe the expectations and concerns that women typically have prior to an event; the questions concluded by asking them to describe the likely aftermath of acquaintance rape (see Appendix for summary). To ensure content validity of the data provided, group members were provided numerous prompts through,;mt the groups to comment on, clarify, amplify, or disagree with comments by others. Group facilitators sought to summarize themes during the discussion to ensure accurate understanding of the discussion topics, and by asking open-ended prompts, such as "What else?" (Swanson-Kauffman, 1986).

Focus groups were tape-recorded without reference to identities and a nonparticipating female research assistant took notes. Recordings of the groups were transcribed verbatim. To ensure accuracy, portions of randomly selected tapes were reviewed by a separate coder. Using a technique described by Corbin (1986a,b), the transcripts were formatted as follows: (a) Transcripts were printed in landscape fashion on legal-sized paper. (b) Four columns were created for the text, an abstract, description of particular facts or incidents, and investigator's comments. (c) Each new paragraph represented a separate speaker. (d) A code for comments or questions by the interviewer was made to the left of the new paragraph and a different code was used to signify comments from participants. (e) All responses were assigned a code-response number. Transcripts from the focus groups were then coded for thematic content and prominence of themes across groups.

As the first step in extracting themes from the transcripts, research assistants read through the text and highlighted all key words and phrases for each response that accurately represented the original intent of the passage. Following the full highlighting of the text, a brief summary abstract was recorded in the adjacent column that kept close to the original data. Salient facts or incidents that illustrated the social interactions between women and men in the Greek system were recorded in the next column. Finally, once the text had been fully summarized, interpretations by the research assistants and investigators were recorded in the final column. Interpretations were formed by weaving together various responses into themes. To ensure the reliability of the theme summaries, coded transcripts were reviewed by a second reviewer affiliated with the project.

Five themes that originated from either the focus groups or questionnaire were closely analyzed for purposes of this study, and include: (a) the social context in which sexual aggression occurs, (b) perception of risk, (c) alcohol effects, (d) psychological barriers to resistance, and (e) the influence of prior victimization. To aid the process of data reduction, three of the original eight transcripts were randomly selected to assess the prominence of these themes across groups. Transcripts were again reviewed, and each response that captured one of the five themes was assigned a corresponding code. Percent agreement between raters of the five themes ranged from 64% to 100% with an overall mean of 81 % . Categorical codes were then counted for each group and compared to other groups. As a final step to ensure construct validity, an independent consultant knowledgeable in the field of sexual aggression research reviewed one of the three coded transcripts to assess the extent to which the transcript's content was reflected in the presentation of the findings.

RESULTS

Social Context

Questionnaire responses indicated that over the preceding month, respondents had attended an average of 11.49 combined sorority-fraternity social events (range = 4-28, SD = 4.89), the most popular described as parties and exchanges at fraternity houses. Focus group discussions elaborated several consistent factors regarding these social events. The first had to do with their situational characteristics. Peer norms at these events are to drink considerable quantities of alcohol, to "get wild," and possibly to "mash" with someone during the evening (mashing includes a wide range of sexual activities, exclusive of sexual intercourse). It is common for men to invite women to other parts of the house for a variety of legitimate and contrived reasons, including to their bedrooms where alcoholic beverages are stored. While sororities may advise their members of the risks of "separating from the group," it nonetheless occurs commonly, if not predictably.

Part of the concept behind events such as exchanges is to regularly provide institutionally approved social forums within which members from selected sorority and fraternity houses can meet and get to know one another. There is a general expectation that, within such activities, sorority members may well meet future husbands; hence, the members see these activities as times to "scope out" men, both to date in the short term and to consider for marriage later on. Although respondents reported becoming more comfortable over time, many expressed shock on initial exposure to the "anything goes" attitude during social events and to the intensity of the "party now before real life begins" ethic.

It's weird because it's like you are sheltered in the Greek system. It's okay to mash, but yet you feel kind of weird. If you went to a movie and mashed on some guy next to you … that wasn't in the Greek system, you know, or you went to a dance club or something, then … I think people would judge you differently. But it's more accepted (in the Greek system): You go to a party and you mash.

