Perceptions of sexual coercion in heterosexual dating relationships: the role of aggressor gender and tactics.

Sexual coercion is a significant problem in college dating relationships. Shook, Gerrity, Jurich, and Segrist (2000) found that in a college sample, 82% reported using verbally coercive and 21% reported using physically coercive behaviors against a dating partner in the last year. Other research has found similarly high rates of sexual coercion (e.g., Jenkins & Aube, 2002; Koss & Oros, 1982; Russell & Oswald, 2001, 2002). Though much of the initial research focused on men as aggressors and women as victims, and statistics indicate that this situation is a major social problem, recently researchers have begun to acknowledge that women are also coercive in their intimate relationships (e.g., Anderson & Struckman-Johnson, 1998; Archer, 2000). This growing body of research suggests that both men and women report engaging in a wide range of coercive tactics to obtain sex within the context of a dating relationship. The goal of this study is to examine participants' perceptions of the actors and their behaviors when they are sexually coercive on a date.

In this study, sexual coercion is conceptualized to include a variety of tactics used to obtain sexual acts or intercourse from an unwilling partner. The behaviors range on a continuum that includes verbal pressure and threats and the use of physical force to obtain sexual acts, such as kissing, petting, or intercourse. Research on college heterosexual dating relationships has found a high rate of coercion and aggression perpetrated against women. Koss and Oros (1982) found that in their sample of college men, 23% reported obtaining sexual intercourse by threatening to end the relationship, 20% reported using some degree of physical force to obtain sex acts, and 3% reported having used physical force to obtain intercourse. More recently, Russell and Oswald (2002) found similar results; 36% of college men in their sample reported engaging in at least one sexually coercive behavior in a dating relationship.

While the research on male aggressors in intimate relationships is fairly extensive, the research examining female aggressors is more limited. However, contrary to initial stereotypes of men as aggressors and women as victims, there is evidence suggesting that women are also coercive in intimate relationships (e.g., Anderson, 1998; Anderson & Struckman-Johnson, 1998; Clements-Shreiber, Rempel, & Desmarais, 1998; Krahe, Waizenhofer, & Moiler, 2003; Lottes & Weinberg, 1996; Russell & Oswald, 2001; Shea, 1998; Struckman-Johnson, 1988). Anderson found that between 26% and 36% of women in their sample reported using sexually coercive behaviors, including use of continual arguments, threats to end the relationship, threats of physical force, actual physical force, and using a weapon to obtain sex acts from a male partner. Obtaining sexual intercourse by getting the male partner drunk or stoned was also a common strategy. Russell and Oswald (2001) found that 18% of women in a college sample reported engaging in sexually coercive behaviors, ranging from verbal threats and pressure to use of physically aggressive tactics.

Men also reported engaging in unwanted sexual activities because of their female partners' tactics (e.g., Russell & Oswald, 2002; Struckman-Johnson, 1988; Struckman-Johnson & Struckman-Johnson, 1998). The prevalence statistics for men as recipients of women's sexual coercion vary depending on the sample and methods used to measure and define sexual coercion; however, a review of the research suggests that between 4% and 44% of men report being sexually coerced in some manner (Struckman-Johnson & Struckman-Johnson, 1998). For example, Struckman-Johnson and Struckman-Johnson (1998) found that 43% of college men sampled reported experiencing a coercive incident, of which 36% reported unwanted touch and 27% reported being coerced into sexual intercourse. Russell and Oswald (2002) found that 44% of college men in their sample reported being the recipient of a sexually coercive tactic. The most common verbal strategies experienced included having unwanted sex because of the partner's continual arguments, lies, threats to end the relationship, or threats of physical force. Men also reported having unwanted sex because of physical strategies, such as the partner becoming so sexually aroused it was useless to stop, having sex because the partner gave them drugs or alcohol, or being the victim of attempted or completed sexual intercourse because of physical force. Across the research studies, there is some similarity in reports by college men and women. Both men and women reported that the two most common coercive behaviors were use of continual verbal pressure and refusing to stop when requested (O'Sullivan, Byers, & Finkelman, 1998; Struckman-Johnson & Struckman-Johnson, 1998). This growing body of research indicates that both men and women use a range of sexually coercive behaviors in their dating relationships. To understand the problem of sexual coercion in dating relationships, it is important to acknowledge that both men and women can be coercive.

Given the prevalence of these tactics in heterosexual dating relationships, one might wonder if college students actually perceive these behaviors as sexually coercive and problematic. It might be the case that some level of coercion has become the social norm for college dating relationships. This may be particularly true of the strategies that might be perceived as "lower level" coercion, such as use of verbal pressure or female-perpetrated strategies. Previous research has found that the specific strategy used is an important factor in determining perceptions of aggressiveness. As the amount of physical force increases, the behavior is more likely to be perceived as coercive (Abbey, BeShears, Clinton-Sherrod, & McAuslan, 2004; Kahn, Jackson, Kully, Badger, & Halvorsen, 2003). In contrast, the involvement of alcohol reduces the likelihood that people hold the aggressor responsible for his or her actions (Hammock & Richardson, 1993) and rate the behavior as assault (Kahn et al., 2003). Furthermore, intoxicated victims, compared to sober counterparts, are more likely to be blamed and held responsible for the incident (Hammock & Richardson; Richardson & Campbell, 1982; Stormo, Lang, & Stritzke 1997). Comparing acceptability rates for various sexually coercive strategies, Struckman-Johnson and Struckman-Johnson (1991) found that verbal pressure and stimulation were rated as more acceptable than use of mock force, intoxication, and physical force. Physical force and intoxication were seen as equally unacceptable. However, men were more accepting of behaviors that were perpetrated by the woman rather than the man, while women were not influenced by the gender of the aggressor.

