Gemini IA a noté :
Dafotec Récupération De Données

Le 2025-04-26 04:36:23

Competitive Aggressiveness, Anger, and the Experience of Provocation in Collegiate Athletes Competitive Aggressiveness, Anger, and the Experience of Provocation in Collegiate Athletes Michael E. Berrebi Follow this and additional works at: https://researchrepository.wvu.edu/etd Recommended Citation Berrebi, Michael E., "Competitive Aggressiveness, Anger, and the Experience of Provocation in Collegiate Athletes" (2018). Graduate Theses, Dissertations, and Problem Reports. 5194. https://researchrepository.wvu.edu/etd/5194 This Dissertation is protected by copyright and/or related rights. It has been brought to you by the The Research Repository @ WVU with permission from the rights-holder(s). You are free to use this Dissertation in any way that is permitted by the copyright and related rights legislation that applies to your use. For other uses you must obtain permission from the rights-holder(s) directly, unless additional rights are indicated by a Creative Commons license in the record and/ or on the work itself. This Dissertation has been accepted for inclusion in WVU Graduate Theses, Dissertations, and Problem Reports collection by an authorized administrator of The Research Repository @ WVU. For more information, please contact researchrepository@mail.wvu.edu. Competitive Aggressiveness, Anger, and the Experience of Provocation in Collegiate Athletes Michael E. Berrebi, M.S. Dissertation submitted to the College of Physical Activity and Sport Sciences at West Virginia University in partial fulfillment of the requirements for the degree of Doctor of Philosophy in Sport & Exercise Psychology Edward Etzel, Ed.D., Chair Jack Watson II, Ph.D. Scott Barnicle, Ph.D. David Mitchell, Ph.D. Department of Sport Sciences Morgantown, West Virginia 2018 Keywords: anger, aggression, aggressiveness, provocation, college athletes © 2018 Michael E. Berrebi ABSTRACT Competitive Aggressiveness, Anger, and the Experience of Provocation in Collegiate Athletes Michael E. Berrebi In sport, aggressive behavior is a potentially harmful byproduct of uncontrolled anger. In addition, it is known that provocation can lead to both anger and aggressive retaliation. However, despite the potential consequences of aggressive behavior, little is known about levels of competitive anger and aggressiveness in athletes, and it is unclear if differences exist by gender or type of sport. Little research has also explored intervention approaches to help athletes better manage anger and aggression. Therefore, the purpose of this study was to explore competitive aggressiveness, anger, and the experience of provocation among collegiate athletes. Participants were 243 male and female contact or collision sport athletes competing at NCAA Division I, II, and III universities across the country. Participants filled out questionnaires assessing both competitive aggressiveness and anger and the experience of provocation. Overall, it was found that male athletes scored significantly higher than female athletes on competitive aggressiveness, as well as experiencing more frequent provocative behavior and more negative and intense responses to provocation. Collision sport athletes were also found to be higher in competitive aggressiveness and anger, regardless of gender. Division I and II athletes were found to be significantly higher than Division III athletes in competitive aggressiveness and anger. TABLE OF CONTENTS iii Page # 1. INTRODUCTION……………………………………………………………………….........1 2. METHODS……………………………………………………………………………….......11 2.1. Participants………………………………………………………………………....11 2.2. Research Design and Sampling…………………………………….……………....11 2.3. Instrumentation…………………………………………………………………......12 2.3.1. Demographic questionnaire…………………………………………........12 2.3.2. Competitive aggressiveness and anger…………………………………...12 2.3.3. Provocation…………………………………………………………….....14 2.3.4. Pilot study…………………………………………………………….......15 2.4. Procedure……………………………………………………………………….......16 2.5. Data Analysis…………………………………………………………………….....17 3. RESULTS……………………………………………………………………....….………....20 3.1. Demographics and Descriptive Statistics ………………………………….............20 3.2. Data Cleaning and Assumption Testing ……………………...…….……………...21 3.3. Bivariate Statistics……………………………………………………………….....22 4. DISCUSSION……………………………………………………………………...….…......27 4.1. Competitive Aggressiveness and Anger……………………………………..…....27 4.1.1. Gender comparisons………………………………………………….......27 4.1.2. Sport type comparisons………………………………………………......28 4.1.3. Division level comparisons…………….....………………………….......32 4.2. The Experience of Provocation………………………………………………........35 iv 4.2.1. Gender comparisons………………………………………………….......35 4.2.2. Sport type comparisons………………………………………………......36 4.2.3. Division level comparisons…………………………………………........38 4.3. Future Research and Directions…………………………………….……......…....39 4.3. Study Strengths and Limitations……………………………....…………......…....42 5. REFERENCES…………………………………………………………………………….....44 6. APPENDICES……………………………………………………………………………......53 6.1. Appendix A: Data Tables………………….…………………………..