It was like a slap in the face, like you know, a wake up to reality. This isn't like … the movies; it's a lot more intense. It's a lot different than high school was, and the attitudes are just totally different. At the parties, it's like, well, "Who cares," you know, an "anything goes" type of attitude.

In summary, it is clear that integration into "Greek life" is associated with a high degree of socializing, which ideally leads to pairing off and to some degree of sexual activity. Although these activities are in and of themselves somewhat normative aspects of college life, they can be associated with high-risk activities, such as alcohol consumption, thus setting the stage for incidents of sexual aggression. In the context of the sororityfraternity system, wherein a woman feels secure among her "sisters" and "brothers," she may not perceive that her risk for being victimized is at least as great as in the rest of the university setting. These points are illustrated below.

Perceptions of Risk and Resistance

Questionnaire responses indicated that the respondents estimated their future risk of encountering sexual aggression in dating as quite unlikely (M = 2.27, SD = 1.33). In envisioning a social situation wherein they would encounter sexual aggression, respondents reported a moderate likelihood of using indirect or gentle messages (M = 1.96, SD = .62) and verbal assertiveness (M = 1.95, SD = 1.07), and a low likelihood of using physical resistance (M = 1.02, SD = 1.01).

Focus group discussions produced several discrepant themes regarding perception of risk and response to a risky situation for themselves versus for their peers. On one hand, respondents conveyed a relatively high degree of awareness of the general risk for sexual aggression and of prevention measures taken to protect other women. This included behavioral specificity about ways that senior sorority members watched for danger signs regarding more junior members (e.g., drinking too much, getting approached by men with "reputations"), as well as strategies for assisting a sister member to avoid or to get out of a potentially dangerous situation (such as buddy systems and hand signals used to surreptitiously request or direct help).

In contrast, respondents were strikingly less specific about how they might anticipate dangerous situations and protect themselves against an assault. The typical response was an assertion that one would not be "dumb enough" to get into a risky situation in the first place. They were "too smart to be raped" (e.g., that the respondent is a good judge of character and would be able to tell if any particular man was a rapist); if threatened they would "just get up and leave." Some recognized that it may not be wise to believe "it just won't happen to me," yet fundamentally, respondents did not harbor concerns for themselves in these situations.

I'm still worried about my sisters. I know everyone thinks, "it won't happen tome."

I'm not more scared of rape at fraternities and stuff. I'm more scared of rape when I'm walking down the street. And some guy, not in a fraternity, grabs me or something.

For myself, I mean, this is gonna sound really fame, but I never really considered myself the type that gets date-raped. I know that's bad because I'm sure a lot of people agree with me, that they think, "Oh, I'm not gonna; that's not going to happen to me," and I know it's terrible to think that because I'm smarter than that, even though I know it probably could. When I go into an exchange or party, I don't think to myself, "I hope I don't get AIDS tonight," or "I hope I don't get date raped tonight." I think, "I hope that I get a chance to meet people."

Alcohol Effects

Alcohol consumption was reported on the questionnaires as being an important part of socializing for most respondents, although roughly 20 % reported abstinence from drinking as a stable pattern. For those who did drink, a generally high level of alcohol was typically consumed on social nights. Approximately half of all respondents had estimated blood alcohol levels (BALs) of .10 (the local legal level of drunkenness for driving) or greater in their preferred social situations, typically fraternity parties. During focus group discussions respondents viewed alcohol as an inherent part of Greek social life, a social lubricant, a convenient topic for conversation, and an excuse for doing things that they might not normally do.

At the Greek (system) parties, you can get away with being really dumb and stupid when you drink. But (at a bar ouside the Greek system), no matter how wasted you are you have to pretend … you have to act a certain way. But at the Greek parties, everyone's getting … acting really dumb.