There is also evidence that people judge men and women differently when they engage in the same aggressive behaviors. A consistent finding is that coercion from men toward women, compared to women's coercion toward men, is considered to be least acceptable and viewed most negatively (Hannon, Hall, Nash, Formati, & Hopson, 2000; Harris, 1991; Harris & Knight-Bohnhoff, 1996). Likewise, people are more likely to label the experience as rape when the man is the aggressor and the woman is the victim than when the woman is the aggressor and man is the victim (Hannon et al., 2000). It has been suggested that men, as compared to women, are perceived as more threatening and capable of doing harm, thus accounting for aggressor gender effects (Struckman-Johnson & Anderson, 1998). In research using vignettes, when women engage in coercive behaviors, they are more likely to be seen as acting romantically, seductively, expressing attraction, and generally rated as less aggressive than are men who use the same behaviors (Struckman-Johnson & Struckman-Johnson, 1991). However, men are judged to be acting aggressively and inappropriately. Researchers have found a general tendency to trivialize aggression from a woman directed toward men (Archer, 2000; Frieze, 2000). For instance, Smith, Pine, and Hawley (1988) found that, compared to female victims, male victims were more likely to be viewed as encouraging the aggressor, feeling less stress, and finding the experience more pleasurable.

Women, similar to men who use these tactics, may not even label their behaviors as coercive. Shea (1998) found that women who had verbally coerced the partner into unwanted sex did not appear to believe that their behaviors were wrong or inappropriate. Shea concluded that these women "simply expect and accept coercion and force as a natural part of some relationships" (p. 101). Furthermore, there is some evidence suggesting that men may not see themselves as victims when they are recipients of women's sexually coercive advances. When imagining hypothetical unwanted heterosexual touch, men, compared to women, respond less negatively and in some cases even slightly positively (Black & Gold, 2003; Struckman-Johnson & Struckman-Johnson, 1993, 1998). In sum, there is a tendency to view women's aggression as less serious than men's aggression. This is unfortunate, as it can result in the denial that men, too, can be victims of sexual coercion (Macchietto, 1998; Muehlenhard, 1998).

Differences in the interpretations of women's and men's aggression may be due in part to sexual scripts of how men and women should act in intimate relationships (Clements-Schreiber et al., 1998; Krahe, 2000). There is double standard that it is not appropriate for women to initiate sex (see Crawford & Popp, 2003). Struckman-Johnson and Struckman-Johnson (1991) hypothesized that women's use of coercion would be viewed as more acceptable than is men's, based on sex-role norms. Specifically, they argued that in traditional sex-role scripts, men are responsible for initiating sexual intercourse, and therefore, their coercive behaviors would be interpreted as exploitive and aggressive. In contrast, women's advances, which violate traditional sex roles, would be interpreted at the individual level and be attributed to romantic attraction. Their research found that men, but not women, were more accepting of female-initiated sexual coercion. It has also been suggested that women's coercion is viewed less negatively due to the common belief that women are physically less able to inflict harm then are men (e.g., Hams & Knight-Bohnhoff, 1996). Alternatively, it has been argued that women who are sexually coercive will be rated especially negatively for violating social norms (e.g., Crawford & Popp). Thus, there are two potential reactions to men and women who engage in sexually coercive strategies. This research will further investigate this topic.

Study Overview

We studied the way college students interpret sexually coercive behaviors of varying strategy when perpetrated by men and women. Participants were randomly assigned to read one scenario about a heterosexual college-student dating couple in which the strategy and aggressor gender was manipulated. The sexually coercive strategies included verbal pressure where the aggressor threatened to end the relationship unless they had sex, purposeful intoxication, and physical force. In the control condition, mutually consensual sexual intercourse occurred. Participants rated the man and the woman on trait characteristics degree to which the person was acting romantically, promiscuously, aggressively, and victimized), the coercive behavior (degree to which the behavior was physically, sexually, and emotionally aggressive), and perceptions of both actors' satisfaction with the relationship. Our overriding research question was, how does the aggressor gender and coercive strategy influence perceivers' ratings of the victim, the aggressor, the behavior, and perceptions of relationship satisfaction?

Given previous findings about the importance of strategy for perceptions of aggression (e.g., Abbey et al., 2004; Kahn et al., 2003; Struckman-Johnson & Struckman-Johnson, 1991), we expected that physical coercion would be rated as most aggressive, followed by alcohol intoxication, verbal aggression, and the control condition. However, perceptions of these behaviors may be qualified by the gender of the perpetrator. If the sexual double standard hypothesis is correct, then we would expect to see that the coercive behaviors when perpetrated by a male actor, compared to the female actor engaging in the same behavior, would be rated more highly in physical and sexual aggressiveness. However, we did not expect that men would be rated as more emotionally aggressive than women when they use the same behaviors, as women are frequently stereotyped as being emotionally aggressive. Similarly, we would expect the man who uses the coercive behavior would be perceived as acting more aggressively rather than romantically or promiscuously. In contrast, we would expect that the woman who uses coercive strategy would be rated as acting more romantically or promiscuously rather than aggressively. If the second alternative is correct and women are evaluated especially negatively for violating social norms, then we might expect to see that women who engage in sexually coercive behaviors are not necessarily rated as different from men who engage in the same behaviors. Thus, the first goal of this research was to test these two arguments to understand how participants perceive men's and women's sexually coercive behaviors that occur in heterosexual college dating relationships.

In addition, participants' ratings of the actors' satisfaction with the relationship might provide insight into how sexually coercive behaviors in college relationships are interpreted. Specifically, we wondered if victims of a partner's sexually coercive strategies would be viewed as being less satisfied with the relationship, as compared to relationships when no coercion occurred. We expected that female victims of sexual coercion would be perceived as being less satisfied with the relationship than women in the control condition. However, based on Smith et al.'s (1988) findings, where men are thought to have positive reactions to women's sexual aggression, we did not expect men's relationship satisfaction to be perceived as decreasing when they experienced sexual coercion. Similarly, we wondered if being the perpetrator of a sexually coercive tactic, as compared to the control condition, was associated with changes in perceived relationship satisfaction. To date no research has examined this question. Thus, the second goal of this research was to examine perceptions of relationship satisfaction for both actors when sexual coercion occurs on a date.

