Acceptability of dating violence among late adolescents: the role of sports participation, competitive attitudes, and selected dynamics of relationship violence.

Merten, Michael J.

Beginning with Makepeace's (1981) pioneering work on dating violence, research has continued to present a sobering picture of the extent to which violence occurs in dating and courtship relationships (Perry & Fromuth, 2005). This violence is not strictly confined to one gender, as both men and women are victimized with women using as much or more violence against their male partners (Marcus & Swett, 2002). For instance, O'Keefe and Treister (1998) report that 45.5% of males and 43.2% of females have experienced at least one incident of physical aggression from dating partners during the course of their dating. The high prevalence rates for dating violence, first noted over 20 years ago by Roscoe (1985), indicate the degree to which violence in dating continues to be acceptable (Pleck, 2004).

A number of factors that may contribute to the acceptance of dating violence have been identified in previous research. These include parental violence experienced as a child (Foshee, Ennett, Bauman, Benefield, & Suchindran, 2005; Lichter & McCloskey, 2004), the seriousness, importance, and length of the dating relationship (Neufeld, McNamara, & Ertl, 1999; O'Keefe, 1997), being humiliated by a partner (Taylor & Sorenson, 2005), and justness of retaliation to violence initiated by a partner (Frieze, 2005). However, additional factors that may be critical to examine based on the extensive media attention they have generated, are those relating to sports participation.

Several studies have focused on the relationship between acceptability of dating violence and sports participation (e.g., Forbes, Adams-Curtis, Pakalka, & White, 2006; Bloom & Smith, 1996; Crosset, Ptacek, McDonald, & Benedict, 1996; Mintah, Huddleston, & Doody, 1999). According to Forbes et al. (2006), males who were active in high school sports were more approving of dating violence and physical aggression. Unfortunately, these studies focused on athletic involvement as a global, uni-dimensional concept. In actuality, athletic involvement is a complex variable, comprising several dimensions, some of which may be more salient to the acceptability of dating violence than others. It may be that only some of the characteristics of athletic involvement are associated with the acceptability of violence.

Interestingly, Rossi, Schuerman, and Budde (1999) found characteristics of the individual (respondent-level characteristics) to be less important than characteristics of the situation (case-level characteristics) in decisions regarding abuse and neglect. That is, characteristics of violent interactions themselves were found to be more important in judgments about violence than the personal characteristics of the respodents making the judgments. This would suggest that characteristics of athletic involvement may be less important predictors of the acceptability of dating violence than characteristics reflecting the dynamics of dating violence interaction.

Because of the complexity of factors influencing individuals' perceptions of violence, Miller and Bukva (2001) argue that respondent- and case-level characteristics need to be examined concurrently. This strategy is followed in the present study. In addition to the three respondent-level characteristics of athleticism (i.e., athletic participation, competitiveness, and need to win), three characteristics of violent couple interactions that may relate to the acceptability of violence are also included in the study: initiator's violence, recipient's reaction, and the initiator-recipient gender combination.

Athletics and Acceptability of Dating Violence

Perhaps no social issue in sports has received more media attention in the past decade than male athletes' violence against women. A significant number of male athletes have made more headlines off the field than on due to their violence in intimate relationships (Crosset, 1999). Media stories about the aggressive acts of high school, collegiate, and professional athletes continue to appear in the news (e.g., "Cases involving athletes and sexual assault," 2003). Surprisingly, even though discussion about athletes' involvement in intimate relationship violence is a common topic within the media, Craig (2000) notes that relatively few empirical studies have focused on this issue.

Although media-based reports suggesting a connection between athletic participation and relationship violence are dramatic, not all athletes are violent in their intimate relationships. The assumption that relationship violence is a result of athletic participation in general may be a misrepresentation. Other characteristics of an athlete may lead to an increased level of acceptability of violence, characteristics that are not simply their level of participation in athletics. In addition to athletic participation, it is important that we examine potential factors such as competitiveness and need to win attitudes, factors that may accompany athletic participation, but have not been specifically examined by previous literature--which may account for variance in the acceptability of violence.

Athletic participation. In the 1990s, theorists began to emphasize the unique way that sports activity legitimizes certain forms of violent behavior (e.g., Nack & Munson, 1995). Studies also began examining athletic participation as a factor in the use of violence in contexts other than sports (e.g., Nixon, 1997; Lenzi, Bianco, Milazzo, Placidi, & Castrogiovanni, 1997). A major focus developed on the "spillover theory of violence," a perspective which suggests that violence used in sports "spills over" into the interpersonal relationships of the athletes. This theory received support in several studies (Forbes et al., 2006; Bloom & Smith, 1996; Crosset et al., 1996).

A critical component of athletes' participation in sport activities is their belief about the acceptability of the use of violence. Cauffman, Feldman, Arnett Jensen, and Jenson Arnett (2000) suggest that high rates of violent behavior reflect the development over time of attitudes regarding the acceptability of violence, fostered by environments that support such development. Thus, exposure to violence may lead to an increased acceptance of violent behavior, and thereby may increase the likelihood of an individual engaging in violent or aggressive activity. This relationship was examined by Gardner and Janelle (2002) who found that aggressive behavior both within and outside sports becomes more acceptable as individuals' years of experience in sports increase. As an example, it could be expected that aggressive behavior within and outside of sports is more acceptable among those who compete in collegiate athletics as opposed to those who have participated in sports at the high school level. Therefore, it is hypothesized that an increase in the level of athletic participation will lead to a greater acceptance of violence in intimate dating relationships among athletes.

However, participation in athletics by itself may not be the only influence on and beliefs about the acceptability of violence. More important than participation per se may be certain factors associated with athletic participation. Two of the most important of these are competitiveness, and the need to win.