The role of alcohol as a risk factor for sexual victimization was supported by questionnaire data for the sample as a whole. Specifically, a higher estimated peak BAL in social situations was significantly correlated with number of recent sex partners (r (62) = .31, p = .01), with the likelihood of embarrassment (r (64) = .26, p = .05), and with alcohol effects (r (64) = .42, p = .0001) as barriers to removing themselves from the situation. Although estimated BAL was not significantly correlated with indirect resistance, it was inversely correlated with physical resistance (r (63) = −.35, p = .01) and with verbal assertiveness (r (64) = −.28, p = .01).

The focus group responses were mixed with respect to alcohol as a risk factor. On the one hand, most discussants agreed that alcohol consumption makes it difficult to discriminate between risky and non-risky cues, and to take needed action to reduce risk of sexual aggression when the woman is intoxicated. They agreed that women who are drinking will likely be seen as more sexually available and that a woman who is intoxi cated would more likely be blamed for being raped. On the other hand, respondents uniformly saw themselves as being able tu i.tay alert and in control in spite of their alcohol consumption.

I tend to be a pretty heavy drinker when I go out. I usually try to anticipate how much I'm going to drink so I don't get too out of control.

I think for me the big concern is AIDS. I think that's more on top than rape. And maybe I'm just trying to be a fearless person, but I honestly think … and who knows, that honestly don't think it will happen to me just because Inever drink to get drunk, just to get belligerently drunk. I always know what I'm doing all the time.

Psychological Barriers to Resistance

In considering psychological barriers lo removing themselves from the situation, questionnaire respondents identified concern about embarrassment (M = 2.16, SD = .89), rejection by the man (M = 1.94, SD = .76), and disabling effects of alcohol consumption (M = 2.03, SD = .99) as having modest significance in pceventing them from removing themselves from the situation. The potential influence of these barriers on behavioral efforts to resist sexual aggression is suggested by correlational analyses. All three types of psychological barriers were positively correlated with indirect resistance and negatively correlated with verbal assertiveness and physical resistance (see Table 1). Although concern regarding the potential for alcohol consumption to impair one's defense efforts correlated less strongly with resistance strategies relative to concern regarding rejection or embarrassment, it was significantly correlated with predictions of the future likelihood of encountering sexual aggression, r (64) = .41, p = .0001, and with concern about embarrassment r (66) = .47, p = .0001) and about rejection (r (66) = .45, p = .0001). Collectively, these findings provide support for hypotheses concerning the relationships between psychological barriers and methods of resistance.

Table 1.

Correlations between psychological barriers to resistance and resistance strategies

	Resistance Strategy
Psychological Barriers	Indirect	Verbal Assertiveness	Direct
Embarrassment	.46***	−.39***	−.30**
Fear of rejection	.44***	−.36**	−.21*
Alcohol incapacitation	.22*	−.21*	−.12

Note: N = 66.

^*p = .05.

^**p = .01.

^***p = .001.

These relations among participants' concerns and their responses to sexual aggression were mirrored in focus group findings. Although fear about being physically hurt in the process of resisting was expressed, far more prominent were expectations of feeling conflicted about being embarrassed or offending the man, fears that because the man is bigger, the woman would not be able to get him to stop, and expectations of feeling cognitively flooded and emotionally overwhelmed in the moment, leading to freezing and confusion. There was a distinct disinclination to use defense methods like yelling, screaming, or physically fighting for fear of drawing the attention of others or of angering the man.

I think people start feeling really uncomfortable, and start to worry, "What am I going to do? How am I going to let him know that he's going too far?" And you're also thinking, "I don't want to feel dumb. I don't, you know, want him to tell his friends, or start calling me a prude, or whatever the case might be …. So, it's kind of like … you are going over in your mind what you should do."

If you possibly like the guy, you don't want him to be just totally offended. You know, you might jokingly say "Stop" or whatever. I think everything is just flooding through your head at the same time. You know, "How do you feel about this person?" and "What's going to happen?" … It's just flooding through all at once.

I think I might be angry … if it (rape) were happening to me. I'd be, like, "Oh my Codi This is actually happening to me. This is never supposed to happen to me." And I'd be frantic, like if a guy were to use physical force. I'm not exactly a scrawny person, but I know that if a guy's bigger than me, he's definitely stronger than me. And if he were to use it, I would not know what to do. I'd probably go into shock.