METHOD

Participants

Students (n = 426; 276 women and 147 men) from a medium-size Midwestern private university and a small Eastern public college completed the study for extra credit. The sample was predominately White (85%, n = 363). One hundred and sixty-eight participants (39.4%) reported being in an exclusive dating relationship, 87 (20.4%) participants were dating casually, 140 (32.9%) participants were single, and 24 (5.6%) participants were married or engaged. The mean age of the participants was 19.63 (SD = 2.74) years.

Vignettes

Eight vignettes of a college student date were developed to manipulate the strategy used to obtain sex (control/consensual, verbal pressure, alcohol intoxication, or physical force) and aggressor gender. In all cases, the relationship was heterosexual. All scenarios provided general information that John and Carla were college students and had been dating for almost 2 months. They had dinner and then returned to one person's apartment to watch a movie. The apartment belonged to the person who was the sexual aggressor in the scenario. In all scenarios except the control/consensual condition, it was indicated that the couple started kissing during the movie, but when the aggressor started to touch the victim "suggestively," the victim clearly stated that he or she did not want to have sexual intercourse. The scenarios then differed by the coercive strategy that the aggressor used to obtain intercourse. In the verbal pressure condition, the aggressor threatened to end the relationship if they didn't have sex. In the intoxication condition, the aggressor gave the partner alcoholic drinks until he or she became drunk. In the physical force condition, the aggressor physically held down the unwilling partner. In all scenarios it was stated that sexual intercourse did eventually occur as a result of the aggressor's actions. It was never indicated that the victim consented to intercourse. It was noted in all scenarios that this was the first time the couple had sex. In the control conditions, the target was described has having responded positively to the partner's initiation of sexual intercourse and never said "no" to sexual intercourse.

Procedure

Participants were randomly assigned to read one scenario and then completed ratings of the man and woman (counterbalanced), ratings of the behavior used to obtain sex, and ratings of the man's and woman's satisfaction with the relationship (counterbalanced). The last survey included manipulation checks and participant demographics.

Measures

Perceptions of the actors. Participants rated both the man and the woman on identical sets of traits as measured on a 7-point Likert scale from 1 (not at all) to 7 (very much). Four subscales for each actor were computed by taking the mean of the subscale items. The first subscale, "romantically interested," consisted of nine traits that were appropriate for a romantically interested partner (e.g., romantic, friendly). The coefficient alphas for the subscales rating the man and woman were .90 and .88, respectively. The second subscale was labeled "victimized" and included seven traits (e.g., scared, upset). The coefficient alphas for perceptions of the man and woman were .85 and .90, respectively. The third subscale was labeled "promiscuous" and consisted of five items (e.g., easy, promiscuous). The coefficient alphas for perceptions of the man and woman were .81 and .88, respectively. The fourth subscale was labeled "aggressive" and consisted of eight items (e.g., threatening, aggressive). The coefficients for rating the man and woman were .94 and .91, respectively.

Perception of strategy used to obtain sex. Participants rated the aggressor's behavior on a variety of traits using a 7-point Likert scale from 1 (not at all) to 7 (very much). Three subscales were computed by taking the mean of the designated items. The first subscale measured perceptions that the behavior was emotionally aggressive (e.g., emotionally coercive, emotionally aggressive, [alpha] = .82). The second subscale measured perceptions that the behavior was sexually aggressive (e.g., sexually aggressive, sexually coercive, [alpha] = .72). The third subscale measured perceptions that the behavior was physically aggressive (e.g., physically manipulative, violent [alpha] = .84).

Perceptions of relationship satisfaction. There were six relationship satisfaction questions that were asked twice, to obtain perceptions of both John's and Carla's satisfaction (e.g., "How satisfied is Carla/John with the relationship?"). Order of presentation of items was counterbalanced. Items were on a 7-point Likert scale from 1 (low) to 7 (high). The alpha coefficients for the perceptions of the man and woman's satisfaction were .91 and .90, respectively.

Demographics and manipulation checks. Participants completed a survey of basic demographics and a set of manipulation checks, which asked participants to indicate if sex occurred on the date, who initiated sex, and if John and Carla consented to sex.

RESULTS

Preliminary Analyses

Based on the manipulation checks, 17 individuals misreported the person who initiated sex or incorrectly answered that sex did not occur. These participants were not included in the analyses.

Perceptions of John and Carla

To analyze the perceptions of John and Carla in the scenario, we conducted a series of 2 (aggressor gender) X 4 (coercive strategy) X 2 (target: rating of John and Carla) ANOVAs, with the target rating as a within-subjects variable. To break down the interactions, we computed simple effect tests and Tukey HSD post-hoc tests as necessary for the comparisons of interest. Descriptive statistics are summarized in Table 1.

Ratings of John and Carla as victimized. For victimization ratings, the Aggressor Gender X Strategy X Target interaction was significant, F(3, 395) = 64.41, p < .001, partial [[eta].sup.2] = .33. For the perception of John as victimized, overall he was perceived as more victimized when the woman was the aggressor (M = 3.65) than when he was the aggressor (M = 2.32). Simple effect tests indicated that when Carla was the aggressor, perceptions of John as victimized differed by coercive strategy (F[3,199] = 68.11, p < .01). Post-hoc tests indicated that he was rated as most victimized in the verbal (M = 4.42) and physical force (M = 4.60) conditions, and victimization scores were lower in the control (M = 2.17) and alcohol (M = 3.62) conditions. All mean differences were significant except between verbal and physical force (Tukey HSD, p < .05).

Simple effect tests examining perceptions of Carla as victimized indicated that overall, she was perceived as more victimized when John was the aggressor (M = 4.13) than when she was the aggressor (M = 2.31). When John was the aggressor, perceptions of the Carla as victim differed significantly by coercive strategy type (F[3,197] = 103.23, p < .001), such that she was rated as least victimized in the control condition (M = 2.17), followed by the alcohol (M = 4.28), verbal (M = 5.13), and physical (M = 5.18) strategies, and all means were significantly different (Tukey HSD, p < .05) except between verbal and physical conditions.