Competitiveness and need to win. Many consider competitiveness to be the core motivating factor underlying athletic participation. However, conceptualizing competitiveness as a single dimension may be an oversimplification. Houston, McIntire, Kinnie, and Terry (2002) emphasize the importance of not treating competitiveness as uni-dimensional and urge researchers to use greater precision when assessing and defining the construct. For instance, according to Gill and Deeter (1988), competitiveness reflects a desire to enter an activity and strive for success, regardless of who wins. They also introduce a second concept which they term "win orientation": the desire to win or avoid losing in participatory activities.

Ryckman and Hamel's (1992) concept of personal development competitiveness is similar to Gill and Deeter's (1988) definition of competitiveness, which reflects a desire to enter an activity and strive for success, regardless of who wins. They define personal development competitiveness as an attitude in which the primary focus is not on the outcome or the desire to win but instead on the enjoyment or mastery of the task. For both Ryckman and Hamel (1992) and Gill and Deeter (1988) the emphasis is on the desire to do one's best and strive for success, not on the outcome of winning or defeating another person.

Morey and Gerber (1995) differentiate between two types of competitiveness: goal competitiveness and interpersonal competitiveness. They define goal competitiveness as the desire to be and do one's best and to excel, a definition consistent with Gill and Deeter's (1988) and Ryckman and Hamel's (1992) concepts. Morey and Gerber (1995) define their other type of competitiveness, interpersonal, as the desire to do better than others and to win or defeat others, similar to Gill and Deeter's (1988) definition of win orientation.

Perhaps the earliest conceptualization of the need to win is the construct originated by Homey (1937) known as hypercompetitiveness, which she defined as the need to win at any cost and to avoid losing. Hypercompetitiveness has been examined in several studies performed by Ryckman and colleagues (Ryckman et al., 1996, 1997, 2002). In one study (Ryckman et al., 2002) they examined the relationship of hypercompetitiveness to various aspects of heterosexual romantic experiences among college students. Results indicated that those who were more hypercompetitive reported higher levels of conflict with their partner, greater need to control, and greater infliction of physical pain. In another study (Ryckman et al., 1996) they also found hypercompetitiveness to be positively related to aggression. Ryckman and colleagues' (1996, 1997, 2002) construct of hypercompetitiveness is similar to Gill and Deeter's win orientation concept. The key underlying theme for both is the desire or need to win and avoid losing, even at great cost.

In the process of revising their competitiveness index, Houston, Harris, McIntire, and Francis (2002) provided evidence for a need to win dimension, albeit indirectly. A total of six items were dropped from their original 20-item scale. The revised scale was positively correlated with the competitiveness subscale of Gill and Deeter's (1988) Sport Orientation Questionnaire (SOQ). Interestingly, each of the six items they omitted includes a common element about winning at all costs and being the best, items resembling those of Gill and Deeter's (1988) win orientation sub-scale.

In summary, a number of studies suggest that competitiveness and the need to win are two distinct dimensions. For instance, some individuals who have a strong desire to be successful in sports activities (competitiveness) may attach relatively little importance to winning (need to win). They may enthusiastically involve themselves in the activity, enjoying their competitor's skillful moves and counter moves as much as their own regardless of the outcome of the contest. Others may be less competitive but have a high need to win. Their interest in the process and skills of the activity may be limited only to those actions that will allow them to win, including unethical and prohibited behaviors. They may even engage only in activities where they are assured they will not lose.

These studies highlight the importance of separating dimensions of competitiveness that reflect a need to win from dimensions that reflect striving for success or enjoyment of competition. Confounding the definition of competitiveness (striving for success) with that of need to win may lead to inaccurate generalizations regarding competitiveness and interpersonal violence. In this study we hypothesize that the need to win is a stronger influence than competitiveness on the acceptability of violence in dating relationships.

Respondent Gender

An additional factor that may influence the acceptability of violence in dating relationships is gender. Previous studies have shown males to be more accepting of violence than females, both within and outside the sports context (Forbes et al., 2006; Gardner & Janelle, 2002). The current study explores differences among males' and females' acceptance of violence in dating relationships.

Respondent-level vs. Case-level Characteristics

Although the acceptability of violence in dating relationships may be associated with athletic participation, competitiveness, and need to win, these respondent characteristics may not account for as much variance in the acceptability of violence in dating relationships as characteristics associated with dating violence interactions themselves. For instance, Rossi et al. (1999) found that case characteristics accounted for significantly more variance than individual characteristics in respondents' decisions regarding abuse and neglect. Miller and Bukva (2001) argue that respondent and case characteristics need to be examined concurrently because of the complexity of factors that enter into and influence judgments about violence. Three characteristics of violent couple interactions that may influence acceptability of violence are examined in this study: severity of initiator's violence, severity of recipient's reaction to the violence, and the initator-recipient gender combination. These case-level characteristics are embedded in written vignettes depicting violence in dating relationships (described below).

Characteristics of Violent Couple Interactions

Initiator's act of violence and recipient's reaction. Walking away from acts of violence including verbal or physical altercations is perhaps the ideal response to interpersonal conflict. However, mutual or bi-directional violence is a common occurrence among dating partners. For instance, Gray and Foshee (1997) found that 66% of dating partners were involved in mutual violence, which is typically characterized by an interactional dynamic in which one of the partners initiates aggressive behavior toward the other who reciprocates in kind. As a result, within a dating couple, a partner can be both perpetrator and victim (Lewis & Fremouw, 2001). Gray and Foshee (1997) report that couples experiencing mutual violence experience more severe aggression and more injuries than those in one-sided violent relationships.

In dating relationships, both the level of the initiators violence as well as the level of the recipient's reaction may be heavily dependent on how acceptable such violent actions and reactions are. Further, the acceptability of the recipient's reaction may be an even more important factor than the acceptability of the initiator's act. This is due to the fact that the level of retaliation is more likely to affect whether the conflict ceases or escalates (Ferguson & Rule, 1988). An accepting attitude by a recipient may actually increase and prolong the violence in the dating relationship.