………........53 6.2. Appendix B: Extended Review of Literature…………………………..……….......75 6.2.1. Part I: Competitive Anger in Sport………………………………….......75 6.2.2. Part II: Aggressive Behavior in Sport…………………………………...88 6.2.3. Part III: Managing Anger and Aggressive Behavior in Sport…….........107 6.2.4. Significance of Study…………………………………………………..115 6.2.5. References…………………………………………………...................121 6.3. Appendix C: Assistant and Head Coach Recruitment Letter……..…………….....139 6.4. Appendix D: Participant Cover Page………………………………..….................140 6.5. Appendix E: Demographic Questionnaire…………………………….…..............141 6.6. Appendix F: Competitive Aggressiveness and Anger Scale………………….......142 6.7. Appendix G: Sport Provocation Questionnaire…………….……………….…….143 6.8. Appendix H: SPQ Pilot Data and Feedback…………………………..………......146 AGGRESSIVENESS, ANGER, AND PROVOCATION IN SPORT 1 Introduction In sport, there is sufficient anecdotal and empirical evidence that suggests being able to manage one’s emotions is a key factor in influencing performance (e.g., Hanin, 2010; Lane, 2007; Woodcock, Cumming, Duda, & Sharp, 2012). In addition to athletic ability, teamwork, and strategy, sport performance also hinges upon the ability of the athlete to regulate emotions. Throughout a competition, athletes experience a variety of positive and negative emotions that can influence motivation and change both physical and cognitive performance (Botterill & Brown, 2002). More recently, it has been suggested that being able to regulate emotions in sport is an important determinant of performance outcomes both for individual athletes (e.g., Lane, Beedie, Jones, Uphill, & Devonport, 2012) and teams (Wagstaff & Weston, 2014). Anger has been described as “an emotional state that consists of feelings that vary in intensity, with associated activation or arousal of the autonomic nervous system” (Spielberger & Reheiser, 2009, p. 281). A key component of this definition is the lack of judgment regarding whether anger is a “good” or “bad” thing, but rather a normal, human emotion. In fact, experiencing anger is somewhat unavoidable, especially in high-stress, pressure-packed environments that competitive sports embody. Whether an athlete’s anger becomes problematic appears to be less about the fact that it is present and more about the nature and severity of behavioral outcomes (Kassinove & Tafrate, 2002). In fact, researchers who have explored whether anger helps or hurts performance in sport have presented mixed results (e.g., Robazza & Bortoli, 2007; Ruiz & Hanin, 2011). It seems the most important factor is how anger is interpreted and managed by athletes (Hanin & Syrja, 1995). Despite a growing body of research focused on exploring emotions in sport, anger has not been thoroughly investigated. This is in spite of the knowledge that anger is one of the most AGGRESSIVENESS, ANGER, AND PROVOCATION IN SPORT 2 common emotions that athletes experience during competitive play (Sofia & Cruz, 2016). While anger is an emotion expressed by a high percentage of athletes, researchers suggest gender, competition level, and type of sport may play a factor in the level of anger experienced by athletes (Maxwell, Visek, & Moores, 2009). Some findings suggest the possibility that athletes with a perfectionist orientation are at a greater risk for experiencing anger when the pressure is on (Vallance, Dunn, & Dunn, 2006). It has also been identified that male and female athletes may cope with anger in similar ways (Bolgar, Janelle, & Giacobbi, 2008). Overall, the nature and degree of differences among gender and level/type of sport is still largely unknown. Anger has been associated with a number of negative performance outcomes such as misuse of energy, a decrease in achievement and motivation, and the possibility of violent behavior (Robazza et al., 2006). The fact that uncontrolled anger can lead to aggressive or violent behavior has been known for decades (e.g., Berkowitz, 1993; Feindler & Ecton, 1994). However, few studies have been published in the years since that explore the relationship between competitive anger levels and aggressive behavior. Maxwell (2004) found that simply ruminating about past experiences that have caused anger could increase the possibility of aggression. In addition, being provoked and having thoughts of revenge have been found to be significantly related to self-reported aggression (Maxwell, Moores, & Chow, 2007). Unfortunately, it remains difficult to assess competitive anger, which is required to provide a more clear understanding of how it impacts aggression. Part of the problem stems from how to accurately assess anger and aggressive behavior. Outside of sport, instruments such as the State Trait Anger Expression Inventory (STAXI; Spielberger, 1988) have been constructed to help assess the experience, expression, and control of anger. However, nearly thirty years since the introduction of the STAXI, it still has not been normed on athletes or become a standard form AGGRESSIVENESS, ANGER, AND PROVOCATION IN SPORT 3 of measurement in sport psychology research. There currently exists no standard method to