The current study examined how participant (i.e., gender) and situational (i.e., couple-specific sexual intercourse history)factors influence attributions about stranger, early dating, late dating, and marital rape. Two hundred undergraduates were randomly assigned to read one of four levels of victim-perpetrator relationship vignettes. In addition, within the two dating relationship conditions, the couple's sexual intercourse history was manipulated. As the degree of acquaintance between the perpetrator and victim increased, the participants incorporated more rape-supportive and sex role stereotypical attributions. Participants also made more negative attributions about the date rapes when the couple was thought to have previously engaged in consensual sexual intercourse. There were no differences in attributions across the date and marital rape conditions when the participants were informed that the dating couples had previously engaged in sexual intercourse, and several gender differences were obtained in these comparisons. The social and legal implications of these findings are discussed.

Analogue research has been used to investigate societal attributions and perceptions of different types of sexual assault. These studies have consistently found that individuals consider sexual assault by a stranger to be a more serious crime that would be more psychologically and physically detrimental to the victim than sexual assault committed by an acquaintance (Barnett, Quackenbush, Sinisi, Wegman, & Otney, 1992; Bell, Kuriloff, & Lottes, 1994; Bridges, 1991; Bridges & McGrail, 1989; Johnson & Jackson, 1988; Klemmack & Klemmack, 1976; L'Armand & Pepitone, 1982; Perry, McLendon, & Foley, 1994; Quackenbush, 1989; Szymanski, Devlin, Chrisler, & Vyse, 1993; Tetreault & Barnett, 1987). Similar results have been found in studies comparing attitudes about stranger and marital rape (Monson, Byrd, & Langhinrichsen-Rohling, 1996; Sullivan & Mosher, 1990). In many of these studies, women have been found to consider all rapes as being more serious than men consider them to be (Barnett et al., 1992; Bell et al., 1994; Bridges, 1991; Monson et al., 1996; Szymanski et al., 1993). However, both male and female perceivers tend to consider rape by an acquaintance compared to rape by a stranger to be less serious and more attributable to the characteristics of the victim than rape by a stranger.

The sex role socialization theory of rape (Burt, 1980) has been advanced as a framework for understanding these differential attributions surrounding stranger versus acquaintance rape (Bridges, 1991; Check & Malamuth, 1983; Monson et al., 1996). This model posits that as a result of the developmental processes involved in learning the appropriate behavior for one's sex, both men and women develop certain expectations regarding acceptable behaviors in sexual interactions (Bridges, 1991). According to this theory, acquaintance rape is more likely than stranger rape to be viewed as an extreme extension of traditional male-female sexual interaction. Furthermore, this model proposes that the process of sex role socialization promotes the acquisition of rape-supportive beliefs, or false beliefs about rape. These rape-supportive beliefs work to mitigate the perceived seriousness of rape, especially rape that occurs between acquaintances.

Research addressing the applicability of the sex role socialization theory of rape to acquaintance rape attributions has followed two parallel lines. The first line of research has focused on comparing date versus stranger rape. In general, increasingly traditional sex role beliefs held by perceivers have been related to an increased likelihood of minimizing the victim's rape experience, attributing greater responsibility to the victim, and a greater hesitancy in defining acquaintance rape situations as rape (Check & Malamuth, 1983; Coller & Resick, 1987; Proite, Dannells, & Benton, 1993; Quackenbush, 1989; Ryckman, Kaczor, & Thornton, 1992; Snell & Godwin, 1993; Willis, 1992). B ridges (1991) created a series of questions measuring sex role expectations and rape-supportive beliefs about the victim and the perpetrator in an analogue study of beliefs about date versus stranger rape. In this study, individuals read one of three scenarios describing an incident of forced sexual intercourse between either strangers, acquaintances on a first date, or steady dating partners. Consistent with hypotheses derived from sex role socialization theory, male and female participants were more likely to incorporate sex role expectations and rape-supportive beliefs in the acquaintance rape conditions compared to the stranger rape condition. Men were also found to hold stronger rape-supportive beliefs in the steady date than in the first date or stranger conditions.

The second line of analogue research has studied the role of sex role socialization processes in attributions of marital versus stranger rape. In a study using a measure of hypermasculinity and guided imagery of stranger and marital rape scenarios, male participants, irrespective of their hypermasculinity classification, consistently minimized the seriousness of marital rape compared to stranger rape (Sullivan & Mosher, 1990). Contrary to the researchers' hypothesis, however, macho men were not more accepting of marital rape than stranger rape. High macho men were more accepting of stranger rape than low macho men, but there were no differences between these groups when considering marital rape. The authors asserted "the acceptance of the rape of wives may be a pernicious and widespread social problem and an attitude which is not limited to a subsample of macho men" (p. 285).

Monson et al. (1996) also investigated the perceptions of marital and stranger rape using two vignettes that manipulated the victim-perpetrator relationship (i.e., stranger and marriage). The questions used to assess rape-supportive beliefs and sex role expectations in the study were similar to those used by Bridges (1991). Monson et al. (1996) found that individuals were consistently more rape supportive when considering marital versus stranger rape. Their results also suggested that sex role expectations related to the perpetrator were more important in the case of marital than stranger rape (i.e., marital perpetrators were considered to be significantly more masculine and to suffer from less psychopathology).