Level of violence. The level or severity of the violent act has been found to explain a significant proportion of variation in individuals' judgments about the acceptability of intimate violence. Miller and Bukva (2001) found that the seriousness of the injury sustained was an important predictor of how seriously violence was perceived by respondents. James, West, Deters, and Armijo (2000) found that pushing, shoving, grabbing, slapping, punching, biting, and kicking were the most common forms of violence among dating couples. Leonard, Quigley, and Collins (2002) found this to be true with college students as well, reporting that the majority of violence perpetrated by college students consisted of these behaviors. Shook, Gerrity, Jurich, and Segrist (2000) found that college students most often experienced pushing, slapping, or hitting with an object as the primary forms of physical abuse in their dating relationships. Slapping and pushing between dating partners represent a somewhat moderate level of violence (Bethke & DeJoy, 1993). Punching and hitting with objects represent higher levels of violence more likely to inflict injury.

Gender vignette combination. Another characteristic of violent dating interactions that may influence the acceptability of violence is gender combination of the individual initiating the violent act and the individual reacting to that act. Men and women have both been found to initiate violence in interpersonal relationships (e.g., Leonard et al., 2002). However, there are differences in the acceptability of violence initiated by either gender (Forbes, Jobe, White, Bloesch, & Adams-Curtis, 2005). Bookwala, Frieze, Smith, and Ryan (1992) note that males' initiation of violence against their female partners is generally viewed as less acceptable than females' initiation of violence against their male partners. O'Keefe (1997) also found that both sexes were more accepting of females' use of dating violence. In addition, violence inflicted by males on their female partners tends to be viewed as more serious by both men and women.

The purpose of this study is to examine both respondent- and case-level characteristics as potential predictors of the acceptability of dating violence. The complexity of factors regarding the acceptability of dating violence requires a multi-level design to examine respondent characteristics as well as case characteristics. First, four respondent-level dimensions are included in the current study: athletic participation, competitiveness, need to win, and gender. It was anticipated that among these four factors there would be different levels of concern with the seriousness of violent acts in a dating relationship. Second, other research suggests that case-level characteristics reflecting the dynamics of the violent interaction itself may be even more influential than respondent-level characteristics. Three important case-level characteristics that may be salient in judgments about acceptability of violence are: severity of initiator's violence, severity of the recipient's reaction, and the genders of initiators and recipients. The working hypotheses in the present study are the following:

1. Individuals with a higher need to win attitude will rate acts of violence in dating relationships to be more acceptable than those with a lower need to win attitude.

2. Males will rate acts of violence in dating relationships to be more acceptable than females.

3. As the level of violence of the initiator's act increases, acceptability of the recipient's reaction will increase.

4. As the level of violence of the recipient's reaction increases, the acceptability of the recipient's response will decrease.

5. A female recipient's response to a male's initial act of violence will be more acceptable than a male recipient's response to a female's initial act of violence.

METHOD

Sample and Procedure

Participants in this study consisted of 266 male and 393 female students at a large public university in the Midwestern United States. Participants' ages ranged from 17-22. Students enrolled in any one of four social science classes at the university completed a 51-item questionnaire regarding attitudes about competitiveness and desire to win. Acceptability of violence was assessed through responses portrayed in nine vignettes of relationship violence. Questions were also included about participation in high school and collegiate athletics along with others asking for demographic information. A modified consent form was solicited and questionnaires were administered during class.

Measures

Acceptability of violence. Acceptability of violence in dating relationships was assessed using a set of 18 vignettes that depicted situations involving violence between dating partners (see Appendix). Many studies examining attitudes and perceptions regarding dating violence have used such written vignettes (Bethke & DeJoy, 1993; Carlson, 1996; Harmon, Hall, Nash, Formati, & Hopson, 2000; Miller & Bukva, 2001). Vignette methodology allows for systematic and controlled manipulation of any number of independent variables. In addition, vignettes also eliminate concern for extraneous factors that may make findings less credible and increase the internal validity of the research design (Campbell, 1957).

The vignettes used in the current study varied along three factors: severity of the initiator's act, severity of the recipient's response, and initiator-recipient gender combination. Severity of initiator's act consisted of three categories: low (yelling at partner), moderate (pushing/ shoving partner), and high (punching the arm]hitting partner in the back). Severity of recipient's response was categorized as: moderate (pushing/shoving partner); high (punching the arm/hitting partner in the back); and very high (kicking partner in stomach/punching partner in the face). Within each dating scenario, the recipient's response to the initiator's violence was at the same level or higher than the level of the initiator's action. The severity range for a recipient's reaction was higher than the severity range for initiator's act due to the current study's focus on the acceptability of increased violence in a recipient's reaction to initial violence. We propose that individuals will respond to an initial act of violence with an equal or higher degree of violence. As a result of this hypothesis, the severity range for a recipient's reaction is greater. The two initiator-recipient gender combinations were male as initiator and female as recipient of violence and female as initiator and male as recipient. All wording in the vignettes other than changes in the levels of the three factors was identical to ensure that only the changes in factor levels would account for differences in acceptability ratings of the different vignettes.

The combination of levels of the three factors produced a 3 x 3 x 2 full factorial design. To reduce redundancy and possible response fatigue, a modified one-half fractional factorial design was used to create four different sets of 9 vignettes. A one-half fractional factorial design allows for administration of fewer vignettes to each person, but still allows us to show all main and two-way interaction effects for all variables. Each set contained all factor level combinations for initiator and respondent violence, and either five M-to-F/four F-to-M gender combinations, or four M-to-F/five F-to-M gender combinations. Each of the eighteen vignettes was systematically pre-assigned to be present in two of the four sets of nine vignettes that were distributed to the participants. The order in which these nine vignettes appeared in each questionnaire was determined by a random selection process. Each respondent received one of the four sets of vignettes that were randomly distributed. Respondents were asked to read each vignette and indicate how acceptable they believed a particular individual's response was to the violence depicted. Responses ranged from 1 (totally unacceptable) to 6 (totally acceptable). Cronbach's alpha for the vignettes was .80.