assess anger or aggressive acts in sport. Being able to utilize trained observers to identify aggression would be ideal. This would help identify what led to the aggressive act, the act itself, and the resulting consequences. This is unfortunately a time-consuming and potentially expensive process. In addition, aggressive behavior may not always be noticeable on a game-to game basis (Maxwell & Moores, 2007). The Competitive Aggressiveness and Anger Scale (CAAS) is a notable instrument used to assess both competitive anger and aggressiveness, or the tolerance of aggressive behavior and inclination to aggress. The CAAS (Maxwell & Moores, 2007) is a 12-item instrument that assesses anger (e.g., frustration from missed calls from referees) and aggressiveness in a sport setting. While aggressiveness is a trait rather than a behavior (like aggression), it is one of the better ways to assess the probability of aggressive behavior in a proactive way (i.e., before it actually happens). Few researchers have utilized the CAAS since its inception (e.g., Visek Maxwell, & Hurst, 2011), but it appears to be an efficient and promising instrument for use in the exploration of anger and aggressiveness in athletes. Another consideration in the study of aggression is the lack of a clear consensus on how to define and differentiate assertive and aggressive behavior, and also what constitutes “violence” (Abrams, 2010; Kirker, Tenenbaum, & Mattson, 2000). Currently, one of the more common methods to help understand aggression is to split these types of behaviors into instrumental and hostile types (Husman & Silva, 1984). The major difference is the distinction that instrumental aggression may cause harm, but has the overarching goal of pursuing a nonaggressive goal (such as scoring points). Hostile aggression suggests a primary intent of injuring another person physically or psychologically. Abrams (2010) suggested that violence AGGRESSIVENESS, ANGER, AND PROVOCATION IN SPORT 4 can be spontaneous or planned, with the end goal to hurt another person. Spontaneous violence seems to be a result of being provoked, while planned violence is intentional and a “complete system failure” (p. 6). Abrams went on to suggest that athletes displaying this type of behavior should be immediately removed from the playing field and, in extreme instances, even prosecuted. With these definitions as resources, hostile and planned violence are the most dangerous. What must happen for athletes to feel it necessary to display these types of behaviors? There is no doubt that feeling frustrated and angry are important factors in understanding aggressive behavior (Berkowitz, 1993; Feindler & Ecton, 1994; Robazza et al., 2006;). However, simply feeling angry does not automatically cause one to lash out aggressively. What other factors come into play? To help answer this question, a number of theories have been put forth to help understand the potential causes of aggression. Examples include the instinct theory, frustration aggression theory (Dollard, Doob, Miller, Mowrer, & Sears, 1939), social learning theory (Bandura, 1973), theory of moral reasoning (Bredemeier, 1994), and revised frustration-aggression theory (Berkowitz, 1965; 1993). Instinct theory suggests aggression is an innate human instinct that builds up until it must be expressed either directly or cathartically (through sports for example). Based on Albert Bandura’s seminal work, social learning theory predicts aggression is learned by observation, and aggression that is reinforced is likely to reoccur if not penalized. The theory of moral reasoning postulates that how likely a person is to aggress is based on their level of moral development. One of the more widely held views, Berkowitz’s revised frustration-aggression theory suggests frustration only leads to aggression when an individual encompasses the social cues that indicate aggression is appropriate in a particular instance. This theory is important in AGGRESSIVENESS, ANGER, AND PROVOCATION IN SPORT 5 understanding how frustration in competition can play an important factor in determining aggressive behavior, based on the individual and his/her environment. One of the least understood experiences related to being frustrated and taking it out aggressively, is the act of provocation. According to Maxwell, Moores, and Chow (2007), provocation is “any behavior [of another person or persons] that is judged by the victim as aversive or unpleasant, normally with intent on the part of the perpetrator implicitly assumed, and rousing feelings of anger, frustration, or fear” (p. 11). This definition suggests that provocation is assumed to involve intent by the perpetrator. This is a key distinction, as it should not be considered provocation if an athlete, simply by competing hard and without harmful intent, frustrates or angers an opponent to the point of him/her lashing out. In his work on revised frustration-aggression hypothesis, Berkowitz (1989) suggested that provocation, along with frustration and aversive stimuli, could lead to aggression through the generation of negative affect that is interpreted by the individual