Both of these lines of research have contributed to the understanding of society's disparate attributions of different types of rape. However, a limitation of this previous work has been the lack of a continuum of relationships between victims and perpetrators in which acquaintance rape occurs. Specifically, studies have not considered attributions about marital, date, and stranger rape concurrently. Therefore, the primary purpose of the present study was to investigate the viability of the sex role socialization theory of rape in different relationship contexts (i.e., stranger, early date, late date, and marriage). Because researchers have proposed that the likelihood of an incident being defined as rape and the tendency for perceivers to employ sex role expectations varies directly with the distance of the participant's interpersonal relationship (Klemmack & Klemmack, 1976; Koss, Dinero, Seibel, & Cox, 1988), the first hypothesis was that as the relationship between the perpetrator and victim increased in intimacy, there would be a corresponding trend in the participant's tendency to endorse rape-supportive and sex role stereotypical attributions.

A second purpose of the current research was to examine the influence of a couple's sexual intercourse history on acquaintance rape attributions. Previous research investigating attributions about date and stranger rape suggests that a prior consensual intercourse history between the victim and perpetrator functions to minimize the seriousness of date rape (L'Armand & Pepitone, 1982; Shotland & Goodstein, 1992). However, to date, few researchers have included a condition in which there is no mention of this variable to assess people's assumptions about this variable when no information is provided to them. Therefore, the second hypothesis was that the participants would endorse the most rape-supportive and sex role stereotypical attributions in the dating conditions in which they were led to believe that there was a history of sexual intercourse between the victim and perpetrator. Intermediately high rape-supportive and sex role stereotypical attributions were expected when there was no mention of the couple's sexual intercourse history, and the least rape-supportive and sex role stereotypical attributions were expected when the participants were led to believe that the couple did not have a history of sexual intercourse.

The third hypothesis tested whether the differential beliefs surrounding marital rape are in part related to an assumed history of previous consensual sexual intercourse between the victim and perpetrator or a product of something specific to the marital relationship. Given the lack of empirical research directly addressing this issue, recent changes in sexual assault statutes were the basis of this hypothesis. There has been a trend to extend traditional marital rape allowances to sexual assault statutes (i.e., husbands are not prosecuted as harshly as other types of nonhusband perpetrators of sexual assault) to nonmarried cohabitants and in some cases nonmarried individuals who have had a history of consensual sexual intercourse within the past year (Sitton, 1993). Therefore, it was hypothesized that there would be no differences in the participant's rape-supportive or sex role stereotypical attributions amongst the early and late dating conditions when the victim and perpetrator had engaged in consensual sexual intercourse and the marital condition in which consensual sexual intercourse is almost invariably assumed.

Participants

A sample of 200 undergraduates, 100 men and 100 women, from a large public university in the Midwest participated in the study. Approximately 90% of the participants were 18 to 21 years of age, and 97% of the sample were single (42% in monogamous relationship, 24% dating multiple partners, 24% not dating, and 10% other). The sample was primarily Caucasian (93%) and heterosexual (99%). An overall multivariate analysis of variance (MANOVA) revealed no significant demographic differences (i.e., age, marital status, educational status, income, and religious involvement) among the students randomly assigned to the four relationship conditions, suggesting initial equivalency across groups, Wilks's lambda: (15, 517) = .91, p > .25.

Research Design

An unbalanced mixed factorial design was used to test the hypotheses. The between-subject factor was the victim-perpetrator's relationship (i.e., stranger, early dating, late dating, and marriage). Twenty-five men and women, respectively, were randomly assigned to each of the four victim-perpetrator relationship levels. Within each of the four victim-perpetrator relationship levels, an additional situational variable was manipulated. However, because of practical considerations, the within-condition situational variable that was manipulated was not the same for each of the four relationship conditions. Specifically, the early and late dating situational variable was the couple's sexual intercourse history, the stranger situational variable was the female's general sexual intercourse history, and the marriage situational variable was the degree to which there was a prior history of physical violence in the marriage. The unbalanced nature of this design was necessary because it was impossible to construct realistic vignettes that manipulated a prior consensual sexual intercourse history between strangers or a lack of consensual sexual intercourse history between spouses. Data from the within-condition factor for the early and late dating relationship levels (i.e., no mention of the couple's sexual intercourse history, previous history, and no previous history) were analyzed for the current study.

Stimuli

Victim-perpetrator relationship manipulation (between-condition vignette). Initially, each participant was randomly assigned to one of four victim-perpetrator relationship conditions. Within these conditions, all participants read one standard between-condition vignette. There was no mention of an additional situational variable in any of these standard between-condition vignettes so as to compare attributions based only on the victim-perpetrator's relationship level. These vignettes were all identical in location, circumstance, and victim and perpetrator's actions except for the level of relationship between the victim and perpetrator. Rape by a stranger was described as perpetrated by a man whom the victim had never seen before. Early date rape was described as perpetrated by a man whom the victim had gone out with a couple of times, whereas late date rape was described as perpetrated by a man whom the victim had been dating for almost a year. Marital rape was described as perpetrated by the victim's husband.