Athletic participation. Respondents' participation in high school and collegiate athletics was assessed using a series of seven questions. These focused on the number of years of participation in high school athletics, number of high school sports, and the favorite high school sport. In addition, individuals were asked whether they were presently participating in intramural or Division I athletics. The athletic participation index developed for this study consisted of a scale ranging from 0 = "No participation" to 6 = "Participation in a Division I sport" (see Table 2 for full list of categories).

Competitiveness and need to win. The Sport Orientation Questionnaire (SOQ) developed by Gill and Deeter (1988) was used to measure both competitiveness and need to win in participatory activities. Competitiveness and win orientation are two major sub-scales of the SOQ. Each five-category item in the index ranges from 1 (strongly disagree) to 5 (strongly agree). Scores for each subscale are derived by summing across their respective items. The 13-item competitiveness subscale measures a desire to enter and strive for success in a participatory activity. This scale includes such items as: "I thrive on competition" and "I look forward to competing." The win orientation subscale consists of 6 items measuring desire to win. Examples of items for this sub-scale include: "I hate to lose" and "The only time I am satisfied is when I win." Cronbach's alpha for the 25-item SOQ administered was .95; alpha coefficients for the competitiveness and need to win sub scales were .94 and .86, respectively. These reliability coefficients compare favorably to reliabilities for the SOQ in previous studies (Gill & Deeter, 1988; Wartenberg & McCutcheon, 1998).

RESULTS

Descriptive Statistics

The sample was predominantly White (91%); other ethnicities included African American (5%), Asian American (2%), and Hispanic (2%). Class standing consisted of Freshmen (19%), Sophomores (21%), Juniors (21%), and Seniors (39%). In addition, approximately one-third of both males and females report their high school graduating class as having 100 students or less. In regard to current relationship status among males, 42% indicated they were not currently in a dating or marital relationship, 18% were in a current relationship of between one month and one year in length, 17% were in a current relationship of 1-2 years duration, 20% were in a current relationship of 2-5 years duration, and 3% of the sample were in a current relationship of over 5 years duration. For females, 36% indicated they were not currently in a dating or marital relationship, 25% were in a current relationship between one month and one year in length, 20% were in a current relationship of 1-2 years duration, 15% were in a current relationship of 2-5 years duration, and 4% of the sample were in a current relationship of over 5 years duration (see Table 1).

Table 2 displays athletic characteristics of the participants in the current study. A high number of respondents reported involvement in high school and college sports activities. Only 8% of males and 15% of females state that they have never participated in sports. The majority of the sample reported continued involvement, with 46% of the males and 25% of the females involved in college intramural sports, and 14% of the males and 9% of the females being current participants in Division I college athletics. For those who reported any sports participation, team-only was the preferred sport for 61% of the males and 51% of the females. The number of sports and the type of sport (e.g., team, individual, football, soccer) were not significantly correlated with the acceptability of dating violence among men and women in this study.

Men's and women's mean scores for competitiveness, need to win, and acceptability of violence are shown in Table 3. Mean competitiveness scores for men (M = 3.96, SD = .79) are higher than those for women (M = 3.44, SD = .83) (t = 8.03, p < .001). Mean need to win scores for men (3.58, SD = .82) are also higher than mean scores for women (M = 3.25, SD = .84) (t = 4.98,p < .001). Scores were averaged across the 18 dating vignettes to produce acceptability of violence means. Table 3 shows that males (M = 2.24, SD = .87) are more accepting of violence than are females (M = 2.01, SD = .81) (t = 3.50, p < .001).

Multi-level random intercept regression analyses were run to examine the independent effects of respondent and case characteristics on the acceptability of violence in dating relationships (Table 4). Due to the complex nature of the research design, which includes a multilevel design (vignettes nested within individuals), this study uses the SAC PROC MIXED procedure for multi-level modeling. This approach allows us to examine the unique influence of both case-level variables and respondent-level variables. The between and within individual variances of acceptability of dating violence were 0.56 and 1.55, respectively. We then estimated several nested models to test hypotheses.

Model 1 in Table 4 examines the four respondent-level characteristics of athletic participation, competitiveness, need to win, and respondent gender. No relationship was found between acceptability of violence and two of the three athletic variables, athletic participation and competition. However, the third athletic variable, need to win, has a significant relationship acceptance of violence in dating relationships ([beta] = .24, SE = .06, p < .001). Respondent gender also had a significant relationship with the acceptability of violence, with males being more accepting of violence than women. Using Raudenbush and Bryk's (2002) method to compute explained variance for respondent-level variables, these four respondent characteristics account for 8% of the variance in acceptability of violence scores.

Model 2 adds the first case-level variable, initiator act, to the equation. This variable is a significant predictor of the acceptability of violence in dating relationships ([beta] = .35, SE = .02, p < .001), with violent responses by the recipient to more aggressive acts initiated by the dating partner being viewed as more acceptable than violent responses to less aggressive initiator acts. The association between the case-level initiator act and acceptability of violence is independent of the association between the respondent-level factors and acceptability of violence previously tested, indicated by the stability of the respondent-level betas and [R.sup.2] when the case-level variable is added to the model. This single case-level initiator act accounts for an additional 5% of the variance in acceptability of violence. The percentage of explained variance accounted for by case-level characteristics was computed using a method derived by Snijders and Bosker (1999).

In Model 3 the second case-level variable is added to the model, recipient's reaction to initiator's act of violence. Recipient's reaction has a significant negative association with the acceptability of violence in dating relationships ([beta] = -.55, SE = .02, p < .001). The direction of the relationship indicates that the recipient's response becomes less acceptable the more severe the recipient's reaction. Recipient reaction accounts for an additional 15% of the total variance of acceptability of violence scores.

