Section: Articles

An ecological framework was used to examine individual, situational, and social predictors of university women's acknowledgment of rape experiences. Only individual and situational factors uniquely predicted acknowledgment. Women were more likely to acknowledge a rape if they experienced higher levels of violence during the rape, possessed factors congruent with an acquaintance rape script rather than a blitz rape script, and blamed their behavior for the rape. All women experienced a generally low amount of self-blame and blamed the perpetrator more than themselves. Implications for rape awareness programs are discussed.

Although approximately 15% of the women on college campuses nationwide reported that they had been rape victims (Koss, Gidycz, & Wisniewski, 1987), most of the women reported feeling victimized but did not acknowledge or label the event as rape (Koss, Dinero, Seibel, & Cox, 1988). As the word rape is so closely connected to beliefs and attitudes in our society, there is probably not a simple explanation of why some women[sup1] label acts as rape whereas others do not. In deciding whether to acknowledge a rape, a woman may be influenced by many factors including her individual history and dispositions, the behaviors of the rapist, the attitudes and reactions of her close friends and family, and sociocultural beliefs and expectations. As multiple factors are probably involved, an ecological model may be a useful framework to investigate rape acknowledgment.

Ecological models have been used as a method to organize findings about child abuse (Belsky, 1980), adolescent sexual aggression (White & Koss, 1991), and domestic violence (Dutton, 1988) and to integrate research on various types of violence against women (Heise, 1998; White & Kowalski, 1988). Belsky's (1980) ecological model contains four interrelated, interacting levels: the macrosystem (sociocultural values and beliefs), the exosystem (formal and informal social networks), the microsystem (interpersonal context of abusive acts), and the ontogenetic level (an individual's personal characteristics and developmental history). This type of model has the advantage of studying various factors separately while recognizing the interrelation of factors within and across levels. Considering each level separately helps avoid the tendency in psychology to focus on individual factors to the exclusion of important contextual, social, and cultural factors. Separating variables is an oversimplification the ecological model recognizes and compensates for by including the interaction between variables within and between levels. To fully understand complex social phenomena, conceptualization and study of interactions between variables is critical (Belsky, 1980; Heise, 1998; White & Kowalski, 1998). This study applies Belsky's (1980) ecological model to rape acknowledgment to investigate individual, situational, and social network factors related to rape acknowledgment.[sup2]

An individual level of analysis may include individual attitudes, beliefs, personality characteristics, and cognitive processes. Specifically, rape acknowledgment may be hampered by at least three individual-level factors: (a) self-blame (Frazier & Seales, 1997; Parrot, 1991; Pitts & Schwartz, 1993), (b) the internalization of blitz rape scripts (Kahn, Andreoli Mathie, & Torgler, 1994), and (c) romantic beliefs (Lloyd, 1991).

General social acceptance of victim-blaming rape myths can contribute to rape victims' feelings of self-blame, guilt, and shame (Katz & Burt, 1988). However, studies examining the effect of selfblame on rape acknowledgment have had mixed results. Two studies have found no difference between acknowledged and unacknowledged rape victims in responsibility felt for the rape (Kahn & Andreoli Mathie, 2000; Layman, Gidycz, & Lynn, 1996), whereas two other studies have suggested that unacknowledged rape victims had more self-blame than acknowledged rape victims (Frazier & Seales, 1997; Pitts & Schwartz, 1993). Although self-blame may be an important issue for acknowledgment, the nature of the relationship between self-blame and rape acknowledgment is still unclear. Previous research has found that the more a woman knew and trusted the perpetrator before the assault, the greater the self-blame after the rape (Katz & Burt, 1988; Koss, 1985; Mynatt & Allgeier, 1990). Therefore, because unacknowledged rape victims are more likely to know their attackers than acknowledged rape victims (Koss, 1985), it was hypothesized that unacknowledged rape victims would suffer more self-blame. It was also expected that women who acknowledged the rape would blame perpetrators more than women who did not acknowledge the rape.

In addition to self-blame, several authors have suggested that possession of a blitz rape script prevents women from accurately labeling an experience as rape (Kahn et al., 1994; Kelly, 1988; Parrot, 1991). A blitz rape is a statistically infrequent type of rape involving a violent attack by a stranger. Other elements common in blitz rape scripts include the use of weapons, physical harm to victims, and a lot of physical and verbal resistance by victims. Kahn and his colleagues (1994) found that the more rape victims endorsed blitz rape scripts, the less likely they were to acknowledge their rape experiences. It is expected that this study will replicate Kahn et al.'s findings and extend their results by examining the importance of blitz rape scripts in acknowledgment when situational and social networks are considered.