Finally, a system-specific barrier emerged. That is, respondents noted concerns about implicating their chapter or that of a favored fraternity in charges of date rape. Terms such as "Greek bashing" were used to depict what they saw as negative press about the Greek system, and these women expressed a sense of responsibility to protect the reputation of their houses.

Prior Victimization

Questionnaire data showed several significant differences between those who had experienced one or more forms of sexual aggression by an acquaintance within the previous year (n = 30) and those who had not (n = 36). As shown in Table 2, those previously victimized reported significantly higher estimates of encountering future sexual aggression, significantly higher likelihood of using indirect methods of resisting and significantly lower likelihood uf using verbal assertiveness and physical resistance. Previously victimized respumlents also reported significantly higher likelihood that embarrassment, fear of being rejected, and effects of alcohol consumption would pose barriers to removing themselves from the threatening situation, as well as higher peak blood alcohol levels and number of recent sex partners.

Table 2.

Mean differences in victim’s versus non-victims’ risk perception, resistance strategies, psychological barriers, and risk factors

Dependent Measures	Victims (SD) (n = 30)		Non-victims (SD) (n = 34)		t
Likelihood of future victimi-  zation	2.91	(1.53)	1.65	(.74)	4.11’**
Resistance strategy
Indirect resistance	2.15	(.53)	1.79	(.64)	2.44**
Verbal assertiveness	1.57	(.85)	2.33	(1.14)	3.00***
Physical resistance	.74	(.70)	1.31	(1.17)	2.31**
Psychological barriers
Embarrassment	2.54	(.89)	1.82	(.76)	3.51***
Fear of rejection	2.15	(.80)	1.75	(.71)	2.11*
Alcohol effects	2.47	(1.04)	1.68	(.79)	3.44***
Computed peak blood alcohol  level	.10	(.06)	.07	(.07)	1.94*
Number of recent sex partners	1.30	(1.26)	.70	(1.18)	1.95*

^*p = .05.

^**p = .01.

^***p = .001.

Because of the way that the study had been described to participants, we did not directly ask participants in the focus group sessions about their personal prior victimizations. Athough participants occassionally made subtle references to "things in their background" and prior "bad personal experiences" in relation to acquaintance sexual aggression, direct discussion of individuals' prior victimization did not spontaneously arise. In response tu opportunities presented to speak in general terms, respondents spoke of fraternity houses wherein sorority women who had encountered bad experiences had acquired "tainted" reputations. However, respondenb also noted the difficulty of sharing information about experiences of aggression, particularly from house to house, and of their reluctance to risk negative attention to their own sorority house's reputation.

DISCUSSION

This study investigated relationships among sorority women's perception of risk for sexual assault, psychological barriers that may impede effective resistance, and the projected use of direct (that is, verbal assertiveness and physical resistance) and indirect resistance strategies. The women in this study perceived themselves to be at low risk for sexual victimization by an acquaintance. They also reported a very low likelihood of using physical resistance and only a moderate likelihood of using verbal assertiveness or indirect resistance if faced with such a situation. Comments during focus group discussions indicated that they maintain a high sense of invulnerability to being victimized (perhaps the most common remark being "I would just walk out"), even though they admit that sexual assault does occur "to other women." However, the context within which these sorority women socialize provides ample opportunity for sexual assault to occur to them as well. The prevalence of heavy alcohol consumption, peer norms related to attracting a man and engaging in sex play, and opportunities for isolation all set the stage for its occurrence. Consequently, these women appear to be in large part unprepared to deal effectively with this form of personal threat.

Two sets of hypotheses were proposed concerning the relationship between risk perception and resistance, and both were, for the most part, supported. First, the greater the presence of psychological barriers in women's cognitions, the more likely they were to endorse the use of indirect resistance strategies, and the less likely to endorse either verbal assertiveness or physical resistance. In particular, fear of embarrassment and concern about rejection by the man were especially noted as impediments to direct resistance. The relationships between alcohol consumption as a psychological barrier and resistance strategies were not as strong. It appears that alcohol as an actual risk factor and women's perception of it as such present a complex, and somewhat contradictory, set of relationships, which will be discussed later.