Model 4 adds the third case-level variable to the model, initiator-recipient gender combination. Results indicate that, controlling for level of response, a female recipient's violence toward a male initiator is more acceptable than a male recipient's violence toward a female initiator ([beta] = .73, SE = .03, p < .001). Initiator-recipient gender combination accounts for a further 10% of the variance in acceptability of violence.

These three case-level variables, reflecting important aspects of violent couple interactions, are independent of the respondent-level variables. This is indicated by the stability of the respondent-level betas and [R.sup.2] values as the case-level variables are successively added to the general model. Together they explain 30% of the variance in the acceptability of violence, which is three times that accounted for by the respondent-level athleticism variables.

As a final step, all interactions among case and respondent characteristics were tested. However, only the three interactions reaching statistical significance are included in the model (Model 5, Table 4). The three interaction effects involve the initiator-recipient gender combination, two with case-level and one with a respondent-level variable. First, the direction of the initiator act by initiator-recipient gender combination interaction coefficient ([beta] = .18, SE = .03, p < .001) indicates that the relationship between initiator act and acceptability of the recipient's reaction is stronger for the M-to-F than for the F-to-M gender combination. That is, the severity of the initiator's violence has a stronger relationship with the acceptability of the recipient's reaction to the violence for male initiators against female recipients than for female initiators against male recipients.

Next, the coefficient for the recipient reaction by initiator-recipient gender combination interaction ([beta] = -.09, SE = .03, p < .01) indicates that the relationship between the recipient's reaction and the acceptability of violence is stronger for the F-to-M combination than for the M-to-F combination. The severity of male reactions to female-initiated violence is more influential in judgments about the acceptability of violence than the severity of female reactions to male-initiated violence.

For the interaction involving the respondent-level variable, need to win by initiator-recipient gender combination, the coefficient ([beta] = .15, SE = .03, p < .001) indicates that need to win's relationship with the acceptability of violence is stronger for the M-to-F combination than for F-to-M. In other words, participants with a high need to win attitude are more supportive of female reactions to male-initiated violence than are those with a low need to win attitude. The decreasing Akaike Information Criterion (AIC), a goodness of fit measure, illustrates the increasing fit to the data across incremental models throughout Table 4. In addition, the respondent-level variable, need to win, remains statistically significant throughout all models indicating that this variable has a significant relationship with the acceptability of violence independently of case-level characteristics.

DISCUSSION

A popular debate has arisen concerning the role of organized sport participation in the development and promotion of prosocial attitudes in young people (Gough, 1998). Participation in sports has been linked positively to self-esteem (Kavussanu & Harnisch, 2000) and achievement attitudes (Curry, Snyder, Cook, Ruby, & Rehm, 1997). However, competitive sports may also be linked to detrimental aspects of development in young people. The current study adds an important dimension to the understanding of athletes and violence in interpersonal relationships. First, as hypothesized, only need to win is related to acceptability of violence, not athletic involvement or competitiveness. This finding supports the notion that competitiveness and need to win should not be considered a single construct, but two distinct constructs. It is individuals' need to win attitude, not their competitiveness or mere participation in athletic activities that leads to a more accepting attitude toward interpersonal violence. These findings suggest that the relationship between sports and dating violence promoted by the media and others may be an overgeneralization, one that unjustly lumps all athletes together.

This popular stereotype linking athletes and dating violence may be harmful in two ways. First, it does not distinguish those athletes who are purely competitive from those who must win the contest. These athletes who enjoy and strive to do their best in their sport activity while being equally appreciative of the best efforts and skills of their opponents, are not distinguished from those athletes who may disregard rules, engage in unethical behavior, and even seek to harm their opponents in their efforts to win.

Second, the need to win may be an attribute evident in a variety of contexts involving gradation and selection, and not limited to the traditional sport environment. For example, the need to win may be a strong motivator for some in the field of music and art, in education and academic settings, in personal dress and acquisition, and especially in relationships requiring decision making, problem solving, and conflict resolution. The relationship between athleticism and the acceptability of violence in interpersonal relationships may be spurious to some degree; rather, it may be the need to win that motivates an individual to be accepting of violence in both sport and non-sport environments. Future studies focusing on interpersonal violence and athletes should examine this dimension further. In particular, studies should attempt to uncover how and when this need to win attitude is instilled in many athletes and why other individuals may not require this attitude at all or to a lesser degree.

However, it is important to note that the respondent-level need to win variable explains only a smaller portion of the acceptability of violence among respondents of this study. Of even more significance are the case-level characteristics of severity of initiator violence, severity of recipient's response, and the initiator-recipient gender combination. Of these three factors, recipient reaction appears to have the most influence, indicating that when assessing acceptability of dating violence, how victims of violence reacts to their partner's initial act may be more important than the initial act of violence. Only slightly less influential were initiator-recipient gender combinations, accounting for 10% of the variation in the acceptability of violence. The initiator act accounted for the least amount of variation in the acceptability of violence (5%) among the three case characteristics examined in this study. The finding that both male and female participants are more accepting of dating violence perpetrated by females than by males is supported by previous research (O'Keefe, 1997). Further, results of this study show males to be even more accepting of interpersonal violence than females in situations where a female reacts to a male's initial act of violence. Overall, the gender structure of the initiator and the recipient of violence appear to have a major impact on the acceptability of violence, both directly and indirectly as it interacts with the other major variables in this study.

A limitation of this study relates to the homogeneous sample population, which consisted predominantly of single, white, adolescents. Future studies that include non-student samples and other ethnic populations are needed to determine the generalizability of the findings of this study to broader populations. Future research should also examine the perceptions of same-sex violence, as we have come to realize that violence is an issue of concern for all types of couples (Beyers, Leonard, Mays, & Rosen, 2000).