Research has suggested that men's use of coercive behaviors toward women was perceived as more harmful than women's use of the same behaviors toward men. To test this, we compared perceptions of victimization of the man when he was a target of the woman's aggression to the woman when she was the target of the man's aggression. There was no difference between ratings of victimization for John (M = 2.17) and Carla (M = 2.17) when sex was consensual (F[1,108] = .002, ns). However, Carla was perceived to be more victimized by John, than was John by Carla, when verbal pressure ([M.sub.woman-victim] = 5.13 vs. [M.sub.man-victim] = 4.42; F[1, 97] = 14.84, p < .01, partial [[eta].sup.2] = .13), alcohol ([M.sub.woman-victim] = 4.28 vs. [M.sub.man-victim] = 3.62; F[1,98] = 10.74, p < .01, partial [[eta].sup.2] = .10), and physical force ([M.sub.woman-victim] = 5.18 vs. [M.sub.man-victim] = 4.60; F [1, 93] = 6.21, p < .02, partial [[eta].sup.2] = .06) were used.

Ratings of aggressiveness. For perceptions of actor aggressiveness, we found a significant three-way Aggressor Gender X Strategy X Target interaction, F(3, 395) = 136.59, p < .001, partial [[eta].sup.2] = .51. Simple effect tests indicated that when John was the aggressor, perceptions of his aggressiveness were lowest in the control condition (M = 2.74) and significantly higher in the verbal (M = 5.93), physical (M = 5.52), and alcohol (M = 5.24) conditions, F(3, 196) = 135.72, p < .001. Similarly, when Carla was the aggressor, perceptions of her aggressiveness were lowest in the control condition (M = 2.73), and she was rated as significantly more aggressive when using verbal (M = 5.49), physical (M = 4.98), and alcohol strategies (M = 5.05), F(3, 199) = 122.75, p < .001. Carla was rated as more aggressive in the verbal condition than in the physical and alcohol conditions (p < .05).

Research suggests that men are rated as more aggressive than are women for the same behavior. We compared perceptions of John and Carla for aggressiveness when they were the aggressors. John was rated as more aggressive than Carla for the physical ([M.sub.woman-aggresor] = 4.98 vs. [M.sub.man-aggressor] = 5.52; F[1, 93] = 8.91, p < .0l, partial [[eta].sup.2] = .09), and verbal conditions ([M.sub.woman-aggressor] = 5.49 vs. [M.sub.man-aggressor] = 5.93; F[1, 97] = 9.07,"p < .01, partial [[eta].sup.2] = .09). However, there was no difference for aggression ratings in the control ([M.sub.woman-aggressor] = 2.73 vs. [M.sub.man-aggressor] = 2.74; F([1,108] = .002, ns) and alcohol [M.sub.woman-aggressor] = 5.05 vs. [M.sub.man-aggressor] = 5.24; F[1, 97] = 1.44, ns) conditions.

Ratings of romantic interest. For ratings of romantic interest, the Aggressor Gender X Strategy X Target interaction was significant, F(3, 395) = 79.62, p < .001, partial [[eta].sup.2] = .38. Overall, John was perceived to be more romantically interested when Carla was the aggressor (M = 4.20) then when he was the aggressor (M = 2.44). Simple effect tests indicated that when John was the aggressor, the perceptions of his romantic interest were significantly higher in the control condition (M = 3.91) than in any of the coercive conditions ([M.sub.verbal] = 1.69; [M.sub.physical] = 1.95; and [M.sub.alcoho1] = 2.02; F[3, 196] = 84.38, p < .001). Perceptions of John's romantic interest did not significantly differ across the coercive conditions when the woman was the aggressor.

Overall, Carla was perceived to be more romantically interested when John was the aggressor (M = 4.23) than when she was the aggressor (M = 3.08). Simple effect tests indicated that when Carla was the aggressor, perceptions of her romantic interest were significantly higher in the control condition (M = 4.13) than in any of the coercive conditions ([M.sub.verbal] = 2.50; [M.sub.physical] = 2.88; and [M.sub.alcohol] = 2.66; F[3, 199] = 36.68, p < .001). However, perceptions of Carla's romantic interest were unaffected by the strategy when she was the victim.

Research suggests that when women aggress, their behaviors are more likely to be attributed to romantic interest than are men's behaviors. When Carla was the aggressor, she was seen as more romantically interested than John when he was the aggressor for the verbal (M.sub.woman-aggressor] = 2 50 vs. [M.sub.man-aggressor] = 1.69; F[1, 97] = 26.29, p < .01, partial [[eta].sup.2]= .21),-physical ([M.sub.woman-aggressor] = 2.88 vs. [M.sub.man-aggressor] = 1.95; F[1, 93] = 28.13, p < .13-1, partial [[eta].sup.2] = .23) and alcohol conditions ([M.sub.woman-aggressor] = 2.66 vs. [M.sub.man-aggressor] = 2.02; F[1, 97] = 15.45, p < .01, partial [[eta].sup.2] = .14]. However, there was no difference between John and Carla when they initiated control/mutual consensual sex ([M.sub.woman-aggressor] = 4.13 vs. [M.sub.man-aggressor] = 3.91; F[1, 108] = 1.41, ns). This suggests that women who utilize sexually coercive behaviors are more likely than men to have their behaviors interpreted as "romantic."

Ratings of promiscuity. The Aggressor Gender X Strategy X Target interaction for promiscuity was significant, F(3, 395) = 62.64, p < .001, partial [[eta].sup.2] = .32. Simple effect tests indicated that perceptions of John as promiscuous were not influenced by his specific strategy when he was the aggressor, F(3,196) = 2.48, ns. However, perceptions of Carla as promiscuous were influenced by the coercive strategy when she was the aggressor (F[3,203] = 7.44, p < .001). Each of the coercive conditions resulted in higher promiscuity ratings compared to the control condition ([M.sub.alcohol] = 6.02, [M.sub.physical] = 5 67, [M.sub.verbal] = 5.45, [M.sub.control] = 5.30 Furthermore, when Carla used alcohol as a strategy, she was viewed as more promiscuous than in the verbal condition. Both John and Carla were perceived as less promiscuous when the partner used a coercive strategy as compared to when they consented to sex in the control condition.