as anger. Provocation is one of the clearest antecedents of aggression in both non-sporting (e.g., Harris, 1993) and sporting (e.g., Huang, Cherek, & Lane, 1999) environments. Outside of sport, researchers have suggested that provocation may cancel out any inhibitory effects that empathy can have on aggressive behavior (Phillips & Giancola, 2007; Stranger, Kavusannu, McIntyre, & Ring 2016). Other research suggests that provocation itself is frequently interpreted as offensive and has been linked to increases in overall anger levels (Mohr et al., 2007). Maxwell (2004) was one of the first researchers to focus on understanding the experience and consequences of provocation in athletics. He reported that provocation might be positively associated with aggression in athletes. This mirrored similar findings of research on norm breaking behaviors in sport, in which Kirker, Tenenbaum, and Matteson (2000) observed that AGGRESSIVENESS, ANGER, AND PROVOCATION IN SPORT 6 mild aggressive acts often followed provocation acts in a vengeful manner. In the worst cases, this sometimes resulted in more severely violent aggression. Maxwell (2004) claimed athletes from team sports report greater frequency of provocation than athletes who compete in individual sports. A few years later, Maxwell and Moores (2007) suggested that males experience provocation at a greater frequency than females, suggesting that males may perceive more incidences as provoking in sport. Maxwell, Visek, and Moores (2009) found that athletes who competed in high contact team sports tended to experience higher provocation while playing sport. Findings of this study suggested that provocation can be seen as a justification for retaliatory aggression, but not always between the original combatants, at least in team sports. Importantly, it has also been suggested that individuals with high trait anger are more likely to feel more readily provoked and endorse aggressive acts as a result (Maxwell, Visek, & Moores, 2009). This hypothesis was originally put forth by Spielberger (1988), who suggested in his state-trait anger theory that high trait anger individuals experience anger more frequently and longer than low trait anger individuals. In addition, these people are more likely to express this anger in an aggressive or harmful manner. Unfortunately, this theory has not been explored in the realm of athletics, and little knowledge exists regarding the characteristics of athletes that are more or less likely to have high trait anger (or respond aggressively when triggered or provoked). Clearly, some research exists that suggests provocation is an important factor in determining anger and aggressive behavior (e.g., Maxwell, Visek, & Moores, 2009; Stranger et al., 2016). However, assessing provocation is difficult since ideally (as with aggressive behavior), provocation is assessed by observation. However, this method is timely and unpredictable, in addition to somewhat subjective. An act perceived as provocative to one athlete may not necessarily be perceived the same way by others. Furthermore, sometimes provocation AGGRESSIVENESS, ANGER, AND PROVOCATION IN SPORT 7 is verbal, which can be difficult to assess by outside observers. Currently, neither observer assessment or provocation nor adequate self-report measures have been documented in the literature. Maxwell and Moores (2006) attempted to meet the need of a self-report assessment in sport by crafting the Provocation in Sport Questionnaire (PSQ), a six-item self-report questionnaire that inquires about incidences of provocation common in many sports. These statements were scored by athletes on a five-point Likert-type scale to help understand the frequency of each provocation type and the corresponding intensity of anger. Unfortunately, this scale is no longer available and no other known instrument exists in which to assess the experience of provocation in sport. Since provocation seems to be a key factor in understanding anger and aggressive behavior in sports, a novel assessment of provocation is needed to help develop this area of sport psychology research. Therefore, the purpose of this study was to explore the relationship between self-reported competitive aggressiveness, anger, and provocation in collegiate male and female contact and collision sport athletes. Contact and collision sport athletes are an important population to investigate because they are frequently in close proximity with opponents during competition and are therefore more likely to engage in aggressive behavior. Non-contact sport athletes were not included in this investigation. Based on the modest amount of research available on anger, aggression, and provocative behavior in sport, this study explored these variables while controlling for gender and type of sport (i.e., contact vs. collision). It has been suggested that gender and type of sport have significant competitive anger and aggression differences based on previous research (e.g., Maxwell, Visek, & Moores, 2009). For example, some studies have suggested that male athletes AGGRESSIVENESS, ANGER, AND PROVOCATION IN SPORT 8 are more aggressive than