Many attitudes have been suggested as factors in rape acknowledgement (e.g., Bateman, 1991; Fisher, 1986). However, acceptance of traditional values, rape myths, sexual aggression, and traditional attitudes toward women, sexuality, and dating have not been found to distinguish acknowledged from unacknowledged rape victims (Koss, 1985) and were therefore not examined in this study. Research has not investigated the hypothesis that women's endorsement of romantic beliefs influences acknowledgment (Lloyd, 1991). A fanciful, idealistic, passionate view of love could lead a woman to disregard an abusive incident as a sign of overzealous love or attraction. Women who endorse attitudes such as "love conquers all" and "love at first sight" may be more likely to minimize intimate violence. If women interpret sexual aggression through a romantic lens, violence could be seen as an obstacle to overcome or as a characteristic of a person that should be accepted. Consequently, it was expected that the greater a woman's acceptance of romantic beliefs, the less likely she would be to acknowledge a rape.

Although the individual level focuses on the person, the situational level of analysis examines the interpersonal dynamics that shape behaviors. Three relational factors about rape have been isolated at the situational level that may influence rape acknowledgment: (a) the relationship of the victim to the perpetrator, (b) the amount of force used in a rape, and (c) the amount the victim resists. Previous findings have been inconsistent concerning the relationship between acknowledgment and degree of perpetratorvictim acquaintance. Whereas Koss (1985) found that unacknowledged rape victims knew the perpetrator better and had a higher level of intimacy prior to the attack than unacknowledged rape victims, Layman and her colleagues (1996) did not find a difference between acknowledged and unacknowledged women. Victims of acquaintance rape, however, acknowledge rape experiences less often than stranger rape victims (Koss et al., 1988). Thus, although it appears that rape involving someone the victim knows, especially romantically, may decrease the likelihood of acknowledgment, replication is needed.

In addition to the relationship between perpetrator and victim, the amount of force used in a rape has been related to acknowledgment in most, but not all, studies. Women who acknowledged rapes are more likely to have experienced physical force during the rapes than those who did not acknowledge their rapes (Gault, 1993; Kahn et al., 1994; Kibler, 1986; Layman et al., 1996; Schwartz & Leggett, 1999; see Koss, 1985, for an exception). Acknowledged rape victims also resisted the perpetrator more than unacknowledged rape victims (Layman et al., 1996; see Koss, 1985; LevineMacCombie & Koss, 1986, for exceptions). It is hypothesized that both greater force and greater resistance increase the likelihood of rape acknowledgment.

Finally, the ecological model looks at the influence of social networks on individual and interpersonal relations. The social network level of analysis examines social structures that directly encroach on an individual's daily life, such as family and peer groups. The peer group has been implicated as an important factor in sexual victimization. Having male friends who are sexually aggressive is an important predictor of female victimization and may legitimize sexually aggressive behaviors and increase a woman's tolerance for sexual victimization (Gwartney-Gibbs & Stockard, 1989). Therefore, it was expected that the more often a woman perceives sexual aggression occurring among her peers, the less likely she would be to acknowledge a rape experience.

This study examined influences on rape acknowledgment of university women using an ecological framework. At the individual level of analysis, acknowledged rape victims were predicted to be less likely to blame themselves for the rapes, more likely to blame the perpetrators for the rapes, less likely to possess blitz rape scripts, and less likely to have less romantic notions of love than unacknowledged rape victims. At the situational level of analysis, it was expected that a higher level of perpetrator force, victim resistance, and victim harm during the rapes and a lower degree of acquaintance with the perpetrators would increase acknowledgement. At the social network level of analysis, acknowledged rape victims were predicted to be less likely to be in peer groups in which sexual aggression is common.

Participants were 109 women who had experienced rape; they were recruited from introductory psychology classes at a medium-sized southeastern university in exchange for class credit. The participants were 80% European American, 16% African American, 2.4% Hispanic, and 1.9% Asian American. The majority of the participants were single (88.4%) and between 17 and 21 years of age (93%).

Women who had rape experiences, assessed by a modified version of the Sexual Experiences Survey (White, 1990), were telephoned, informed about the study, and asked to till out a questionnaire. The questionnaire included the Modified Sexual Experience Survey followed by questions about their most serious victimization experiences, questions about their relationships to the perpetrators, and questions about victim and rapist responsibility, blame, and causality. The Characterological and Behavioral Self-Blame Questionnaire (Hill & Zautra, 1989), the Romantic Beliefs Scale (Sprecher & Metts, 1989), an assessment of perceptions of sexual aggression in the peer group (GwartneyGibbs & Stockard, 1989), and a blitz rape script survey were also included. All participants were debriefed afterward. To anticipate the possibility that completing the questionnaire might cause distress after the debriefing, each participant was given a list of local and regional mental health resources.

The female version of the Sexual Experiences Scale (SES) (Koss et al., 1987) modified by White (1990) was used. Although evidence regarding the reliability and validity of the SES has been established (Koss & Gidycz, 1985; Koss et al., 1987; Koss & Oros, 1982), the original version of the SES consists of items that confound behavioral tactic and outcome. The modified version of the SES contains 28 questions that ask how often, since the age of 14, each of four types of sexual behavior have occurred (sex play, attempted intercourse, intercourse, and sex acts) across seven types of situations (consensual, false use of flattery, verbal pressure, use of authority, threat of force, use of force, and intoxication). The test-retest reliability for the modified SES is .65 (White & Hoecker, 1994).