We compared perceptions of John and Carla's promiscuity when they were the aggressors. Carla, compared to John, was perceived as more promiscuous for the verbal ([M.sub.woman-aggressor] = 5.45 vs. [M.sub.man-aggressor] = 4.88; F[1, 97] = 8.84, p < .01, partial [[eta].sup.2] = .08), physical ([M.sub.woman-aggressor] = 5.67 vs. [M.sub.man-aggressor] = 5.24; F[1, 93] = 5.31, p < .05, partial [[eta].sup.2] = .05), and alcohol ([M.sub.woman-aggressor] = 6.02 vs. [M.sub.man-aggressor] = 5.41; F[1, 97] = 12.43, p < .01, partial [[eta].sup.2] = .11) conditions. However, there were no differences in rating the aggressors as promiscuous in the control ([M.sub.woman-aggressor] = 5.30 vs. [M.sub.man-aggressor] = 5.20; F[1, 108] = .31, ns) condition.

Comparisons across ratings. Within-subjects ANOVAs were also conducted to examine perceptions of the target across ratings of aggressiveness, victimization, promiscuity, and romance. When Carla was the aggressor, she was perceived most highly as promiscuous (M = 5.61), followed by aggressive (M = 4.50), romantically interested (M = 3.08), and victimized (M = 2.31), F(3,597) = 584.44, p < .001, partial [[eta].sup.2] = .75; all means were significantly different, p < .01. When Carla was the target of aggression she was perceived to be victimized (M = 4.13), romantically interested (M = 4.23), promiscuous (M = 3.57), and aggressive (M = 2.10; F[3, 591] = 288.60, p < .001, partial [[eta].sup.2] = .59). All means were significantly different except between romantically interested and victimized (HSD, p < .01).

When John was the aggressor, he was perceived as being promiscuous (M = 5.17), aggressive (M = 4.80), and equally low in victimization (M = 2.32) and romantic interest (M = 2.44), F(3, 588) = 622.04, p < .01, partial [[eta].sup.2] = .76. Post-hoc tests indicated all means were significantly different except victimization and romantic ratings (p < .01). However, when John was the victim, he was perceived primarily as romantically interested (M = 4.20), followed by victimized (M = 3.65), promiscuous (M = 3.60), and aggressive (M = 2.24); F(3, 597) = 163.59, p < .01, partial [[eta].sup.2] = .45. All means were different from each other except for ratings of victimization and promiscuity, p < .01.

Rating of the Strategy Used to Obtain Sex

To examine the impact of aggressor gender and coercive strategy on perceptions of the behavior, we computed a 2 (Aggressor Gender) x 4 (Strategy) MANOVA with the three behavioral ratings as dependent variables. The interaction (Wilks) [lambda] = .94, F[9, 954] = 2.99, p < .01, partial [[eta].sup.2] = .02), as well as the main effects for strategy (Wilks [lambda] = .39, F[9, 954.18] = 50.11, p < .001, partial [[eta].sup.2] = .26) and aggressor gender (Wilks) [lambda] = .91, F[3, 392] = 12.84, p < .001, partial [[eta].sup.2] = .09) were all significant. Univariate ANOVAs with Tukey post-hoc tests were conducted as follow-ups to the MANOVA. For ratings of emotional aggressiveness, there was a main effect for strategy, F(3, 394) = 63.32, p < .001, partial [[eta].sup.2] = .33. Physical force (M = 5.25), alcohol (M = 5.28), and verbal (M = 5.81) strategies were rated as more emotionally coercive than the control condition (M = 3.44). Verbal aggression was rated as more emotionally aggressive than physical and alcohol strategies. Main effects for aggressor gender and the interaction were not significant.

For ratings of sexual aggressiveness, there was a main effect for strategy, F(3,394) = 60.29, p < .001, partial [[eta].sup.2] = .32. Post-hoc tests indicated that verbal (M = 6.27), physical (M = 6.31), and alcohol (M = 6.19) coercion were perceived as more aggressive than the control/mutual (M = 4.63) behavior. There was no difference in aggression between the coercive strategies. The main effect for gender and the interaction was not significant.

For the ratings of physical aggressiveness, the interaction was significant, F(3, 394) = 6.40, p < .001, partial [[eta].sup.2] = .05. Simple effect tests indicated that John and Carla were perceived as equally physically aggressive in the control ([M.sub.man] = 2.11 vs. [M.sub.woman] = 2.32; F[1, 113] = 2.32, n.s.) condition. However, the behavior was rated as more physically aggressive when utilized by John in the verbal ([M.sub.man] = 5.59 vs. [M.sub.woman] = 4.30; F[1, 100] = 23.63, p < .001), alcohol ([M.sub.man] = 5.46 vs. [M.sub.woman] = 4.77; F[1,104] = 5.88, p < .05), and physical ([M.sub.man] = 6.13 vs. [M.sub.woman] = 5.46; F[1, 99] = 7.82, p < .01) conditions.

Rating of Perceived Relationship Quality

To test for the impact of aggressor gender and strategy on the man's and woman's perceived relationship satisfaction, we computed a 2 (Aggressor Gender) X 4 (Strategy) X 2 (Actor Perspective: ratings of John and Carla) ANOVA with the last variable as a within factor (see Table 2). The three-way interaction was significant (F[3, 396] = 10.05, p < .001, partial [[eta].sup.2] =. 10). Simple effect tests indicated that John's relationship satisfaction did not vary by coercive strategy when the female was the aggressor. However, when he was the aggressor, John was perceived as most satisfied in the control condition (M = 4.50) and significantly less satisfied when he used any of the coercive strategies ([M.sub.verbal] = 2.73; [M.sub.physical] = 2.98; [M.sub.alcoho1] = 3.36; F[3,196] = 28.80, p < .001).

Simple effect tests indicated that when Carla was the aggressor, her relationship satisfaction varied by strategy such that she was perceived as most satisfied in the control condition (M = 5.07) and significantly less satisfied when she used any of the coercive strategies ([M.sub.verbal] = 3.67; [M.sub.physical] = 3.98; [M.sub.alcohol] = 3.74; F[3,200] = 14.44, p < .001). When she was the victim, Carla s satisfaction differed by coercive strategy (F[3,196] = 28.80, p < .001) such that she was perceived as more satisfied in the control condition (M = 5.05) than in the physical (M = 4.17) or alcohol (M = 4.00) conditions. The control and verbal (M = 4.60) conditions did not differ.