their female counterparts, both on and off the field (e.g., Burton & Marshall, 2005; Coulomb-Cabagno & Rascle, 2006). Male athletes have also been found to perceive aggression as more legitimate than females (e.g., Bredemeier, 1985; Gardner & Janelle, 2002; Tucker & Parks, 2001). Other research has disputed these findings (e.g., Bolgar, Janelle, & Giacobbi, 2008; Keeler, 2007). Ultimately, aggression studies that focused on gender have revealed conflicting findings, with no significant differences being reported between males and females (Kimble et al., 2010). With this knowledge in mind, the first major research question in the study was: “Does competitive anger, aggressiveness, and experience of provocation vary between male and female college athletes?” The second major research question was: “Does competitive anger, aggressiveness, and provocation vary between contact and collision sport college athletes?” Based on prior research, there were two main hypotheses in this study. The first was that male athletes would score higher on the CAAS anger and aggressiveness subscales, in addition to experiencing (and responding negatively to) more provocation than female athletes. The second was that collision sport athletes would score higher on the CAAS anger and aggressiveness subscales, in addition to experiencing more provocation and more frequently responding to provocation than contact athletes. There were a number of assumptions made by the researcher in this study. For example, although prior research was ambiguous (e.g., Coulomb-Cabagno & Rascle, 2006, Keeler, 2007), it was assumed that gender does play a significant role in the relationship between anger, aggressiveness, and provocation. It was also assumed that contact and collision sport athletes differed in aggressiveness, anger, and/or experiences of provocation. Closely related to that is the assumption that provocation played a significant role in determining one’s level of competitive AGGRESSIVENESS, ANGER, AND PROVOCATION IN SPORT 9 anger and aggressiveness. Finally, it was assumed that the original sport provocation questionnaire being used in this study adequately assessed the experience of provocation in collegiate sport. There is currently a major gap in the understanding of how anger, aggressive behavior, and provocation interact in sport settings. There exists no valid assessment tool for the experience of provocation in sport. In order to develop interventions to mitigate anger responses to provocation, a more clear understanding of the interaction of aggression, anger and provocation is required. Understanding the characteristics of athletes who are prone to high anger levels, aggressiveness, and experiencing provocation will help in the development of more focused interventions for these individuals. Male athletes may not actually be more aggressive than female athletes, despite popular perception, and therefore interventions for one gender could be utilized effectively for the other. However, anger levels and aggressiveness may be significantly different, in which case gender-specific programming and interventions may be required. While the current study is largely exploratory, increasing the understanding of provocation can be helpful for athletes, coaches, sport psychology consultants, and even referees. Being able to understand how frequent provocative behavior occurs, how it is being displayed, how it affects athletes, and how often it is reciprocated is valuable both theoretically and practically. Team sport contact and collision athletes can better understand what to expect in competition and coaches may be able to better prepare athletes for provocative behavior. Coaches of certain types of teams (e.g., male teams or collision sports) need better information, education, and programming to help promote an environment more conducive to safe, sanctioned play. AGGRESSIVENESS, ANGER, AND PROVOCATION IN SPORT 10 It is vital to understand who the at-risk athletes may be so they can be provided better education and emotional regulation resources, as it is possible high trait anger athletes are more likely to be involved in aggressive on-field acts (Maxwell, Visek, & Moores, 2009). Similarly, coaches and sport psychology consultants can use the findings to better understand aggression and where and why it is likely to occur. Referees could foster a better sense of how to officiate sport to eliminate provocative behavior before it turns into dangerous reciprocation. Nearly all stakeholders involved in collegiate sport could benefit from the findings of this study, as a better understanding of the relationship between provocation and competitive anger and aggression in sport is necessary. AGGRESSIVENESS, ANGER, AND PROVOCATION IN SPORT 11 Methods Participants Participants were 243 (M = 124) male and female athletes from NCAA Division I, II, and III universities throughout the continental United States. The age range of participants was between 18-23 years. Participants were sampled from both contact and collision team sports. To be considered a contact sport, in-game contact is allowed, but extreme contact or direct