Consistent with past research on acknowledgment (Kahn et al., 1994, Koss, 1985; Koss et al., 1988; Layman et al., 1996), participants were classified as rape victims if they indicated that they had had vaginal, oral, or anal intercourse against their will due to physical force, threats of physical force, or physical incapacitation from alcohol or drugs.

After taking the SES, participants were asked to indicate the number of the question in the SES that reflected the most serious experience that had happened to them. If an experience occurred more than once, they were asked to answer further questions for the most recent experience. For that experience, participants rated, on a 5-point scale (1 = not at all, 5 = a great deal), how well they knew the perpetrator, the relationship to the perpetrator, the degree of physical harm suffered, and the degree of verbal and physical resistance they had offered.

After answering questions about their most serious experience, the question, "Have you ever been raped?" was asked. The women who indicated that they had experienced a rape and responded affirmatively to this question were considered "acknowledged rape victims." Those who responded negatively to this question but who had been raped were labeled "unacknowledged rape victims."

The Characterological and Behavioral Self-Blame Questionnaire (Hill & Zautra, 1989) is a measure of how much a woman blames aspects of her personality for a rape (e.g., "I'm a bad person," "I trust people too much") and how much she blames her behavior for a rape (e.g., "I didn't resist"). The internal consistency was adequate to good (.68 and .95). Designed to assess the effect of self-blame on adjustment in rape victims, the original Characterological and Behavioral Self-Blame Questionnaire asks the person to rate on a 5-point scale the frequency of each self-blaming thought in the past month. Acknowledgment of rape, unlike adjustment, may be related to the occurrence of self-blaming thoughts rather than the frequency. Therefore, the rating scale was altered to assess how much each self-blaming statement was a factor in their rapes (1 = definitely not a factor, 5 = definitely a factor).

Shaver and Drown (1986) criticized measures of behavioral and characterological self-blame for assessing causality and responsibility instead of blame. They suggest that the confounding of causality, blame, and responsibility has resulted in contradictory results in self-blame research. For this reason, six questions were designed to reflect causality, responsibility, and blame separately for both the participants and the perpetrators. Participants rated perpetrator and self-causality, responsibility, and blame for the incident on a 5-point scale (1 = not at all, 5 = a great deal).

The Romantic Beliefs Scale has 15 items that measure four distinct aspects of romantic beliefs about intimate relationships: love finds a way, one and only, idealization, and love at first sight (Sprecher & Metts, 1989). Three-week test-retest reliability, internal consistency, and concurrent validity were assessed and found to be adequate (Sprecher & Metts, 1989).

Kahn and his associates (1994) found that women who acknowledged rape experiences were less likely to demonstrate possession of blitz rape scripts. The features of blitz rape scripts generated by women in the aforementioned study were put into a questionnaire format to assess the possession of blitz rape scripts. Women were instructed to check the characteristics that fit what they thought of as a "typical" rape.

A sexually aggressive peer group was assessed by two questions asking individuals how many of their (a) male friends were sexually aggressive and (b) female friends had been sexually victimized (Gwartney-Gibbs & Stockard, 1989).

Consistent with past research, the majority (64%) of the women did not acknowledge their rape experiences (Kahn et al., 1994; Kahn & Andreoli Mathie, 2000; Koss, 1985; Pitts & Schwartz, 1993; Schwartz & Leggett, 1999). All noncategorical variables were approximately normally distributed. As there was no association between the demographic variables (age, race, marital status) and acknowledgment, they were not included in any further analyses.

The correlations of variables that were conceptually related were calculated to see if these variables could be combined. The attribution variables, the romantic beliefs, and the sexually aggressive peer group variables had small (.02 to .29) to moderate (.30 to .60) intercorrelations and thus were not combined. Levels of physical and verbal resistance were highly correlated (.85) and combined into a new variable, level of violence (described below).

Each hypothesis was tested separately using logistic regression analyses, and then a stepwise logistic regression was conducted to determine which variables uniquely contributed to the prediction of acknowledgment. Because level of violence was such a strong predictor of acknowledgment, it was used as a covariate in all analyses unless otherwise indicated.

Acknowledged rape victims were more likely to report a higher level of physical force during the rape than unacknowledged rape victims, X[sup2] = 32.77, p < .0001. In fact, less than 10% of the women who had experienced lower levels of force (threat of force or incapacitation) acknowledged their experiences. In contrast, 62% of women who had experienced physical force labeled the experiences as rape.

Acknowledged rape victims reported that they resisted more verbally and physically and suffered more physical harm than unacknowledged rape victims (see Table 1). A principal components analysis of verbal resistance, physical resistance, and physical harm revealed one factor that explained 76% of the variance in acknowledgment, with factor loadings ranging from .75 to .93. A new factor, level of violence, was created by multiplying the mean of these three items and the ordinal variable from the SES that assessed the type of force used in the rape (1 = incapacitation due to alcohol or drugs, 2 = threat, 3 = force). The intercorrelations of the composite variable with the individual items ranged from .70 to .84, suggesting a high amount of shared variance among the measures. This composite variable, level of violence, was used as a covariate in further analyses.