DISCUSSION

Given the prevalence of aggression in college dating relationships, we sought to examine how college students react to men and women who use sexually coercive behaviors to obtain sex from an unwilling partner, how the behaviors are interpreted, and whether or not the people are thought to be satisfied with the relationship. The pattern of results across these variables provides insight into how college students view sexually coercive behaviors in dating relationships.

These findings tell us that coercive strategies and the person who uses them are not viewed as particularly aggressive. Likewise, the targets of the coercive strategy are not rated very highly as victimized. Although participants did rate the actors in the coercive conditions as more aggressive and victimized than in the control condition, the means were not high. On a seven-point scale (with high numbers indicating more aggressiveness and victimization), the highest mean for aggressiveness was 5.93, and the highest perception of victimization was 5.18. Thus, participants did not perceive the aggressor as far above the scale midpoint for aggressiveness. Similarly, participants did not perceive the target as highly victimized. These findings are especially shocking for the scenario where the man uses physical force to obtain sex from the unwilling woman. This scenario follows the prototypical script of date rape, and one would expect that it would be rated as unambiguously aggressive.

This suggests that in college dating relationships these behaviors might be considered to be socially normal. This is troublesome because behaviors seen as normal or "expected" in college dating relationships are likely to continue. Recent findings by other researchers have also noted the importance of social normative beliefs for dating violence. For example, an adolescent's own experiences with dating violence are predicted by having friends who are involved in interpersonally aggressive dating relationships (Arriaga & Foshee, 2004). Similarly, adolescents who were exposed to marital violence as children have been found to develop attitudes that condone interpersonal violence (Lichter & McCloskey, 2004). Thus, this acceptance and social norming of interpersonally-coercive behaviors may be a contributing factor to aggression in dating relationships.

We also found that participants' characterization of the man and woman when using sexually coercive strategies reflected gender stereotypes. The man's use of coercion was perceived as more aggressive than the woman's use of the same behavior. The man was rated as aggressive when he used verbal, alcohol, or physical strategies; however, the woman's coercive behaviors were more likely to be viewed as promiscuous. To the extent that a woman was rated as aggressive, we found that this occurred when she used a verbal strategy, not when she used physical force. This is consistent with the stereotype that women, compared to men, cannot inflict as much physical harm, and thus women's verbal coercion might appear to be more aggressive. Note that the verbal strategy was also the behavior that was rated as most emotionally aggressive. To the extent women were perceived as aggressive, it was emotional, not physical.

There is a double standard in how sexually coercive behaviors are viewed when perpetrated by a man and woman. He is aggressive; she is promiscuous. Likewise, there is a double standard in how the victims are viewed. In the conditions where the woman was the victim, she was perceived as primarily victimized; however, when the man was the victim, he was primarily viewed as romantically interested. These findings support previous research (e.g., Hannon et al., 2000; Harris, 1991; Harris & Knight-Bohnhoff, 1996) which argued that there is trivialization of women's coercion against men. However, the double standard that women shouldn't initiate consensual sex (Crawford & Popp, 2003) was not found in this study, as men and women who initiated sex in the consensual conditions were viewed in essentially the same manner. This suggests that in college dating relationships, the stereotypes are shifting so that it is more acceptable for women to initiate consensual sex than it was in the past.

Another striking finding is that participants did not necessarily perceive the victim of sexual coercion as experiencing a decrease in relationship satisfaction. Rather, having to use a coercive behavior to obtain sex was viewed as an indicator of the aggressor having lower satisfaction with the relationship. John being the victim of sexual coercion did not result in participants' perceptions of his satisfaction decreasing. However, Carla was perceived as having lower satisfaction when she was the victim of the physical and alcohol strategies. This finding was consistent with Smith and colleagues' (1988) research, where participants expected men to have positive reactions to women's sexually coercive behaviors. These findings are also reflective of the reported experience of victims. Research has shown that women, but not men, who are the victims of dating coercion experience a decrease in relationship satisfaction (Katz et al., 2002). Nonetheless, this finding is troubling because one would hope that people perceive sexual coercion as a reason for the victim, regardless of his or her gender, to be dissatisfied with a relationship. If male victims do not believe it is appropriate to feel upset or distressed by the event, they may continue to stay in an unhealthy relationship.

Taken together, these results indicate that college students are not viewing sexually coercive behaviors as particularly aggressive or troublesome in heterosexual college dating relationships. Given the high prevalence rates of sexual coercion found in previous research and the results of the current study, we suspect that a certain level of sexual coercion is considered normal in college student dating relationships. If college students do not perceive the aggressors as being highly "aggressive" and the victims as being "victimized," one might argue that researchers shouldn't label the behaviors as such either. However, we argue that just because college students don't perceive a behavior as aggressive doesn't mean that it isn't problematic. Just because people don't label themselves as a victim doesn't mean that they don't experience negative implications. Research has clearly shown that women who have been victims of sexual aggression experience negative reactions (Browne, 1993). Similarly, men who have been the targets of women's sexually coercive advances also experience a variety of negative reactions (Byers & O'Sullivan, 1998; Struckman-Johnson & Struckman-Johnson, 1998). Not acknowledging a person as a victim could interfere with the person seeking help or treatment (Macchietto, 1998; Muehlenhard, 1998) or result in the person staying in an unhealthy relationship. Not identifying a person as coercive when he or she is using verbal threats, purposeful intoxication, or physical force to obtain sexual intercourse could prevent appropriate interventions and result in the person continuing to engage in these behaviors. Indeed, that college students do not label the aggressors and these behaviors as highly coercive, or see the targets as victims, indicates the seriousness of the problem on college campuses and highlights the need for proper education about the various forms of sexual coercion perpetrated by both men and women.