Additional situational variable manipulation (within-condition vignette). In addition to the between-condition vignette, all participants read two more vignettes within their assigned relationship condition. The two standard additional within-condition vignettes each varied slightly in circumstance and used different victim and perpetrator's names. This was done to make each vignette seem different while decreasing the chance for demand characteristics for the within-condition manipulations.

The two standard within-condition vignettes in the early and late dating conditions were included to manipulate the victim-perpetrator's consensual sexual intercourse history. One vignette depicted the victim as having a sexual intercourse history with the perpetrator, whereas the other indicated that there was no sexual intercourse history between the victim and perpetrator.

Twenty-one independent observers rated the comparability of the three standard within-condition vignettes. Seven dimensions of each vignette were assessed, and the vignettes were found to be equivalent on all of the dimensions.(n1) Each participant randomly received one of three different orders of the three vignettes to control for potential order effects. Analyses within each relationship condition revealed no evidence of order effects.(n2)

In all vignettes, names such as Jenny and Kevin were used, versus man and woman or victim and perpetrator, to avoid bias and to enhance readability. Also, the woman was described as "persistently resisting the sexual interaction" in each vignette to characterize the nonconsensual nature of the interaction. The phrase completed the act of sexual intercourse was used versus raped or sexually assaulted to avoid bias associated with either of these phrases. Consistent with previous research (Tetreault & Barnett, 1987), there was no mention of the type or level of violence engaged in to complete the rape so as to draw on the participants' beliefs about sexual violence rather than their attitudes about physical assault.

Procedure

The participants were recruited from a participant pool made up of general psychology students who received partial course credit for their participation in the study. All participants were told that the researchers were interested in their perceptions of various heterosexual social interactions. The participants completed the self-report packets in groups of approximately 15 to 20 students to facilitate anonymity, and they were informed that all of their answers would be completely confidential. After obtaining informed consent, the participants were randomly assigned to one of the four relationship conditions and completed the self-report packet. At the conclusion of the study, participants were debriefed and asked to refrain from discussing the study for a minimum of 1 month to reduce the likelihood of influencing other students who might later be participants in the study.

Measures

After reading each of the three vignettes in their packet, the participants were asked to answer a number of questions assessing their attributions about the sexual interaction on 10-point rating scales of intensity anchored by 1 = minimal and 10 = maximum. Consistent with previous research, these questions comprised two scales (Bridges, 1991; Monson et al., 1996).

Rape-Supportive Attributions Scale (RAPE-SUPPORT). Rape-supportive attributions were assessed with the following four questions: (a) How violent do you feel this situation was? (b) How psychologically damaged do you feel "Jenny" will be from this experience? (c) To what degree were "Kevin's" actions a violation of "Jenny's' rights? and (d) How certain are you that this incident would be considered rape? The participants' responses to these four questions were reverse-scored and summed to create the RAPE-SUPPORT scale. Higher scores on this scale reflect the endorsement of greater rape-supportive attributions. The alpha reliability coefficient for this scale within the present sample was .82.

Sex Role Stereotypical Victim Blame Attributions Scale (BLAME). The participants' sex role stereotypical attributions about the victim's blame were assessed with the following four questions: (a) How much control did "Jenny" have in this situation? (b) How much did "Jenny" enjoy this situation? (c) How obligated was "Jenny" to engage in sexual relations in this case? and (d) How interested was "Jenny" in having sexual relations? The participants' responses to these four questions were summed to create the BLAME scale. Higher scores on this scale reflect the endorsement of greater sex role stereotypical attributions about the victim's blame. The alpha reliability coefficient for this scale within the present sample was .64.

Data Analysis Procedures

A series of MANOVAs, using the RAPE-SUPPORT and BLAME scales as dependent variables, were conducted to test the hypotheses regarding the: (a) influence of the victim-perpetrator relationship on rape attributions, (b) influence of a couple's history of sexual intercourse within dating relationships on rape attributions, and (c) role of an assumed sexual intercourse history in attributions about marital rape. Significant multivariate effects were followed up with univariate ANOVAs, and Tukey's Honestly Significant Differences group comparisons were then used for significant univariate effects. An error rate of p < .05 was adopted for all of the analyses.

Influence of the Victim-Perpetrator Relationship on Rape Attributions

The mean ratings for the four standard between-condition vignettes (i.e., stranger, early dating, late dating, and married) were analyzed via a 2 (participant's gender) x 4 (relationship level) between-groups MANOVA, using the RAPE-SUPPORT and BLAME scales as the dependent variables. As hypothesized, a significant main effect for relationship condition was found, Wilks's lambda: (6, 382) = .73,p < .0001. Although no significant main effect for participant's gender was found, Wilks's lambda: (2, 191) = .98, p = .22, a significant relationship by participant's gender multivariate interaction effect was revealed, Wilks's lambda: (6, 382) = .93, p < .05.