The hypothesis that acknowledged rape victims would be less familiar with the perpetrators and less close to the perpetrators before the rape than unacknowledged rape victims was not supported. Acknowledged and unacknowledged rape victims did not differ on how well they knew perpetrators, X[sup2](3, N = 89) = .71, p < .87. The majority of the victims classified perpetrators as friends (29%) or boyfriends (39%); 27% of the women responded that perpetrators were casual acquaintances; only 6% said the perpetrators were strangers, and no one said they were family members. The degree of closeness victims felt toward the perpetrators also failed to predict acknowledgment in a logistic regression (see Table 1).

Contrary to prediction, acknowledged rape victims engaged in more characterological and behavioral self-blame than did unacknowledged rape victims (see Table 1). For characterological and behavioral self-blame, the overall means (M = 2.39 and M = 2.07, respectively) suggest that both groups engage in little self-blame and that both acknowledged and unacknowledged rape victims engaged in more characterological than behavioral self-blame, t[subl09] = 5.13, p < .0001.

Blame, responsibility, and causality attributed to the self and to the perpetrator did not predict acknowledgment when level of violence was controlled (see Table 1). Women's self-attributions were generally low. For both acknowledged and unacknowledged rape victims, the women made greater attributions toward the perpetrators than toward themselves: cause, t[sub106] = 10.08, p < .0001; responsibility, t[sub106] = 9.36, p < .0001; blame, t[subl05] = 6.34, p < .0001.

Romantic beliefs did not predict acknowledgment when controlling for level of violence (see Table 1).

As expected, unacknowledged rape victims were more likely than acknowledged rape victims to endorse several items consistent with blitz rape scripts: physical attacks, threats of weapon use, weapon use by perpetrators, screams by victims, and severe physical harm to victims (see Table 2). Acknowledged rape victims were also significantly more likely than unacknowledged rape victims to endorse items consistent with acquaintance rape scripts: physical restraint, mild physical harm, and drinking by the victims. A composite variable, created by averaging the items that unacknowledged rape victims endorsed more frequently, predicted acknowledgment when controlling for the use of violence, Wald x[sup2](1, N = 89) = 12.91, p < .0003, log-odds -2.15. This composite variable was used in the stepwise logistic regression.

The prediction that acknowledged rape victims would perceive their peer groups as containing more sexually aggressive men and sexually victimized women was not supported. When the level of violence of the rape was controlled, the number of rape victims who reported having sexually aggressive male peers did not predict acknowledgment, Wald x[sup2](1, N = 88) = .24, p < .62, logodds = -.07 (acknowledged M = 1.95, SE = .13; unacknowledged M = 1.86, SE = .09). There was a trend toward significance in the number of victimized female peers' prediction of acknowledgment when level of violence was controlled in a logistic regression analysis, Wald x[sup2](1, N = 88) = 2.69, p < .10, log-odds = .27. Contrary to expectation, the acknowledged rape victims reported knowing mote women who had been sexually victimized (M = 2.74, SE = .13) than did the unacknowledged rape victims (M = 2.33, SE = .08).

A stepwise logistic regression assessed the unique predictive ability of each variable. Level of violence, characterological self-blame, behavioral self-blame, possession of blitz rape scripts, and number of sexually victimized female peers were put into the model as predictors of acknowledgment. Of these variables, level of violence, possession of blitz rape scripts, and behavioral self-blame were unique contributors to the prediction of acknowledgment and correctly classified 86.2% of the acknowledged rape victims (see Table 3).

Consistent with past research, the majority (64%) of the women did not acknowledge their rape experiences (Kahn et al., 1994; Kahn & Andreoli Mathie, 2000; Koss, 1985; Pitts & Schwartz, 1993; Schwartz & Leggett, 1999). Although variables at all three levels predicted acknowledgment, situational and individual variables were the best predictors. This study supports previous findings that a combination of a belief in specific elements of blitz rape scripts and relatively nonviolent sexual assault experiences appear to be the most important factors hindering women from acknowledging rape (Kahn et al., 1994). The combined variable, perceived violence of a rape, reflected in this study by a combination of victims' self-reports of the amount of perpetrator force, physical harm experienced, and verbal and physical resistance was the best predictor of acknowledgment. Women who perceived their rapes as more violent were more likely to acknowledge the rapes.

Another situational variable, perpetrator-victim relationship, failed to predict acknowledgment. Although consistent with some research (Kibler, 1986; Layman et al., 1996), two other studies have found that the closer the relationships between women and the perpetrators, the less likely they were to acknowledge the rapes (Gault, 1983; Koss, 1985). Future research is needed to further clarify the nature of the association between acknowledgment and the victims' relationships with perpetrators.