This study provides insights into how college students interpret sexually coercive behaviors in dating relationships, but there are limitations. One limitation is that we only examined heterosexual dating relationships. This study should be replicated and extended to same-sex relationships. A second limitation is that we used study vignettes. While this allowed us to experimentally manipulate the variables of interest, participants had little relatively information to use in making their ratings of the targets and behavior. Future research could ask people about real situations that they have encountered and to examine their perceptions of these incidents.

In this study, we chose to examine perceptions of coercion initiated by both men and women. By including women as perpetrators of sexual coercion, we did not mean to detract from the serious problem of men's aggression towards women. Coercion from any person in a relationship, regardless of his or her gender, is problematic for the relationship and the people involved. Furthermore, relationships that are mutually coercive have more frequent and severe behaviors than those that display one-sided aggression (e.g., Gray & Foshee, 1997). We seek to further our understanding of the whole picture of sexual coercion and how college students interpret and perceive these behaviors from both individuals. Ultimately, to reduce interpersonal aggression in dating relationships, we must understand how people interpret these behaviors from all individuals in the relationship. We hope this research helps us understand this complex problem.

REFERENCES

Abbey, A., BeShears, R., Clinton-Sherrod, A., & McAuslan, P. (2004). Similarities and differences in women's sexual assault experienced based on tactics used by the perpetrator. Psychology of Women Quarterly, 28, 323-332.

Anderson, P. B. (1998). Women's motives for sexual initiation and aggression. In P. Anderson & C. Struckman-Johnson (Eds.), Sexually aggressive women: Current perspectives and controversies (pp. 79-93). New York: Guilford.

Anderson, P. B., & Struckman-Johnson, C. (Eds.). (1998). Sexually aggressive women: Current perspectives and controversies. New York: Guilford.

Archer, J. (2000). Sex differences in aggression between heterosexual partners: A meta-analytic review. Psychological Bulletin, 126, 651-680.

Arriaga, X. B., & Foshee, V. A. (2004). Adolescent dating violence: Do adolescents follow in their friends', or their parents', footsteps? Journal of Interpersonal Violence, 19, 162-184.

Black, K. A., & Gold, D. J. (2003). Men's and women's reactions to hypothetical sexual advances: The role of initiator socioeconomic status and level of coercion. Sex Roles, 49, 173-178.

Browne, A. (1993). Violence against women by male partners: Prevalence, outcomes, and policy implications. American Psychologist, 48, 1,077-1,087.

Byers, E. S., & O'Sullivan, L. F. (1998). Similar but different: Men's and women's experiences of sexual coercion. In P. Anderson & C. Struckman-Johnson (Eds.), Sexually aggressive women: Current perspectives and controversies (pp. 144-167). New York: Guilford.

Clements-Schreiber, M. E., Rempel, J. K., & Desmarais, S. (1998). Women's sexual pressure tactics and adherence to related attitudes: A step toward prediction. The Journal of Sex Research, 35, 197-205.

Crawford, M., & Popp, D. (2003). Sexual double standards: A review and methodological critique of two decades of research. The Journal of Sex Research, 40, 13-26.

Frieze, I. H. (2000). Violence in close relationships--development of research area: Comment on Archer (2000). Psychological Bulletin, 126, 681-384.

Gray, H. M., & Foshee, V. (1997). Adolescent dating violence: Differences between one-sided and mutually violent profiles. Journal of Interpersonal Violence, 12. 126-141.

Hammock, G. S., & Richardson, D. R. (1993). Blaming drunk victims: Is it just world or sex role violation? Journal of Applied Social Psychology, 23, 1,574-1,586.

Hannon, R., Hall, D., Nash, H., Formati, J., & Hopson, T. (2000). Judgments regarding sexual aggression as a function of sex of aggressor and victim. Sex Roles, 43, 311-322.

Harris, M. B. (1991). Effects of sex of aggressor, sex of target, and relationship on evaluations of physical aggression. Journal of Interpersonal Violence, 6, 174-186.

Harris, M. B., & Knight-Bohnhoff, K. (1996). Gender and aggression I: Perceptions of aggression. Sex Roles, 35, 1-24.

Jenkins, S. S., & Aube, J. (2002). Gender differences and gender-related constructs in dating aggression. Personality and Social Psychology Bulletin, 28(8), 1,106-1,118.

Kahn, A. S., Jackson, J., Kully, C., Badger, K., & Halvorsen, J. (2003). Calling it rape: Differences in experiences of women who do or do not label their sexual assault as rape. Psychology of Women Quarterly, 27, 233-242.

Katz, J., Kuffel, S. W., & Coblentz, A. (2002). Are there gender differences in sustaining dating violence? An examination of frequency, severity, and relationship satisfaction. Journal of Family Violence, 17, 247-271.

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

Krahe, B. (2000). Sexual scripts and heterosexual aggression. In T. Eckes and M. Harms (eds.). The developmental social psychology of gender. New Jersey: Lawrence Earlbaum.

Krahe, B., Waizenhofer, E., & Moller, I. (2003). Women's sexual aggression against men: Prevalence and predictors. Sex Roles, 49, 219-232.

Langhinrichsen-Rohling, J., Shlien-Dellinger, R., Huss, M., & Kramer, V. (2004). Attributions about perpetrators and victims of interpersonal abuse: An Analogue Study. Journal of Interpersonal Violence, 19, 484-498.

Lichter, E. L., & McCloskey, L. A. (2004). The effects of childhood exposure to marital violence on adolescent gender-role beliefs and dating violence. Psychology of Women Quarterly, 28, 344-357.

Lottes, I., & Weinberg, M. (1996). Sexual coercion among university students: A comparison of the United States and Sweden. The Journal of Sex Research, 34, 67-76.

Macchietto, J. G. (1998). Treatment issues of adult male victims of female sexual aggression. In E Anderson & C. Struckman-Johnson (Eds.), Sexually aggressive women: Current perspectives and controversies (pp. 187-204). New York: Guilford.

Muehlenhard, C. L. (1998). The importance and danger of studying sexually aggressive women. In E Anderson & C. Struckman-Johnson (Eds.), Sexually aggressive women: Current perspectives and controversies (pp. 19-48). New York: Guilford.