Rape-supportive attributions. As predicted and shown in Table 1, there was a significant linear trend in the participants' RAPE-SUPPORT scores across the relationship levels, with lowest scores obtained in the stranger condition and the highest scores for the marital condition, F(3, 192) = 25.17, p < .0001. Furthermore, a significant relationship by participant's gender interaction effect was found for the RAPE-SUPPORT scale, F(3, 192) = 2.60, p < .05. As shown in Figure 1, women rated the stranger and early dating conditions similarly on the RAPE-SUPPORT scale. Their RAPE-SUPPORT scores for the late dating and marital conditions were also similar and significantly higher for these conditions compared to the stranger and early dating conditions. In contrast, the men's scores on the RAPE-SUPPORT scale were lowest for the stranger and late dating conditions (not significantly different from each other), significantly higher for the early dating condition, and highest in the marital condition.

Sex role stereotypical victim blame attributions. Univariate follow-ups revealed that the participants blamed the victim in the stranger condition least, made intermediately blaming attributions in the early and late dating conditions (not statistically different from each other), and blamed the victim the most in the marital condition, F(3, 192) = 6.07, p < .001 (see Table 1). Furthermore, a significant relationship by participant's gender interaction effect was found, F(3, 192) = 3.62, p < .01. Similar to the relationship by participant's gender interaction found for the RAPE-SUPPORT scores, women rated the stranger and early dating conditions similarly on the BLAME scale. Their BLAME scores for the late dating and marital conditions were also similar and significantly higher for these conditions compared to the stranger and early dating conditions. The men's scores on the BLAME scale were lowest for the stranger and late dating conditions (not significantly different from each other), significantly higher for the early dating condition, and highest in the marital condition (see Figure 2).

Influence of a Couple's Sexual Intercourse History on Date Rape Attributions

As hypothesized, a 2 (participant's gender) x 2 (early and late dating relationship levels) x 3 (couple's sexual intercourse history) mixed factorial MANOVA, with the RAPE-SUPPORT and BLAME scales as dependent variables, revealed significant main effects for the dating couple's sexual intercourse history, Wilks's lambda: (4, 93) = .68, p < .0001, and participant's gender, Wilks' lambda: (2, 95) = .91, p < .01. No main effect for length of dating relationship was found, Wilks's lambda: (2, 95) = .95, p = .11. However, a significant three-way interaction was found, Wilks's lambda: (4, 93) = .89, p < .05. None of the two-way multivariate interaction effects were significant.

Rape-supportive attributions. As predicted and shown in Table 1, participants endorsed the most rape-supportive attributions when they were told that the couple had a previous history of consensual sexual intercourse as compared to the other two conditions (i.e., no mention and no previous history), F(2, 192) = 18.12, p < .0001. However, mean ratings on the RAPE-SUPPORT scale did not differ significantly between the no mention and the no previous history conditions. Men in general attributed less seriousness to rape in the dating relationships than did women, F(1, 96) = 6.28, p < .01 (see Table 2). Follow-up univariate analysis of the significant multivariate three-way interaction was not significant for the RAPE-SUPPORT scale, F(2, 192) < 1, p = .45.

Victim blame attributions. As hypothesized and shown in Table I, participants blamed the victim the most when they were told that the couple had a previous history of consensual sexual intercourse as compared to the other two conditions (i.e., no mention and no previous history), F(2, 192) = 8.77, p < .0001. However, mean ratings on the BLAME scale did not differ significantly between the no mention and no previous history conditions. In general, men attributed more blame to the victim in these date rapes than did the women, F(1, 96) = 8.09, p < .01 (see Table 2). Follow-up univariate analysis of the significant multivariate three-way interaction was not significant for the BLAME scale, F(2, 192) = 2.171, p = .12.

The Role of an Assumed Sexual Intercourse History in Attributions About Marital Rape

To determine if the differential attributions surrounding marital rape were the result of a presumed history of sexual intercourse between the victim and perpetrator, attributions from the marital rape vignette in which no additional situational variable was manipulated were compared to the early and late date rape conditions in which participants were informed the victim and perpetrator had a sexual intercourse history. A 2 (participant's gender) x 3 (history of sexual intercourse in early and late dating, marital between-condition) between-groups factorial MANOVA revealed a significant main effect for relationship level, Wilks's lambda: (4, 286) = .93, p < .05, and participant's gender, Wilks's lambda: (2, 143) = .92, p < .01. No significant multivariate interaction effect was found, Wilks's lambda (4, 286) = .96, p = .25.

Rape-supportive attributions. Univariate follow-ups did not reveal differences in the participants' ratings on the RAPE-SUPPORT scale across the relationship conditions, F(2, 144) < 1, p = .67. As shown in Table 2, in general, men reported higher RAPE-SUPPORT scores across all three consensual sexual intercourse history vignettes compared to women, F(1,144) = 6.61, p < .01.

Sex role stereotypical attributions. Again, univariate follow-ups did not reveal differences in the participants' ratings on the BLAME scale across the relationship conditions, F(2, 144) = 1.50, p = .23. As shown in Table 2, in general, men reported higher BLAME scores compared to women, F(1, 144) = 11.88, p < .001.