The finding that possession of specific elements of blitz rape scripts uniquely predicted acknowledgment supports and extends previous research (Kahn et al., 1994). The greater the endorsement of blitz rape scripts, the less likely a woman is to acknowledge the rape, even when the violence of the rape and other predictors of acknowledgment are considered. Possession of a blitz rape script may reduce the likelihood of acknowledgment because most rapes do not fit with elements of the script. This is supported by qualitative research that lack of acknowledgment in rape victims was partly related to a lack of congruence between their ideas of rape and the complexity of their personal experiences (Phillips, 1995). Because the data are correlational, however, it is not clear whether the possession of blitz rape scripts causes the lack of acknowledgment. It is possible that the labeling of experiences as "not rape" may lead women to redefine the concept of rape in more extreme terms to support this decision. For unacknowledged rape victims, the possession of blitz rape scripts could be either a justification for denial or a hindrance to labeling.

Inconsistent with Kahn et al.'s (1994) findings, a number of features of blitz rape scripts did not differentiate acknowledged and unacknowledged rape victims in this study (i.e., whether the rapist was a stranger or acquaintance and whether the rape occurred indoors or outdoors). These findings suggest that the most important aspect of blitz rape scripts is the belief in a high level of force, not who the perpetrators are or where the rapes occurred. It is possible that the differences between the two studies are related to different methodologies. Kahn et al. had women write scenarios, whereas women in this study responded to a checklist. Replication is needed to ascertain the important elements of blitz rape scripts across samples and methodologies.

Counter to what was hypothesized, acknowledged rape victims were higher in both behavioral and characterological self-blame after controlling for the amount of violence in the rapes. The prediction that unacknowledged rape victims would blame themselves more was based on previous findings that the closer the relationship with the perpetrator, the greater was the self-blame (Katz & Burt, 1988) and that unacknowledged rape victims usually know their attackers better than acknowledged rape victims (Koss, 1985). However, in this study, unacknowledged rape victims did not report knowing their attackers better than acknowledged rape victims. Thus, when level of relationship with perpetrators does not distinguish acknowledged and unacknowledged rape victims, blame is greater for acknowledged rape victims. It is also possible that when women label an experience "rape," she thereby affirms the seriousness of the sexual abuse and its aftermath. This difference in seriousness of the experience may lead acknowledged rape victims to blame themselves more than unacknowledged rape victims. As acknowledged rape victims report more posttraumatic stress symptoms than unacknowledged rape victims (Layman et al., 1996) and as self-blame has been associated with depressive symptoms and demoralization in rape victims (Frazier, 1990, 1991; Hill & Zautra, 1989), increased self-blame in acknowledged rape victims may be associated with poorer adjustment after the rape. Furthermore, the concept of rape may be so loaded with implications for victim culpability in our society that labeling the experience as rape is concomitant with higher self-blame. Although assumed to be a positive step for rape victims, acknowledgment also may carry the burden of self-deprecation for rape victims because of the negative associations attached to them in our culture (Lebowitz & Roth, 1994). For unacknowledged rape victims, labeling the event as "not rape" may be a strategy to avoid blame.

As behavioral and characterological self-blame were correlated and behavioral self-blame was a stronger predictor of acknowledgment, only behavioral self-blame remained in the stepwise regression model predicting acknowledgment. The overlap between the two measures in the prediction of acknowledgment is consistent with the argument that characterological and behavioral self-blame may be inseparable in the case of rape victims (Katz & Burt, 1988) and that characterological blame implies behavioral blame (Frazier, 1990; Pitts & Schwartz, 1993).

The amount of self-blame in this study was generally low, and self-attributions were lower for all the women than were their attributions to the perpetrators. Although self-attributions are assumed to be a central issue with which rape victims deal (Katz & Burt, 1988; Lebowitz & Roth, 1994; Pitts & Schwartz, 1993), the level of self-blame in this study, especially when compared to the level assigned to the perpetrator, raises questions about the importance of self-blame to rape victims. Abbey (1987) found self-attributions to occur at a low frequency among acknowledged rape victims from a rape crisis center. Self-blame may be a less important factor for nonclinical samples of rape victims, perhaps especially for women whose enrollment in a university suggests they are able to withstand the multiple demands and pressures of higher education. Women more seriously affected by rape may be lower functioning and choose to not attend college or to withdraw from college after rape. Another possibility is that rape education programs are encouraging women not to blame themselves for sexual violence. Future research is needed to explore how attributions are related to acknowledgment, postrape adjustment, and rape prevention efforts.

Romantic beliefs failed to predict acknowledgment. This is consistent with findings that personality and attitudinal variables do not predict acknowledgment (Koss, 1985). Labeling may have more to do with the specific context of rapes and scripts about rape than about broader attitudes and beliefs.