The acknowledgement in questionnaire and focus group responses of psychological barriers to resisting sexual assault is especially noteworthy. It indicates that women are placed in conflict when socializing with men. On the one hand, traditional sex roles promote behaviors intended to attract a man (e.g., dressing and acting sexy) and to assume responsibility for the quality of their social interaction (e.g., to be flattering of him; to smooth ruffled feelings and awkward moments between them). On the other hand, the high prevalence of sexual assault by acquaintances requires women to be alert to risk and self-protective with the same men they are expected to attract. In addition to pursuing simultaneously conflicting goals of affiliation and safety, women also face the strain of expressing their sexuality in a sociopolitical context that supports a double standard for men and women (see, for example, Muehlenhard & McCoy, 1991). Thus, women must "walk a cognitive tightrope" whenever they are in a dating or other social situation (Nurius & Norris, in press).

In addition, the findings suggest that the phenomenon of positivity bias warrants consideration regarding challenges in perceiving and coping with acquaintance sexual aggression. Social psychological findings indicate that rather than rationally objective perceptions of oneself, the world, and the future, people more typically operate on the basis of cognitive biases, such as unrealistic optimism, exaggerated perceptions of mastery, and overly positive self-evaluations (Friedland, Keinan, & Regev, 1992; Greenwald, 1980; Scheier & Carver, 1985; Taylor, 1989; Taylor & Brown, 1988). Although such biases may in general be associated with well-being, as can be seen in the present study, they incline women to underestimate their risk of encountering acquaintance sexual aggression; to overlook or misinterpret aggression cues when encountered; and to overestimate their own efficacy in resisting. Thus, a woman may have abstract knowledge about the general risk to women of sexual aggression by an acquaintance, but biases such as these incline her to appraise her own risk as low and her individual capacity to handle a threat as high (cf. Smith & Lazarus, 1993). This tendency was evident in the present study, suggesting that education about risk is insufficient to effect successful rape resistance and that greater understanding of the dynamics of women's appraisal processes relative to their own risk and protection is needed.

An additional element influencing women's cognitive mediation of cues related to sexual aggression involves limits of "working knowledge." Although people may amass a large repertoire of beliefs and coping responses, only a very limited amount of one's total cognitive repertoire can be activated in awareness at any given moment (Anderson, 1983; Nurius, 1993). Activated cognitive constructs tend to be mood congruent, making it far easier to access constructs consistent with one's prevailing mood and difficult to access those at odds with one's emotional state (Brown & Taylor, 1986; Mayer, Gaschke, Braverman, & Evans, 1992; Nurius & Markus, 1990). Thus, in a familiar and comfortable social situation, a woman's cognitions will be consistent with the "good time" aspects of the situation, and cognitions contrary to this will be difficult to access. For example, male behaviors that in a different context would be viewed as danger cues would instead more likely be interpreted in a situation- and moodcongruent way (e.g., as joking, showing off, or even as seduction). All in all, the combination of the stress associated with conflicting goals and the need to eliminate the uneasiness that these cause make the context ripe for errors in cognitively processing danger cues relative to sexual aggression by an acquaintance.

The findings regarding the role of alcohol in risk perception and resistance are particularly troubling. The relationships between participants' reported use of alcohol and projected resistance strategies, as well as number of sex partners, indicate that heavy alcohol consumption puts them at risk for sexual victimization. These findings are consistent with those of other researchers (e.g., Koss et al., 1987; Muehlenhard & Linton, 1987). That women's alcohol consumption is also related to an acknowledgement of alcohol as a barrier to effective resistance and to greater embarrassment may provide some insight to the way in which alcohol increases risk. Besides being physically incapacitated and thus less able to fend off an attacker, women in social situations may also succumb to the view that a drunk woman deserves to be violated. Thus, by blaming herself for inviting the assault, she may feel that she has no right to resist strongly. Yet the correlation between alcohol as a barrier and projected resistance (greater use of indirect methods; lower use of verbal assertiveness and physical resistance) appears to be a recognition that women understand the impairing potential of heavy alcohol consumption. Why would women drink heavily, knowing that it places them at increased risk?