The present results confirm and extend two parallel lines of research investigating attributions about acquaintance rape by investigating a continuum of victim-perpetrator relationships. Consistent with hypotheses derived from the sex role socialization theory of rape, the perceived seriousness of the sexual assault (i.e., rape-supportive attributions) decreased in a linear fashion as the level of acquaintance between the victim and perpetrator increased. Participants were more likely to make sex role stereotypical attributions about the victim's blame in cases of acquaintance rape and particularly marital rape when compared to stranger rape. These results were found although the victim and perpetrator's actions were described exactly the same way in the four relationship conditions.

This study also elucidated an important factor in people's tendency to make rape-supportive and sex role stereotypical attributions in cases of acquaintance rape: the couple's sexual intercourse history. Irrespective of gender, when the participants were informed that the dating couple had consensual sexual intercourse in the past, the participants made more rape-supportive and sex role stereotypical attributions about the rape. No differences were found when there was no mention of this variable and when told that there was no history of consensual sexual intercourse. Moreover, when the participants were informed that there had been a history of sexual intercourse between the victim and perpetrator in the dating relationships and these vignettes were compared to the marital relationship vignette, there were no statistical differences in the participants' rape-supportive or sex role stereotypical attributions.

These results are encouraging because they suggest that people do not automatically assume that because a couple is dating they have engaged in sexual intercourse and consequently that rape within that relationship is less serious and the victim is more blameworthy. On the other hand, these results suggest that once a female consents to sexual intercourse in a relationship and this is known, people's perceptions of the seriousness of the rape decreases as the victim's culpability increases. Also, these analyses argue that the differential attributions surrounding marital rape are not necessarily a product of the legal institution of marriage, the privacy accorded this relationship, or based on the historical notion that women are the sexual chattel of their husbands. Instead, consistent with Shotland and Goodstein's (1992) research on date rape, these analyses highlight how powerful a factor a couple's previous sexual intercourse history appears to be in people's attributions about acquaintance rape. Shotland and Goodstein (1992) found that individuals consistently minimized the seriousness of the rape and attributed more blame to the victim when the victim was reported to have engaged in sexual intercourse with her dating partner on prior occasions. These authors argued that a history of sexual relations within a relationship establishes a "sexual precedence" that leads to the belief that each member of the couple is entitled to a legitimate claim on the other's sexuality. They argued that this sexual precedence applies more generally to consensual sexual relations, and to men and women alike, in dating relationships. Using this framework, women are considered no more the sexual property of their partners than vice versa.

Two points should be raised regarding the gender neutrality of implied consent to sexual relations in intimate relationships. First, women in general were significantly less rape supportive and sex role stereotypical in their attributions across the intimate relationships. These findings are consistent with sex role socialization theory and previous research (Barnett et al., 1992; Szymanski et al., 1993); however, there was a gender by relationship level interaction found with men's attributions. Of note, men made the least rape-supportive and sex role stereotypical attributions in the stranger and late dating relationships, intermediately rape-supportive and sex role stereotypical attributions in the early dating condition, and the most rape-supportive and sex role stereotypical attributions in the marital condition. One possible explanation for these findings could be that in the early dating situation, men make more sex role stereotypical attributions because they consider the female victim to be more permissive, or even promiscuous, in allowing the perpetrator any type of contact (i.e., blaming the victim), whereas they consider rape to be more of a breach of trust and understanding within the context of a long-term relationship. In contrast, women are less sure of the role of the victim's consent and revocation of consent as the relationship intimacy increases. Further research is needed to corroborate and expand these findings. There were also significant differences in the men's and women's attributions when the participants were explicitly told there was a history of sexual intercourse between the victim and perpetrator. The current results, as well as previous research (Shotland & Goodstein, 1992), suggest that women are more likely to consider a rape a rape regardless of the sexual intercourse history or status of the relationship. Moreover, these gender differences may contribute to the gender differences found in the rates of rape perpetration.

Second, legally and psychologically, sexual precedence in a relationship should not preclude a right to change the precedence. Practically, however, a woman saying "no" to sexual relations in an intimate relationship after she has said "yes" might not necessarily carry the same meaning for men and women.

One limitation of the present study is the use of college-aged individuals. However, the high incidence of acquaintance rape in this cohort suggests that this is an appropriate population for study (Proite et al., 1993). Future researchers in this area might recruit individuals from the community at large or from jury rolls to assess the attributions of older men and women. It might be hypothesized that a community sample, with diverse levels of education and age, would hold even greater rape-supportive and sex role stereotypical attributions about acquaintance rape. Furthermore, it may be useful to recruit individuals from diverse ethnocultural backgrounds. Past studies have found racial differences in the acceptance of rape myths (Giacopassi & Dull, 1986).

Researchers in this area should consider the difficulties of independent variable manipulation and dependent variable measurement. One common criticism of rape attribution research is the use of written vignette stimuli and the generalizability of results. In this particular study, it was difficult to construct vignettes that provided a realistic and equivalent context for the victim and perpetrator's interaction across the different relationships. For example, people might make different attributions in the case of a woman opening a door for her husband versus a stranger. Future research using alternative modalities for presenting the stimuli (e.g., videotape or audiotape) and studies specifically addressing the construction of stimuli for this type of research are needed. As with most research in this area, the potential for extreme responding is high and can potentially lead to ceiling effects in statistical analyses. On the other hand, the fact that rape-supportive and sex role stereotypical attributions were relatively low in the present study is encouraging. Also, although no effect for the order in which the vignettes were presented was found in the present within-subject research design, future between-group studies should also be undertaken to address the replicability of the present findings.