It was predicted that the perception of a greater number of sexually aggressive male peers and sexually victimized female peers would increase the tolerance for sexual victimization and thereby reduce the likelihood of acknowledgment. Instead, the perceived number of sexually aggressive male peers had no predictive ability for acknowledgment and the number of sexually victimized female peers was marginally higher for acknowledged rape victims than for unacknowledged rape victims when level of violence in the rape was controlled. This study does not confirm the hypothesis that a greater perception of sexual aggression in the peer group is associated with an increased tolerance for rape in the form of lower rape acknowledgment. Although it may be that some women in peer groups with sexually aggressive and sexually victimized peers have a higher tolerance for sexual aggression, this does not appear to be the case for actual rape victims. Perhaps the knowledge of peer victimization creates a supportive climate that facilitates acknowledgment. Women who acknowledge a rape may be more likely to discuss the rape with their peers, thus increasing the opportunities for reciprocal disclosures. Both social learning and social comparison theories could operate in such an environment. Having friends define victimization experiences as rape may serve as a role model to other victims who, through the process of social comparison, decide to label their experiences as rape. As this was a marginal finding, replication is needed as well as research into the role of social support, awareness, and modeling in acknowledgment.

The ecological model was useful for conceptualizing factors involved in rape acknowledgment. Although factors in all three categories predicted acknowledgment, only situational and individual factors were unique predictors. It is possible that individual and situational attributes of rape may have more impact on acknowledgment because they are more directly related to the rape experience. Having sexually victimized women in a peer group may change women's thoughts about rape in general or about her rape in particular but may be less powerful as a predictor than a direct measure of their beliefs. It is also possible that the weak prediction of social network variables was due to the small number of variables assessed at this level and the measures' lack of psychometric properties (i.e., one-item measures). Rather than concluding that variables at the social network level are less important to rape acknowledgment, further investigation into whether social network factors affect women's cognitions about rape is warranted. Future research should include more variables at the social network level, including social support (Pitts & Schwartz, 1993), to determine the importance of factors at this level.

The study is limited to university women between the ages of 18 and 25 who are primarily White and middle class. Older women may be more likely to acknowledge rape experiences due to greater life experience. Exploration of acknowledgment across a more diverse ethnic sample may reveal the influence of cultural values. Latina and Indian women, for instance, may be less likely to acknowledge rape due to stronger expectations for premarital virginity and marital fidelity (Reid & Bing, 2000). The results are also limited by the correlational nature of this study. A longitudinal study would permit the temporal sequencing of the relationships between the variables. Labeling rape experiences may change over time as awareness of peers' victimization experiences, reactions of friends, self-blame, and concepts of rape change. It is possible that some of the women who did not acknowledge will later define their experiences as rape. Continuing efforts to isolate the factors that affect acknowledgment will help guide rape awareness educators. At present, rape awareness educators should focus on educating women about the incongruity between blitz rapes and typical rape experiences while continuing to send the message that women are not to blame for rape.

AUTHOR'S NOTE: This study, partially funded by the Grants-in-Aid Program of the Society for Psychological Study of Social Issues, represents a part of the author's dissertation at the University of North Carolina at Greensboro. Thanks go to Jacquelyn W. White, David Herr, Jean Malone, Ernest Lumsden, and Herbert Wells for their helpful comments.

[sup1]. Although men can be raped, this article will focus only on female rape victims, not because male rape victims are less important but because there are likely significant differences between female and male rape victims.

[sup2]. The sociocultural level, although important, was outside the realm of this study.

Legend for chart:

A1=Resist verbally[supa]
A2=Resist physically[supa]
A3=Physical harm[supa]
A4=Closeness to perpetrator[supa]
A5=Characterological
A6=Perpetrator caused
A7=Perpetrator responsible
A8=Perpetrator to blame
A9=Love at first sight
                  Acknowledged   Unacknowledged
Variable           M     SE      M    SE  Wald X[sup2]     Log-Odds
A1                3.46  .16     2.29  .17  15.51[sup****]   .38
A2                3.54  .17     2.27  .19  15.61[sup****]   .34
A3                3.00  .19     1.50  .12  21.58[sup****]   .59
A4                2.58  .30     2.64  .21   2.63           -.17
A5                2.66  .12     2.23  .11   5.37[sup*]     -.48
Behavioral        2.32  .10     1.83  .09   8.61[sup**]    -.76
A6                4.72  .14     4.54  .13    .85            .18
A7                4.59  .14     4.56  .12    .01            .02
A8                4.42  .19     4.19  .17    .86            .13
Self-caused       2.81  .21     2.67  .19    .17            .17
Self-responsible  2.94  .21     2.52  .19   1.62           -.15
Self to blame     3.31  .23     2.86  .21   1.49           -.12
Love finds a way  3.68  .13     4.02  .11   2.57           -.29
One and only      3.11  .17     3.48  .15   2.10           -.19
Idealization      2.83  .14     2.74  .12    .23            .08
A9                2.74  .16     2.74  .14    .03            .02
[supa]. Logistic regression analyses. Violence used as a covariate,
except in analyses with superscript.
[sup*]p < .05.
[sup**]p < .01.
[sup****]p < .0001. Two-tailed tests.