The response conflict model of alcohol effects (Steele & Josephs, 1990) provides an appropriate framework for interpreting these findings. This model proposes that individuals perform socially inappropriate behaviors when intoxicated if, when sober, cues related to the inappropriateness would present a conflict. Through its detrimental effects on cognitive processes, alcohol relieves such conflict by focusing the individual on permissive cues (e.g., desire to feel relaxed and be seen as fun and sexy; encouragement by others to "party hardy"). Alcohol consumption thus provides one means of decreasing women's sense of conflict between social and safety goals, often leading unfortunately to adverse consequences. In addition, the role of expectancies associated with alcohol's effects cannot be ignored. Although these were not assessed in the questionnaire, focus group discussions brought forth many comments about alcohol's anticipated positive effects, especially with regard to social lubrication and providing an excuse for desired behavior that may otherwise be seen as imprudent or inappropriate (see Leigh, 1989 for a discussion). In addition to the physical effects of alcohol on cognitive processing, such anticipated positive effects of alcohol consumption further bias the individual to focus on situational cues consistent with positive expectancies rather than on cues contradictory to these expectancies.

The second set of hypotheses concerned the relationship between prior victimization and perceptions of risk and resistance. Indeed, women who had experienced at least one incident of sexual aggression in the last year experienced psychological barriers to a higher degree than non-victims and indicated that they would be less likely to resist sexual assault effectively. They also indicated a greater number of sex partners and higher BALs when socializing, even though they acknowledged the potentially disabling effects of alcohol to a greater level than nonvictimized respondents. Perhaps realistically, they perceived their risk of future victimization to be higher than that of nonvictims. Our findings that these women reported engaging in more, rather than fewer, behaviors and circumstances that may put them at greater risk is consistent with previous findings of higher rates of victimization for this group (e.g., Gidycz, Coble, Latham, & Layman, 1993; Himelein, 1995; Wyatt, Guthrie, & Notgrass, 1992) and of lower effectiveness of rape prevention interventions (Hanson & Gidycz, 1993). This suggests that prior experience of victimization is a significant risk factor for future victimization. Moreover, when including both childhood and adult forms of sexual assault, multiple victimization appears to be more common than not (Sorenson, Siegel, Golding, & Stein, 1991). Our data did not allow examination of the presence of risk factors, such as higher alcohol consumption and psychological barriers, before the earlier victimization experience. Their occurrence after victimization may have been a response to it, setting into motion a circular pattern of highrisk behaviors and victimization. More in-depth research is needed to address this issue. That is, how does the experience of victimization affect a woman's perception of her ability to perceive danger and resist it? Does she view herself with every victimization experience as less able to protect herself and thus become fatalistic in her behavior? If so, how can her perceptions and efficacy be reversed so that women who have experienced acquaintance sexual aggression may come to feel that they have learned from the experience and are now better prepared to contend with future threat?

Some of the findings presented here may be relatively distinct to sorority women as a group (e.g., safety assumptions based on familial concepts of the Greek system; concern about tarnishing the reputation of one's house or chapter) relative to college students who have no such clear reference group-although similar effects may be true for women who closely identify with other cultural, religious, or organizational groups. With respect to the basic patterns of cognitive appraisals, however, we have no reason to expect that these women are either developmentally or socially so unique that their perceptions, concerns, and expectations would not generalize. Because the women studied may have underreported victimization experiences to present the Greek system in a positive light, the findings associated with this measure should be viewed as conservative. Furthermore, the relatively small sample size employed here may not give an accurate representation of the true incidence of victimization either in the Greek system or among college women in general. Our goal, however, was to begin investigation of the underlying cognitive processes that may affect women's perception of risk and response to sexual aggression.

Further investigation is needed to ascertain the stability of these findings and their generalizability, as well as a more in-depth analysis of cognitive and contextual factors involved in detecting risk and resisting aggression. The relatively small number of participants in this study limited analyses of findings to bivariate techniques. However, research is underway that will recruit a sample of sufficient size to employ multivariate techniques, including causal modeling. Hence, the relative contributions of both immediate cognitive mediation factors, such as psychological barriers, and background factors, such as prior victimization, to predicting responses to acquaintance sexual aggression will be examined.