Because many states have allowances to their sexual assault statutes based on current marital status (Sitton, 1993), vignettes manipulating the spouse's living arrangements and/or current marital status are an important avenue for future research. The addition of this situational variable would help determine when perceivers consider implied consent to sexual relations between spouses to end (e.g., separation and divorce). An additional situational variable worthy of research attention is the level of violence depicted in the scenarios. Previous studies have shown the level of violence to influence people's attributions about the seriousness of acquaintance rape (Schneider, Ee, & Aronson, 1994). In this study, participants were left to their own assumptions about how violent the nonconsensual sexual intercourse would be in each of the situations. Also, when manipulating the couple's prior sexual intercourse history in the early and late dating conditions in the present study, the participants were told only whether they had previously engaged in consensual sexual intercourse. Given the previously found influence of the number of sexual intercourse experiences on attributions of date rape (Shotland & Goodstein, 1992), future research manipulating the amount of times the couple had previously engaged in sexual intercourse is warranted.

The differential attributions about stranger, dating, and marital rape found in the current study have a number of prevention and sociolegal implications. The influence of a couple's history of sexual intercourse on attributions about rape, especially for men, appears to be a critical topic for rape awareness/prevention workshops. One study in this area has shown that these workshops have more impact on changing men's beliefs about hypothetical acquaintance rape victims and perpetrators than women's, perhaps because of women's already high antirape attributions (Szymanski et al., 1993).

Fortunately, all states have abolished the traditional marital rape exemption to their sexual assault statutes that allowed husbands to rape their wives with immunity (Sitton, 1993). However, a number of jurisdictions have replaced the traditional exemption with marital rape allowances, which create a lesser crime for wife rape than for other rapes. Consequently, states adopting these allowances have maintained the legal right of a husband to rape his wife under certain circumstances. An additional disturbing trend has been for states to expand these allowances for marital rape perpetrators to some unmarried perpetrators who are intimately involved with their victims prior to the rape. For example, several states have made allowances for perpetrators who are cohabiting with their victims but not married to them. In Delaware, if a rape victim has engaged in voluntary sexual intercourse with the assailant at some point during the year preceding the assault, the defendant cannot be prosecuted under the first-degree statue unless the victim suffers "serious injury" (Sitton, 1993). This trend of considering implied consent to current sexual relations because of past sexual relations reinforces the minimization of marital and nonmarital acquaintance rape.

The disparate attributions society holds for acquaintance rape may account for the tendency for victims to avoid reporting their crimes and seeking medical and psychological assistance. Furthermore, individuals who do report the crime and enter the criminal justice system may still experience the impact of these false beliefs about intimate rape. For example, acquaintance rape cases appear to be difficult to bring beyond the investigatory phase, perhaps because police officers are vulnerable to the same faulty attributions about acquaintance rape as is society in general. Prosecutors' personal biases about acquaintance rape may affect their decision to prosecute, and they may be dissuaded from prosecuting such cases because of the difficulties in obtaining convictions (Augustine, 1991). In fact, one study found prosecutors were significantly less likely to believe that maximum charges would be filed in marital rape cases than in identical rape cases involving strangers, particularly when no serious injury occurred (Jeffords, 1984). Moreover, these social attributions potentially influence judges' decisions regarding bail setting and the admissibility of evidence. Jurors holding these attributions may be reluctant to convict acquaintance rapists and more likely to hand down shorter prison sentences in these trials (Augustine, 1991). A final and important implication is that these attributions presumably aid in the perpetuation of sexual violence against women.

Authors' Note: Correspondence concerning this article should be addressed to Candice M. Monson, Ph.D., who is now at Dartmouth Medical School, White River Junction VA Medical Center, Mental Health/Behavioral Sciences (116-E), 215 N. Main, White River Jct., VT, 05009; telephone: (802) 295-9363, ext. 6072; e-mail: Candice.M.Monson@Dartmouth.edu.

(n1.) The equivalency of the three standard vignettes (i.e., between-and within-condition vignettes) was determined by 21 independent raters. There was no mention of the relationship of the perpetrator to the victim, and no additional situational variables were manipulated in the vignettes. This was done to assess only the equivalency of the context of the three vignettes. Each rater randomly received one of the three vignettes and was asked to rate on 10-point rating scales of intensity (1 = minimal and 10 = maximum) the vignette on the following dimensions: readability, believability, victim likability, perpetrator likability, level of victim risk, likelihood of the event occurring in the real world, and level of violence. Nonparametric comparisons (i.e., Kruskal-Wallis tests using chi-square analyses of the mean ratings of the vignettes) revealed no differences in the observers' ratings of the three vignettes on any of the dimensions: readability, Chi[sup 2](2, N= 21) = .27, p = .87; believability, Chi[sup 2](2, N = 21) = .77, p = .68; victim likability, Chi[sup 2](2, N = 21) = .91, p = ,64; perpetrator likability, Chi[sup 2](2, N = 21) = 2.19, p = .33; level of victim risk, Chi[sup 2](2, N = 21) = 4.01, p = .13; likelihood of the event occurring in the real world, Chi[sup 2](2, N = 21) = 2.10, p = .35; and level of violence, Chi[sup 2](2, N = 21) = 1.14, p = .57. All vignettes used in this study are available from the first author.