This study provided important new information about the acceptability of violence in dating relationships among late adolescents. The unique relationships of both respondent and case-level characteristics with the acceptability of dating violence call for additional studies to implement research designs that account for the multi-level factors that influence perceptions of interpersonal violence. Multi-level research designs helps us better identify and understand the various factors that help us to advance our knowledge of interpersonal violence. These designs allow us to assess simultaneously the relative importance that multiple factors from various domains (i.e., family, individual, or case-level variables) may have on individuals' perceptions of violence.

APPENDIX

Acceptability of Dating Violence Vignettes

1. Adam pushed Sara away when she yelled at
him about his new job hours.

How acceptable is Adam's response?

 1 2 3 4 5 6

Totally Totally
Unacceptable Acceptable

2. Sara pushed Adam away when he yelled at
her about her new job hours.

How acceptable is Sara's response?

 1 2 3 4 5 6

Totally Totally
Unacceptable Acceptable

3. Jason hit Amy in the back when she
yelled at him about the way he was acting at the
party.

How acceptable is Jason's response?

 1 2 3 4 5 6

Totally Totally
Unacceptable Acceptable

4. Amy hit Jason in the back when he
yelled at her about the way she
was acting at the party.

How acceptable is Amy's response?

 1 2 3 4 5 6

Totally Totally
Unacceptable Acceptable

5. Josh punched Heather in the face while
she was yelling at him about his recent behavior.

How acceptable is Josh's response?

 1 2 3 4 5 6

Totally Totally
Unacceptable Acceptable

6. Heather punched Josh in the face while
he was yelling at her about
her recent behavior.

How acceptable is Heather's response?

 1 2 3 4 5 6

Totally Totally
Unacceptable Acceptable

7. Matt shoved Lisa after she pushed him
away while the two were
arguing about the cost of the trip.

How acceptable is Matt's response?

 1 2 3 4 5 6

Totally Totally
Unacceptable Acceptable

8. Lisa shoved Matt after he pushed her away
while the two were arguing about the cost of
the trip.

How acceptable is Lisa's response?

 1 2 3 4 5 6

Totally Totally
Unacceptable Acceptable

9. James got up and punched Melissa in
the arm after she pushed him into the chair
following his comments about one of her friends.

How acceptable is James's response?

 1 2 3 4 5 6

Totally Totally
Unacceptable Acceptable

10. Melissa got up and punched James in the arm
after he pushed her into the chair following her
comments about one of his friends.

How acceptable is Melissa's response?

 1 2 3 4 5 6

Totally Totally
Unacceptable Acceptable

11. John kicked Rebecca in the stomach
after she shoved him away during a dispute
over where they were going to spend Thanksgiving.

How acceptable is John's response?

 1 2 3 4 5 6

Totally Totally
Unacceptable Acceptable

12. Rebecca kicked John in the stomach after
she shoved him away during a dispute over where
they were going to spend Thanksgiving.

How acceptable is Rebecea's response?

 1 2 3 4 5 6

Totally Totally
Unacceptable Acceptable

13. Kevin pushed Julie away after she
walked over and punched him in
the arm during a recent disagreement.

How acceptable is Kevin's response?

 1 2 3 4 5 6

Totally Totally
Unacceptable Acceptable

14. Julie pushed Kevin away after he
walked over and punched her in
the arm during a recent disagreement.

How acceptable is Julie's response?

 1 2 3 4 5 6

Totally Totally
Unacceptable Acceptable

15. Brian turned around and punched Jennifer
in the arm after she hit him in the back when
she realized the keys were locked in the car.

How acceptable is Brian's response?

 1 2 3 4 5 6

Totally Totally
Unacceptable Acceptable

REFERENCES

Bethke, T. M., & DeJoy, D. M. (1993). An experimental study of factors influencing the acceptability of dating violence. Journal of Interpersonal Violence, 8(1), 36-51.

Beyers, J. M., Leonard, J. M., Mays, V. K., & Rosen, L. A. (2000). Gender differences in the perception of courtship abuse. Journal of Interpersonal Violence, 15(5), 451-466.

Bloom, G. A., & Smith, M. D. (1996). Hockey violence: A test of cultural spillover theory. Sociology of Sport Journal, 13, 65-77.

Bookwala, J., Frieze, I. H., Smith, C., & Ryan, K. (1992). Predictors of dating violence: A multivariate analysis. Violence and Victims, 7(4), 297-311.

Campbell, D. T. (1957). Factors relevant to the validity of experiments in social settings. Psychological Bulletin, 54, 297-312.

Carlson, B. E. (1996). Dating violence: Student beliefs about consequences. Journal of Interpersonal Violence, 11, 3-18.

"Cases involving athletes and sexual assault." (2003, December 22). USA Today, p. A.08

Cauffman, E., Feldman, S. S., Arnett Jensen, L., & Jensen Arnett, J. (2000). The (un)acceptability of violence against peers and dates. Journal of Adolescent Research, 15(6), 652-673.

Craig, K. M. (2000). Defeated athletes, abusive mates: Examining perceptions of professional athletes who batter. Journal of Interpersonal Violence, 15(11), 1224-1232.

Crosset, T. W. (1999). Male athletes' violence against women: A critical assessment of the athletic affiliation, violence against women debate. Quest, 51, 244-257.

Crosset, T. W., Ptacek, J., McDonald, M. A., & Benedict, J. R. (1996). Male student-athletes and violence against women: A survey of campus judicial affairs offices. Violence Against Women, 2(2), 163-179.

Curry, L. A., Snyder, C. R., Cook, D. L., Ruby, B. C., & Rehm, M. (1997). The role of hope in academic and sport performance. Journal of Personality and Social Psychology, 73, 1257-1267.