Although the prevalence of male sexual aggression toward women needs to be addressed at a societal level, effective individual resistance by women is unquestionably an essential component in preventing victimization. Here we are arguing for a need to better understand the cognitive challenges for women: how they anticipate and enter dating situations and how they are inclined to interpret what they encounter. This inquiry should help in understanding acquaintance rape as a social process to be analyzed and undermined as it occurs (Marcus, 1991), and in informing development of rape prevention strategies that "start where the woman is."

APPENDIX

1. Introductory script for focus groups

"As you know, we will be talking about the relationship between alcohol consumption and sex. It is quite common for people to socialize in a group setting like a party, have a few drinks, and become sexually intimate with someone during the night. Many times these experiences are pleasurable. Sometimes, however, they are instead painful and hurtful. We are particularly interested in finding out what kinds of things lead to having positive experiences, and the kinds of things that lead to having negative experiences. I will be asking a progressive series of questions from the expectations you have going into a social situation where you might meet someone you find attractive to the specific situations that might make a sexual experience either negative or positive. We'll progress like movies do, going from the first events and moving through each event to the last scene.

"Many people have different experiences. We are interested in getting specific information and experiences of those participating in this study. We have found that it is often times best to use your own experience or the experiences of friends close to you as a frame of reference when describing what you think. In other words, we know that there are lots of women who have lots of different experiences. We are interested in getting feedback about how people like you view these issues."

2. Stardard focus group questions

• What did you expect Greek life would be like before you joined? What were some of your expectations?

• Part of what we want to understand are the kinds of events that commonly occur in the Greek system where females and males socialize together. Are these events formal or informal? Do they occur in the evening? How many people attend, and from how many different houses? How important of a role does alcohol play?

• Now, I'd like to get a sense of what constitutes a great social event, one that for you personally is just about perfect. It may so happen that it doesn't actually happen a lot. Take a few moments to think about what constitutes a great (Greek System) social event. After you have a clear picture in your mind, what does it look like? What are some of the concerns you might have for the evening? What role does alcohol play in this?

• Now let's move on to the event itself. Let's talk for a moment about what you might do to help make the evening turn out the way you'd like it to. What might others around you notice about you? Also, what might you do to avoid the possible negative outcomes? What role does alcohol play in this?

• Now let's imagine that we are at the event, and you want to express your interest in someone you find attractive or really nice. Let's slow the motion of the camera down to carefully identify exactly what you and he communicate and how. Think for a moment about what you and he do and express verbally and nonverbally. Providing a sequential string of the events, what happens? What role does alcohol play here?

• Let's now say that the evening has further progressed and you and the guy you are with have communicated your interest in one another. You have been both talking and hanging out together. The situation then begins to heat up. Think for a moment about how this typically happens. Who initiates? How? What is said verbally and through actions? How do couples in this situation communicate sexual interest and initiate sexual play? How would you signal or communicate your desire to slow down or continue on? What role does alcohol play here?

• Let's say that things have become "steamy" and you two are kissing and maybe feeling each other. So far, it's been okay with you. However, at some point, he begins coming on too strong, moving faster than you are ready for. You may or may not have actually found yourself in this situation, but many women can imagine what such a situation might be like. Think for a moment about what you might be thinking. What might you feel? How do you realistically see yourself responding at first? How do you imagine he would respond to your actions? Say he keeps pressing and you feel really uncomfortable. What might you do next? How do you imagine you would feel or be thinking? What role does alcohol play here?

• Now we are out of the party and into the next day. Say that unwanted sex did occur. What might you be feeling? What would your thoughts be: about yourself, about him, about what others might think, about what happened? Would you talk to anyone? If you could go back in time and change the situation, what would you do differently?

• Final question: If you could create a prevention program that you believe would be helpful in averting a situation like the one described above, what do you think would help prevent unwanted sex from occurring?