O'Sullivan, L. F., Byers, E. S., & Finkelman, L. (1998). A comparison of male and female college students' experiences of sexual coercion. Psychology of Women Quarterly, 22, 177-195.

Richardson, D., & Campbell, J. L. (1982). Alcohol and rape: The effect of alcohol on attributions of blame for rape. Personality and Social Psychology Bulletin, 8, 468-476.

Russell, B. L., & Oswald, D. L. (2001). Strategies and dispositional correlates of sexual coercion perpetrated by women: An exploratory investigation. Sex Roles, 45, 103-115.

Russell, B. L., & Oswald, D. L. (2002). Sexual coercion and victimization of college men: The role of love styles. Journal of Interpersonal Violence, 17, 273-285.

Shea, M. E. (1998). When the tables are turned: Verbal sexual coercion among college women. In P. Anderson & C. Struckman-Johnson (Eds.), Sexually aggressive women: Current perspectives and controversies (pp. 94-104). New York: Guilford.

Shook, N. J., Gerrity, D. A., Jurich, J., & Segrist, A. E. (2000). Courtship violence among college students: A comparison of verbally and physically abusive couples. Journal of Family Violence, 15, 1-22.

Smith, R., Pine, C., & Hawley, M. (1988). Social cognition about adult male victims of female sexual assault. The Journal of Sex Research, 24, 101-241.

Stormo, K. J., Lang, A. R., & Stritzke, W. G. (1997). Attributions about acquaintance rape: The role of alcohol and individual differences. Journal of Applied Social Psychology, 27, 279-305.

Struckman-Johnson, C. (1988). Forced sex on dates: It happens to men, too. The Journal of Sex Research, 24, 234-241.

Struckman-Johnson, C., & Anderson, P. B. (1998). "Men do and women don't": Difficulties in researching sexually aggressive women. In E Anderson & C. Struckman-Johnson (Eds.), Sexually aggressive women: Current perspectives and controversies (pp. 9-18). New York: Guilford.

Struckman-Johnson, C., & Struckman-Johnson, D. (1993). College men's and women's reactions to hypothetical sexual touch varied by initiator gender and coercion level. Sex Roles, 29, 371-385.

Struckman-Johnson, C., & Struckman-Johnson, D. (1998). The dynamics and impact of sexual coercion of men by women. In P. Anderson & C. Struckman-Johnson (Eds.), Sexually aggressive women: Current perspectives and controversies (pp. 121-143). New York: Guilford.

Struckman-Johnson, D., & Struckman-Johnson, C. (1991). Men and women's acceptance of coercive sexual strategies varied by initiator gender and couple intimacy. Sex Roles, 25, 661-676.

Manuscript accepted September 22, 2005

Note. We thank Kristine Chapleau, Kara Lindstedt, Angela Pirlott, and Sara Thimsen for their assistance with data collection.

Address correspondence to Debra L. Oswald, Marquette University, Department of Psychology, P.O. Box 1881, Milwaukee, WI, 53201; e-mail: [e-mail].

Debra L. Oswald

Marquette University

Brenda L. Russell

Castleton State College 

Table 1. Rating of John and Carla for the 2 (Gender of aggressor) x 4
(strategy) x 2 (target) ANOVAS

                             John is aggressor

                  Control      Verbal       Alcohol     Physical

Aggressive
  John          2.74 (1.07)  5.93 (.79)   5.24 (.85)   5.52 (.87)
  Carla         1.91 (.71)   1.92 (.57)   2.26 (.70)   2.41 (.75)
Victimized
  John          1.62 (.60)   2.84 (.94)   2.40 (.95)   2.50 (1.02)
  Carla         2.17 (.95)   5.13 (.93)   4.28 (1.10)  5.18 (1.06)
Romantic
  John          3.91 (.99)   1.69 (.67)   2.02 (.78)   1.95 (.77)
  Carla         4.26 (.80)   4.20 (.99)   3.99 (.80)   4.51 (.89)
Promiscuous
  John          5.20 (1.04)  4.88 (1.04)  5.41 (.94)   5.24 (1.00)
  Carla         4.86 (.86)   3.06 (.83)   3.15 (.98)   2.99 (1.18)

                             Carla is aggressor

                  Control      Verbal       Alcohol     Physical

Aggressive
  John          2.11 (.95)   2.26 (.61)   2.11 (.92)   2.49 (.79)
  Carla         2.73 (.96)   5.49 (.62)   5.05 (.73)   4.98 (.88)
Victimized
  John          2.17 (.78)   4.42 (.89)   3.62 (.96)   4.60 (1.21)
  Carla         2.07 (1.02)  2.61 (.80)   2.40 (1.05)  2.23 (.89)
Romantic
  John          4.11 (1.02)  4.15 (1.00)  4.37 (.99)   4.18 (.81)
  Carla         4.13 (.90)   2.50 (.90)   2.66 (.84)   2.88 (.91)
Promiscuous
  John          4.71 (.97)   3.15 (.93)   3.08 (1.10)  3.32 (1.07)
  Carla         5.30 (.88)   5.45 (.83)   6.02 (.79)   5.67 (.82)

Note. Means are on a scale of 1 to 7 (7 = high) and standard
deviations are reported in parentheses.

Table 2. Rating of Relationship Satisfaction for the 2
(gender of aggressor) x 4 (strategy) x 2 (target) ANOVAS

                                  John is aggressor

                       Control      Verbal       Alcohol     Physical

John's satisfaction  4.50 (1.14)  2.73 (1.02)  3.36 (1.11)  2.98 (.98)
Carla's
  satisfaction       5.05 (.98)   4.60 (1.13)  4.00 (1.46)  4.17 (1.28)

                                  Carla is aggressor

                       Control      Verbal       Alcohol     Physical

John's satisfaction  4.53 (1.20)  4.32 (1.20)  4.30 (1.14)  4.70 (.98)
Carla's
  satisfaction       5.07 (1.21)  3.67 (1.18)  3.74 (1.18)  3.98 (1.43)

Note. Means are on a scale of 1 to 7 (7 = satisfied) and standard
deviations are reported in parentheses.