(n2.) The potential for order effects was assessed within each of the relationship levels using the RAPE-SUPPORT scale. A series of 3 (order of vignette presentation) x 3 (additional situational variable) MANOVAs revealed no main effect for order: stranger, Wilks's lambda: (8, 88) = .99, p = .96; early dating, Wilks's lambda: (8, 88) = .85, p = .50; late dating, Wilks's lambda: (8, 88) = .85, p = .50; marriage, Wilks's lambda: (8, 88) = .85, p = .47. There were also no order by additional situational variable interaction effects in any of the relationship levels: stranger, Wilks's lambda: (16, 80) = .90, p =. 18; early dating, Wilks's lambda: (16, 80) = .65, p = .28; late dating, Wilks's lambda: (16, 80) = .69, p = .45; and marriage, Wilks's lambda: (16, 80) = .76, p = .74. Thus, it was concluded that the order in which the vignettes were presented to the participants was not a significant factor in their ratings.

Legend for Chart:

A - Relationship Condition
B - Sexual Intercourse History No Information Provided M
C - Sexual Intercourse History No Information Provided SD
D - Sexual Intercourse History No Previous Intercourse
    History M
E - Sexual Intercourse History No Previous Intercourse
    History SD
F - Sexual Intercourse History Previous Intercourse History
    M
G - Sexual Intercourse History Previous Intercourse History
    SD

A                       B        C              D         E
                                                F         G

RAPE-SUPPORT scale

Stranger              5.40   2.19[sup a]         --
                                                 --

Early dating          8.38   5.21[sup b,1]     8.90   5.67[sup 1]
                                              11.68   9.10[sup 2]

Late dating           9.62   5.33[sup c,1]     8.34   4.07[sup 1]
                                              13.01   5.33[sup 2]

Marriage             13.22   6.59[sup d]         --
                                                 --
BLAME scale

Stranger              6.58   3.04[sup a]         --
                                                 --

Early dating          7.26   3.72[sup b, l]    8.26   4.63[sup 1]
                                              10.86   6.21[sup 2]

Late dating           7.00   3.05[sup b, 1]    7.24   3.58[sup 1]
                                               9.42   5.34[sup 2]

Marriage              9.54   5.22[sup c]         --
                                                 --

NOTE: Higher scores on the RAPE-SUPPORT and BLAME scales
represent greater endorsement of rape-supportive and sex role
stereotypical blame attributions. Each relationship condition
included 50 (25 men, 25 women) randomly assigned participants.
Different superscripts denote statistically significant mean
differences (alphabet superscripts are related to comparisons
of victim-perpetrator relationship levels, and numeric
superscripts are related to comparisons of sexual intercourse
histories).

Legend for Chart:

A - Scale
B - Dating Relationships Only[a] Men (N = 50) M
C - Dating Relationships Only[a] Men (N = 50) SD
D - Dating Relationships Only[a] Women (N = 50) M
E - Dating Relationships Only[a] Women (N = 50) SD
F - Dating and Marital Relationships[b] Men (N = 75) M
G - Dating and Marital Relationships[b] Men (N = 75) SD
H - Dating and Marital Relationships[b] Women (N = 75) M
I - Dating and Marital Relationships[b] Women (N = 75) SD

A                       B        C        D        E
                        F        G        H        I

RAPE-SUPPORT          11.39     6.37     8.77     5.65
                      14.33     8.02    11.06     7.42

BLAME                  9.45     5.06     7.36     3.22
                      11.51     6.34     8.47     4.16

[a.] Means and standard deviations from analyses in which the
couple's sexual intercourse history was examined in the early
and late dating conditions.

[b.] Means and standard deviations from analyses in which the
early and late dating conditions in which the couple was
reported to have a history of sexual intercourse were compared
to the marital condition in which there was no mention of an
additional situational variable.

Legend for Chart:

A - Victim-Perpetrator Relationship
B - Males
C - Females

A                  B       C

Stranger          5.28    5.32
Early Dating     10.17    6.07
Late Dating       8.74   12.21
Marital          13.4    13.56

Legend for Chart:

A - Victim-Perpetrator Relationship
B - Males
C - Females

A                  B       C

Stranger          5.89    6.98
Early Dating      8.96    5.95
Late Dating       6.51    8.03
Marital          10.88    8.91

Augustine, R. (1991). Marriage: The safe haven for rapists. Journal of Family Law, 29, 559-590.

Barnett, M. A., Quackenbush, S. W., Sinisi, C. S., Wegman, C. M., & Otney, K. L. (1992). Factors affecting reactions to a rape victim. Journal of Psychology, 126, 609-620.

Bell, S. T., Kuriloff, P. J., & Lottes, I. (1994). Understanding attributions of blame in stranger rape and date rape situations: An examination of gender, race, identification, and students' social perceptions of rape victims. Journal of Applied Social Psychology, 24, 1719-1734.