Ferguson, T. J., & Rule, B. G. (1988). Children's evaluations of retaliatory aggression. Child Development, 59(4), 961-968.

Forbes, G. B., Adams-Curtis, L. E., Pakalka, A. H., & White, K. B. (2006). Dating aggression, sexual coercion, and aggression-supporting attitudes among college men as a function of participation in aggressive high school sports. Violence Against Women, 12(5), 441-455.

Forbes, G. B., Jobe, R. L., White, K. B., Bloesch, E., & Adams-Curtis, L. E. (2005). Perceptions of dating violence following a sexual or nonsexual betrayal of trust: Effects of gender, sexism, acceptance of rape myths, and vengeance motivation. Sex Roles, 52(3/4), 165-173.

Foshee, V. A., Ennett, S. T., Bauman, K. E., Benefield, T., & Suchindran, C. (2005). The association between family violence and adolescent dating violence onset: Does it vary by race, socioeconomic status, and family structure? Journal of Early Adolescence, 25, 317-344.

Frieze, I. H. (2005). Female violence against intimate partners: An introduction. Psychology of Women Quarterly, 29(3), 229-240.

Gardner, R. E., & Janelle, C. M. (2002). Legitimacy judgments of perceived aggression and assertion by contact and noncontact sport participants. International Journal of Sport Psychology, 33, 290-306.

Gill, D. L., & Deeter, T. E. (1988). Development of the sport orientation questionnaire. Research Quarterly for Exercise and Sport, 59(3), 191-202.

Gough, R. W. (1998). A practical strategy for emphasizing character development in sport and physical education. The Journal of Physical Education, Recreation & Dance, 69(2), 18-22.

Gray, H. M., & Foshee, V. (1997). Adolescent dating violence: Differences between one-sided and mutually violent profiles. Journal of Interpersonal Violence, 12(1), 126-141.

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

Horney, K. (1937). The neurotic personality of our time. New York: Norton.

Houston, J. M., Harris, P., McIntire, S., & Francis, D. (2002). Revising the competitiveness index using factor analysis. Psychological Reports, 90, 31-34.

Houston, J. M., McIntire, S. A., Kinnie, J., & Terry, C. (2002). A factorial analysis of scales measuring competitiveness. Educational and Psychological Measurement, 62, 284-298.

James, W. H., West, C., Deters, K. E., & Armijo, E. (2000). Youth dating violence. Adolescence, 35(139), 455-4565.

Kavussanu, M., & Harnisch, D. L. (2000). Self-esteem in children: Do goal orientations matter? British Journal of Educational Psychology, 70(2), 229-242.

Lenzi, A., Bianco, I., Milazzo, V., Placidi, G., & Castrogiovanni, P. (1997). Comparisons of aggressive behavior between men and women in sport. Perceptual and Motor Skills, 84, 139-145.

Leonard, K. E., Quigley, B. M., & Collins, R. L. (2002). Physical aggression in the lives of young adults: Prevalence, location, and severity among college and community samples. Journal of Interpersonal Violence, 17(5), 533-550.

Lewis, S. F., & Fremouw, W. (2001). Dating violence: A critical review of the literature Clinical Psychological Review, 21(1), 105-127.

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.

Makepeace, J. M. (1981). Courtship violence among college students. Family Relations, 30, 97-102.

Marcus, R. R., & Swett, B. (2002). Violence and intimacy in close relationships. Journal of Interpersonal Violence, 17(5), 570-586.

Miller, J., & Bukva, K. (2001). Intimate violence perceptions: Young adults' judgments of abuse escalating from verbal arguments. Journal of Interpersonal Violence, 16(2), 133-150.

Mintah, J. K., Huddleston, S., & Doody, S. G. (1999). Justifications of aggressive behavior in contact and semicontact sports. Journal of Applied Social Psychology, 29, 597-605.

Morey, N., & Gerber, G. L. (1995). Two types of competitiveness: Their impact on the perceived interpersonal attractiveness of women and men. Journal of Applied Social Psychology, 25(3), 210-222.

Nack, W., & Munson, L. (1995). Sports' dirty secret. (Special report). Sports Illustrated, 83(5), 62-73.

Neufeld, J., McNamara, J. R., & Ertl, M. (1999). Incidence and prevalence of dating partner abuse and its relationship to dating practices. Journal of Interpersonal Violence, 14(2), 125-137.

Nixon, H. L. (1997). Gender, sport, and aggressive behavior outside sport. Journal of Sport and Social Issues, 21(4), 379-391.

O'Keefe, M. (1997). Predictors of dating violence among high school students. Journal of Interpersonal Violence, 12(4), 546-568.

O'Keefe, M., & Treister, L. (1998). Victims of dating violence among high school students. Violence Against Women, 4(2), 195-223.

Perry, A. R., & Fromuth, M. E. (2005). Courtship violence using couple data: Characteristics and perceptions. Journal of Interpersonal Violence, 20(9), 1078-1095.

Pleck, E. H. (2004). Domestic tyranny: The making of American social policy against family violence from colonial times to the present. Urbana, IL: University of Illinois Press.

Raudenbush, S. W., & Bryk, S. A. (2002). Hierarchical linear models. Thousands Oaks, CA: Sage.

Roscoe, B. (1985). Courtship violence: Acceptable forms and situations. College Student Journal, 19, 389-393.

Rossi, P. H., Schuerman, J., & Budde, S. (1999). Understanding decisions about child maltreatment. Evaluation Review, 23(6), 579-598.

Ryckman, R. M., & Hamel, J. (1992). Female adolescents' motives related to involvement in organized team sports. International Journal of Sport Psychology, 23, 147-160.

Ryckman, R. M., Hammer, M., Kaczor, L. M., & Gold, J. A. (1996). Construction of a personal development competitive attitude scale. Journal of Personality Assessment, 66(2), 374-385.