collisions are not implicit or required by the rules of the sport (Keeler, 2007; Silva, 1983). For the purposes of this study, men’s and women’s basketball, field hockey, women’s lacrosse, and men’s and women’s soccer were all considered contact sports. For collision sports, collisions are necessary and integral, and they are also considered a predesigned aspect of appropriate goal-directed behavior in that sport (Keeler, 2007; Silva, 1983). Collision sports sampled for this study included football, men’s lacrosse, men’s and women’s rugby, and men’s and women’s ice hockey. Research Design and Sampling This study employed a quantitative, survey-based approach to investigate the relationships between three phenomena of interest: 1) competitive anger, 2) aggressiveness, and 3) provocation in collegiate sport. The researcher used purposive sampling to select participants who met the inclusion criteria for the study (Creswell, 2014). This criterion was being a current NCAA Division I, II, or III athlete and competing in selected contact or collision sports. This was also a sample of convenience due to the researcher using personal contacts at various universities to gain access to athletes competing at those universities. Sampling began during the 2017 summer “off-season” period and ended in October of the Fall 2017 competitive season. After the initial sampling period, additional attempts at recruiting participants were implemented AGGRESSIVENESS, ANGER, AND PROVOCATION IN SPORT 12 every two weeks. This entailed follow-up email reminders to coaches requesting that they pass along the survey link to their athletes. Instrumentation Coaches were contacted via email and asked to forward the survey link to athletes on their team (see Appendix C). If participants were willing to partake, they clicked on the survey link to find an informational cover page (see Appendix D), followed by the online survey that consisted of a short demographics questionnaire, a sports-based competitive anger and aggressiveness assessment, and a provocation questionnaire. The order of the questionnaires was randomized for participants to help reduce the possibility of an order effect (Creswell, 2014). The provocation assessment was an original questionnaire that focused on the participant’s experience of provocation during athletic competition. It was detailed in the participant cover page that consent was implied by completing the online survey. All three sections of the online survey was predicted to take approximately fifteen minutes to fill out. Demographic questionnaire. The researcher used a demographic survey to gather the following seven variables of interest: 1) age, 2) gender, 3) race/ethnicity, 4) year in school, 5) type of sport, 6) name of university, and 7) NCAA division level. While some identifying information such as school of enrollment was collected, all information was kept confidential and secured online using password-protected software. Once downloaded, study data was stored in encrypted files on the researcher’s personal computer. Demographic information was also collected at the beginning of the survey, which has been shown to increase item response rate for participants who begin the survey (Teclaw, Price, & Osatuke, 2011). (See Appendix E) Competitive aggressiveness and anger. The Competitive Aggressiveness and Anger Scale (CAAS; Maxwell & Moores, 2007) is a 12-item self-report questionnaire designed to AGGRESSIVENESS, ANGER, AND PROVOCATION IN SPORT 13 assess competitive anger and aggressiveness, or the tolerance of aggression and inclination to aggress, in sport settings. The CAAS is divided into two subscales: a) anger, and b) aggressiveness subscales, with six items in each subscale rated on a five-point Likert type scale from 1=almost never to 5=almost always. A sample item from the anger subscale is: “I find it difficult to control my temper during a match”. A sample item from the aggressiveness subscale is: “It is acceptable to use illegal physical force to gain an advantage”. Using confirmatory factor analysis (CFA), the authors reported good internal consistencies for each subscale and the total scale score. These Cronbach alphas include: anger (α = .78), aggressiveness (α = .84), and total (α = .87; Maxwell & Moores, 2007). These values all fall within the generally accepted reliability levels as determined by Nunnally and Bernstein (1994). Concurrent validity was established with subscales of the Buss-Perry Aggression Questionnaire (BPAQ; Buss & Perry, 1992). In addition, the authors found adequate one-month test-retest statistics for the subscales and total scale score. Discriminant validity was established using peer perception of aggressive orientation. The anger and aggressiveness subscales were also found to be moderately correlated to each other (e.g., α = .59 and .60), suggesting that they are related but not too similar in nature. This is in agreement with literature that suggests a relationship exists between anger and aggression (e.g., Buss & Perry, 1992; Maxwell, 2004), in addition to a clear link with aggressiveness. When crafting and examining the psychometric properties