Bridges, J. S. (1991). Perceptions of date and stranger rape: A difference in sex role expectations and rape-supportive beliefs. Sex Roles, 24, 291-307.

Bridges, J. S., & McGrail, C. A. (1989). Attributions of responsibility for date and stranger rape. Sex Roles, 21, 273-286.

Butt, M. R. (1980). Cultural myths and supports for rape. Journal of Personality and Social Psychology, 38, 217-230.

Check, J.V.P., & Malamuth, N. M. (1983). Sex role stereotyping and reactions to depictions of stranger versus acquaintance rape. Journal of Personality and Social Psychology, 45, 344-356.

Coller, S. A., & Resick, P. A. (1987). Women's attributions of responsibility for date rape: The influence of empathy and sex-role stereotyping. Violence and Victims, 2, 115-125.

Giacopassi, D. J., & Dull, R. T. (1986). Gender and racial differences in the acceptance of rape myths within a college population. Sex Roles, 15, 63-75.

Jeffords, C. R. (I 984). Prosecutorial discretion in cases of marital rape. Victimology, 9, 415-425.

Johnson, J. D., & Jackson, L. A. (1988). Assessing the effects of factors that might underlie the differential perception of acquaintance and stranger rape. Sex Roles, 19, 37-45.

Klemmack, S. H., & Klemmack, D. L. (1976). The social definition of rape. In M. J. Walker & S. L. Brodsky (Eds.), Sexual assault (pp. 135-147). Lexington, MA: D. C. Health.

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

L'Armand, K., & Pepitone, A. (1982). Judgments of rape: A study of victim-rapist relationship and victim sexual history. Personality and Social Psychology Bulletin, 8, 134-139.

Monson, C. M., Byrd, G. R., & Langhinrichsen-Rohling, J. (1996). To have and to hold: Perceptions of marital rape. Journal of Interpersonal Violence, 11, 410-424.

Perry, L., McLendon, K., & Foley, L. (1994). Male and female perceptions of date rape. Journal of Social Behavior and Personality, 9, 421-428.

Proite, R., Dannells, M., & Benton, S. L. (1993). Gender, sex-role stereotypes, and the attribution of responsibility for date and acquaintance rape. Journal of College Student Development, 34, 411-417.

Quackenbush, R. L. (1989). A comparison of androgynous, masculine sex-typed and undifferentiated males on dimensions of attitudes toward rape. Journal of Research in Personality, 23, 318-342.

Ryckman, R. M., Kaczor, L. M., & Thornton, B. (1992). Traditional and nontraditional women's attributions of responsibility to physically resistive and nonresistive rape victims. Journal of Applied Social Psychology, 22, 1453-1463.

Schneider, L. J., Ee, J.S.C., & Aronson, H. (1994). Effects of victim gender and physical versus psychological trauma/injury on observers' perceptions of sexual assault and its aftereffects. Sex Roles, 30, 793-808.

Shotland, R. L., & Goodstein, L. (1992). Sexual precedence reduces the perceived legitimacy of sexual refusal: An examination of attributions concerning date rape and consensual sex. Personality and Social Psychology Bulletin, 18, 756-764.

Sitton, J. (1993). Old wine in new bottles: The "marital" rape allowance. North Carolina Law Review Association, 72, 261-289.

Snell, W. E., & Godwin, L. (1993). Social reactions to depictions of casual and steady acquaintance rape: The impact of AIDS and stereotypical beliefs about women. Sex Roles, 29, 599-616.

Sullivan, J. P., & Mosher, D. L. (1990). Acceptance of guided imagery of marital rape as a function of macho personality. Violence and Victims, 5, 275-286.

Szymanski, L. A., Devlin, A. S., Chrisler, J. C., & Vyse, S. A. (1993). Gender role and attitudes toward rape in male and female college students. Sex Roles, 29, 37-57.

Tetreault, P. A., & Barnett, M. A. (1987). Reactions to stranger and acquaintance rape. Psychology of Women Quarterly, 11, 353-358.

Willis, C. E. (1992). The effect of sex role stereotype, victim and defendant race, and prior relationship on rape culpability attributions. Sex Roles, 26, 213-226.

~~~~~~~~

By Candice M. Monson; Jennifer Langhinrichsen-Rohling and Tisha Binderup

Candice M. Monson, Ph.D., is an assistant professor of psychiatry at Dartmouth Medical School. She is a graduate of the University of Nebraska--Lincoln Clinical Psychology Training Program and completed a postdoctoral forensic fellowship at the University of Missouri--Kansas City School of Medicine. Her specific research interests are in trauma, typologies of intimate violence perpetrators, the integration of sexual and nonsexual violence perpetration, and rape attributions

Jennifer Langhinrichsen-Rohling, Ph.D., is an associate professor in the Department of Psychology at the University of South Alabama. Her research interests include marital violence, marital rape, and the long-term consequences of growing up in a violent family context. She has published extensively on these topics

Tisha Binderup received her B.A. in psychology from the University of Nebraska. Portions of this research project served as her undergraduate honors project.