Ryckman, R. M., Libby, C. R., van den Borne, B., Gold, J. A., & Lindner, M. A. (1997). Values of hypercompetitive and personal development of competitive individuals. Journal of Personality Assessment, 69(2), 271-283.

Ryckman, R. M., Thornton, B., Gold, J. A., & Burckle, M. A. (2002). Romantic relationships of hypercompetitive individuals. Journal of Social and Clinical Psychology, 21(5), 517-530.

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), 1-22.

Snijders, T., & Bosker, R. (1999). Multilevel analysis: An introduction to basic and advanced multilevel modeling. London: Sage.

Taylor, C. A., & Sorenson, S. B. (2005). Community-based norms about intimate partner violence: Putting attributions of fault and responsibility into context. Sex Roles, 53(7/8), 573-589.

Wartenberg, L., & McCutcheon, L. (1998). Further reliability and validity of data on the sport orientation questionnaire. Journal of Sport Behavior,. 21(2), 219-221.

Requests for reprints should be sent to Michael J. Merten, Department of Human Development and Family Science, Oklahoma State University, 1111 Main Hall, Tulsa, Oklahoma 74106. E-mail: [email protected]

Table 1. Demographic characteristics of the participants included
in the study, by gender

 Males Females
Age
 17-18 14% 15%
 19-20 43% 45%
 21-22 38% 37%
 Older than 22 5% 3%
Year in School
 Freshman 17% 25%
 Sophomore 22% 21
 Junior 19% 22%
 Senior 42% 32%
Ethnicity
 White-Caucasian 90% 92%
 African-American 5% 4%
 Asian-American 4% 2%
 Hispanic 1% 2%
High School Graduating Class
 50 or less 9% 16%
 51-100 25% 19%
 101-250 28% 24%
 251-350 18% 17%
 351 or more 20% 24%
Current Relationship Status
 Not currently dating anyone 42% 36%
 Less than 1 year 18% 25%
 1-2 years 17% 20%
 2-5 years 20% 15%
 More than 5 years 3% 4%

Note: N=266 for males; N=393 for females.

Table 2. High school and collegiate athletic participation
characteristics, by gender

 Males Females

Highest level of participation (a)
 No participation 8% 15%
 Freshman (HS) 3% 5%
 Sophomore (HS) 3% 7%
 Junior (HS) 4% 8%
 Senior (HS) 22% 31%
 College Intramurals 46% 25%
 College Division I 14% 9%
Number of high school sports
 None 11% 19%
 One 20% 21%
 Two 22% 21%
 Three 23% 20%
 Four 17% 12%
 Five or more 7% 7%
Favorite type of participation (b,c)
 Individual/team sport 39% 49%
 Team-only sport 61% 51%
 (n = 236) (n = 319)
Favorite individual/team sport (d)
 Cross-country 14% 14%
 Golf 18% 11%
 Gymnastics 2% 4%
 Swimming/Diving 9% 16%
 Tennis 8% 13%
 Track & Field 17% 23%
 Wrestling 26% --
 Other 6% 19%
 (n = 94) (n = 155)

Table 3. Means for competitiveness, need to win, and acceptability
of violence, by gender

 Males Females t
Competitiveness
 Mean 3.96 3.44 8.03 ***
 Standard Deviation. 0.79 0.83
 (n = 266) (n = 393)
Need to win
 Mean 3.58 3.25 4.98 ***
 Standard Deviation 0.82 0.84
 (n = 266) (n = 393)
Acceptability of Violence
 Mean 2.24 2.01 3.50 ***
 Standard Deviation 0.87 0.81
 (n = 266) (n = 393)

* p < .05 ** p < .01 *** p < .001

Table 4. Unstandardized multi-level random intercept regression
coefficients for respondent and case characteristics on the
acceptability of dating violence (standard errors in parentheses)

Predictor M1 M2 M3

Respondent Characteristics
 Athletic Participation -.01 -.01 -.01
 (.02) (.02) (.02)
 Competitiveness -.06 -.06 -.06
 (.06) (.06) (.06)
 Need to Win .24 *** .24 *** .24 ***
 (.03) (.03) (.03)
 Gender (0=male) -.20 ** -.20 ** -.20 **
 (.03) (.04) (.04)
Case Characteristics
 Initiator Act .35 *** .35 ***
 (.02) (.02)
 Recipient Reaction -.55 ***
 (.02)
 Gender Combination (0=F-M)
Interactions
 Initiator Act*Gender
 Combination
 Recipient Reaction*Gender
 Combination
 Need to Win*Gender
 Combination
Respondent-level variance
 (null model=.56) .50 .51 .53
 Explained variance, [R.sup.2] .09 .09 .09
Case-level variance
 (null model=1.55) 1.54 1.46 1.23
 Explained variance, [R.sup.2] .00 .05 .20
AIC 20,366 20,059 19,169

Predictor M4 M5

Respondent Characteristics
 Athletic Participation -.02 -.02
 (.02) (.02)
 Competitiveness -.06 -.06
 (.06) (.06)
 Need to Win .24 *** .21 ***
 (.03) (.02)
 Gender (0=male) -.20 ** -.13 *
 (.04) (.02)
Case Characteristics
 Initiator Act .35 *** .35 ***
 (.02) (.02)
 Recipient Reaction -.55 *** -.55 ***
 (.02) (.02)
 Gender Combination (0=F-M) .73 ** .73 **
Interactions
 Initiator Act*Gender .18 ***
 Combination (.03)
 Recipient Reaction*Gender -.09 **
 Combination (.03)
 Need to Win*Gender .15 ***
 Combination (.03)
Respondent-level variance
 (null model=.56) .54 .54
 Explained variance, [R.sup.2] .10 .10
Case-level variance
 (null model=1.55) 1.08 1.07
 Explained variance, [R.sup.2] .3 .31
AIC 18,503 18,435

* p<.05 ** p<.01 *** p<.001