Legend for chart:

A1=Unacknowledged
A2=Wald x[sup2]
A3=3.55[sup*]
A4=8.84[sup**]
A5=4.70[sup*]
A6=5.57[sup*]
A7=4.33[sup*]
A8=10.55[sup***]
A9=5.09[sup*]
A10=5.40[sup*]
A11=Victim tells friend and/or relative
A12=Avoids perpetrator afterward[supb]
A13=Perpetrator threatens others
                            Acknowledged   A1
Blitz Rape Items             M     SE      M    SE  A2    Log-Odds
Stranger                    .09   .07     .26  .06  2.94   -.60
Outdoors                    .16   .08     .34  .07  2.53   -.45
Verbal coercion             .90   .07     .70  .07  2.95    .58
Physical threat             .75   .09     .61  .08  1.10    .27
Physical restraint          .73   .09     .48  .08  A3      .49
Physical attack             .38   .09     .81  .07  A4     -.95
Threat of weapon            .44   .09     .74  .08  A5     -.56
Perpetrator uses weapon     .13   .08     .44  .07  A6     -.70
Victim screams              .54   .08     .84  .07  A7     -.57
Victim cries[supb]         1.00   .00    1.00  .00    NA     NA
Victim tries to leave       .91   .06     .83  .05   .71    .33
Victim protests[supb]       .83   .04     .99  .04    NA     NA
Victim struggles[supb]      .93   .03    1.00  .00    NA     NA
Severe physical harm        .14   .09     .63  .07  A8    -1.06
Mild physical harm          .91   .08     .62  .07  A9       .66
Victim drinking             .46   .08     .16  .07  A10      .68
Perpetrator drinking        .58   .09     .56  .08   .05    .06
Victim flees                .72   .08     .81  .07   .47   -.21
Victim tells police         .03   .05     .12  .04  1.25   -.51
A11                         .51   .09     .37  .08  1.12    .26
Victim tells no one[supb]  0     0                    NA     NA
A12                         .96   .03     .99  .02    NA     NA
A13                         .32   .09     .59  .09  3.19   -.46
Perpetrator flees           .76   .07     .92  .06  2.80   -.59
[supa]. Logistic regression used for all analyses with violence as
a covariate.
[supb]. Logistic regression could not estimate parameters because
of a lack of variability in at
least one group.
[sup*]p < .05.
[sup**]p < .01.
[sup***]p < .001. Two-tailed tests.

Legend for chart:
A1=Wald X[sup2]
A2=Behavioral self-blame
                                                         Percentage
                                                          Correctly
Variable               B   SE B   A1      p    Log-Odds  Classified
Step 1                                                     78.7
Level of violence     .40   .08  24.85  .0001    .17
Step 2                                                     81.8
Level of violence     .51   .11  22.43  .0001    .22
Blitz rape script   -4.96  1.38  13.03  .0003  -2.15
Step 3                                                     86.2
Level of violence     .68   .15  21.56  .0001    .30
Blitz rape script   -5.73  1.60  12.79  .0003  -2.52
A2                  -2.24   .81   7.73  .005     .97
NOTE: Five variables were used as predictors: level of violence,
characterological self-blame, behavioral self-blame, possession
of blitz rape scripts, and number of sexually
victimized peers.

Abbey, A. (1987). Perceptions of personal avoidability versus responsibility: How do they differ? Basic and Applied Psychology, 8, 3-19.

Bateman, P. (1991). The context of date rape. In B. Levy (Ed.), Dating violence: Young women in danger (pp. 94-99). Seattle, WA: Seal Press.

Belsky, J. (1980). Child maltreatment: An ecological integration. American Psychologist, 35, 320-335.

Dutton, D. (1988). The domestic assault of women: Psychological and criminal justice perspectives. Boston: Allyn & Bacon.

Fisher, G. J. (1986). College student attitudes toward forcible date rape: Vol. 1. Cognitive predictors. Archives of Sexual Behavior, 15, 457-467.

Frazier, P. (1990). Victim attributions and post-rape trauma. Journal of Personality and Social Psychology, 59, 298-304.

Frazier, P. (1991). Self-blame as a mediator of postrape depressive symptoms. Journal of Social and Clinical Psychology, 10, 47-57.

Frazier, P. A., & Seales, L. M. (1997). Acquaintance rape is real rape. In M. D. Schwartz (Ed.), Researching sexual violence against women: Methodological and personal perspectives (pp. 54-64). Thousand Oaks, CA: Sage.

Gault, C. A. (1993). An application of the theory of reasoned action to the understanding and prediction of disclosure of sexual victimization. Unpublished master's thesis, Auburn University.

Gwartney-Gibbs, P., & Stockard, J. (1989). Courtship aggression and mixed-sex groups. In M. Pirog-Good & J. Stets (Eds.), Violence in dating relationships (pp. 185-204). New York: Praeger.

Heise, L. L. (1998). Violence against women: An integrated, ecological framework. Violence Against Women, 4, 267-290.

Hill, J. L., & Zautra, A. J. (1989). Self-blame attributions and unique vulnerability as predictors of post-rape demoralization. Journal of Social and Clinical Psychology, 8, 368-375.

Kahn, A. S., & Andreoli Mathie, V. (2000). Understanding the unacknowledged rape victim: A lifespan perspective. In C. B. Travis & J. W. White (Eds.), Sexuality, society, and feminism (pp. 377-403). Washington, DC: American Psychological Association.

Kahn, A. S., Andreoli Mathie, V., & Torgler, C. (1994). Rape scripts and rape acknowledgment. Psychology of Women Quarterly, 18, 53-66.

Katz, B., & Burt, M. (1988). Self-blame in recovery from rape: Help or hindrance. In A. W. Burgess (Ed.), Sexual assault (Vol. 2, pp. 345-358). New York: Garland.

Kelly, L. (1988). Surviving sexual violence. Minneapolis: University of Minnesota Press.

Kibler, K. J. (1986). Differential labeling of acquaintance rape. Unpublished master's thesis, Auburn University.

Koss, M. P. (1985). The hidden rape victim: Personality, attitudinal, and situational characteristics. Psychology of Women Quarterly, 9, 193-212.

Koss, M. P., Dinero, T. E., Seibel, C. A., & Cox, S. L. (1988). Stranger and acquaintance rape: Are there differences in the victim's experiences? Psychology of Women Quarterly, 12, 1-24.

Koss, M. P., & Gidycz, C. A. (1985). Sexual Experiences Survey: Reliability and validity. Journal of Consulting and Clinical Psychology, 53, 422-423.

Koss, M. P., Gidycz, C., & Wisniewski, N. (1987). The scope of rape: Incidence and prevalence of sexual aggression and victimization in a national sample of higher education students. Journal of Consulting and Clinical Psychology, 55, 162-170.

Koss, M. P., & Oros, C. (1982). Sexual Experiences Survey: A research instrument investigating sexual aggression and victimization. Journal of Consulting and Clinical Psychology, 50, 455-457.

Layman, M. J., Gidycz, C. A., & Lynn, S. J. (1996). Unacknowledged versus acknowledged rape victims: Situational factors and posttraumatic stress. Journal of Abnormal Psychology, 105, 124-131.

Lebowitz, L., & Roth, S. (1994). "I felt like a slut": The cultural context and women's response to being raped. Journal of Traumatic Stress, 7, 363-390.

Levine-MacCombie, J., & Koss, M. P. (1986). Acquaintance rape: Effective avoidance strategies. Psychology of Women Quarterly, 50, 311-320.

Lloyd, S. A. (1991). The dark side of courtship: Violence and sexual exploitation. Family Relations, 40, 14-20.

Mynatt, C. R., & Allgeier, E. R. (1990). Risk factors, self-attributions, and adjustment problems among victims of sexual coercion. Journal of Applied Social Psychology, 20, 130-153.

Parrot, A. (1991). Medical community response to acquaintance rape-recommendations. In A. Parrot & L. Bechhofer (Eds.), Acquaintance rape: The hidden crime (pp. 304-316). New York: John Wiley.

Phillips, L. (1995). Flirting with danger: A study of the multiple meanings of male aggression in women's hetero-relations. Unpublished doctoral dissertation, Graduate School of Education, University of Pennsylvania.

Pitts, V. L., & Schwartz, M. D. (1993). Promoting self-blame in hidden rape cases. Humanity and Society, 17, 383-398.

Reid, P. T., & Bing, V. M. (2000). Sexual roles of girls and women: An ethnocultural lifespan perspective. In C. B. Travis & J. W. White (Eds.). Sexuality, society, and feminism (pp. 141-166). Washington, DC: American Psychological Association.

Schwartz, M. D., & Leggett, M. S. (1999). Bad dates or emotional trauma? The aftermath of campus sexual assault. Violence Against Women, 5, 251-271.

Shaver, K. G., & Drown, D. (1986). On causality, responsibility, and self-blame: A theoretical note. Journal of Personality and Social Psychology, 50, 697-702.

Sprecher, S., & Metts, S. (1989). Development of the "romantic beliefs scale" and examination of the effects of gender and gender-role orientation. Journal of Social and Personal Relationships, 6, 387-411.

White, J. W. (1990). A modification of the Koss and Oros Sexual Experiences Survey. Unpublished manuscript, University of North Carolina, Greensboro.

White, J. W., & Hoecker, K. S. (1994). [Reliability of the Modified Sexual Experiences Survey]. Unpublished raw data.

White, J. W., & Koss, M. P. (1991). Adolescent sexual aggression within heterosexual relationships: Prevalence, characteristics, and causes. In H. E. Barbaree, W. L. Marshall, & D. R. Laws (Eds.), The juvenile sexual offender (pp. 182-202). New York: Guilford.

White, J. W., & Kowalski, R. M. (1998). Male violence toward women: An integrated perspective. In R. G. Geen & E. D. Donnerstein (Eds.), Human aggression: Theories, research, and implications for social policy (pp. 203-228). San Diego: Academic Press.

~~~~~~~~

By Barrie Bondurant, Lyon College rtland

Barrie Bondurant, Ph.D., is an assistant professor of psychology at Lyon College. Her research focuses on gender issues in interpersonal aggression, including sexual assault, domestic violence, and sexual harassment.