of the scale, the authors found differences in gender and type of sport, with males and contact sport athletes reporting a higher tendency to aggress than females and non-contact athletes (Maxwell & Moores, 2007). The CAAS is one of the only known sport-specific scales to assess competitive anger and aggressiveness in athletes. However, the factor structure of the CAAS has been found to be AGGRESSIVENESS, ANGER, AND PROVOCATION IN SPORT 14 replicable with both Americans and English-speaking Chinese athletes. It has been identified as an appropriate way of assessing athletes most likely to display acts of aggression (Visek, Maxwell, Watson, & Hurst, 2011). The CAAS was intended to be a trait measure, so it does not take into account fluctuations in state anger and aggressiveness throughout an athletic competition. Other limitations include that it lacks a social desirability check and was constructed using a non-elite sample of athletes. (See Appendix F) Provocation. To assess participants’ experience of provocation during competition, an original questionnaire was utilized in this study. This questionnaire is based on the Provocation in Sport Questionnaire (PSQ; Maxwell & Moores, 2006), which was a scale that contained six short statements representing incidences of provocation that are common in sports. In the only confirmed published study utilizing it, the PSQ was translated to use in Chinese (Maxwell, Moores, & Visek, 2009). The scale measured the frequency at which respondents experienced various types of perceived provocation and the corresponding self-reported intensity of associated anger on five-point Likert type scales. The PSQ was the only known instrument to study the experience of provocation in sport. Unfortunately, the full version of the PSQ is no longer available for use as it cannot be located. Instead, a new provocation questionnaire was created to assess the experience and response of various incidences of provocation (i.e., verbal, gestures, and physical). The Sport Provocation Questionnaire was constructed using the original PSQ by Maxwell and Moores (2006) as a foundation, in addition to information collected from current and former athletes about their experiences. The author’s own experiences and observations of athletic competition also factored into the construction of the questionnaire. Specifically, the items assess the AGGRESSIVENESS, ANGER, AND PROVOCATION IN SPORT 15 frequency of experienced provocation, the frequency of one’s response to provocation, the level of anger felt by provocation, and the intensity of one’s response to being provoked. For each item, scores range from 1 (lowest) to 5 (highest). Higher scores represent greater frequency of provocation and response to provocation, as well as more intense anger felt by provocation. The intensity of provocation is assessed as ordinal data and will be analyzed separately from the other categories of the questionnaire. An example of an item assessing frequency of provocation is: “In competitive sports, how likely do you experience the following types of verbal provocation? a) the use of curse words or verbal abuse; b) the use of racial/ethnic slurs; c) the use of violent threats”. An example of an item assessing anger level from provocation is: “In competitive sports, what is your level anger when an opponent: a) aggressively or inappropriately touches you; b) purposefully shoves or trips you; c) punches or kicks you; d) purposefully strikes you with an instrument (like a helmet or stick)”. (See Appendix G). Pilot study. After obtaining IRB approval, the sport provocation questionnaire was piloted with a sample of twenty-two former high school and collegiate athletes. The mean age for all participants was 25.23 (SD = 4.33); 7 participants were male. Nearly half (45.5%) of all pilot participants played at the NCAA collegiate level, while the remaining played high school sports. Just over two-thirds of participants (68.2%) played contact sports, while the second biggest group (18.2%) played collision sports. The majority (63.4%) of participants played soccer, while the rest played a variety of contact or collision team sports such as basketball, hockey, or football. All participants filled out the questionnaire based on their prior sport experiences involving provocation. Open-ended feedback obtained from the pilot research was used to AGGRESSIVENESS, ANGER, AND PROVOCATION IN SPORT 16 improve the content and structure of the questionnaire and can be found in Appendix H. The four provocation scales were found to have adequate internal reliability, and the scales were moderate-highly correlated to each other (Field, 2009). These values suggest the scales are related to each other but still assessing different concepts in the experience of provocation. This information, in addition to the descriptive and correlational data obtained from the pilot study, can also be found in the tables in Appendix H. At this time, no standardized assessment of provocation in sport exists. With adequate internal reliability and sufficient face validity, it is expected that this new questionnaire will help build the foundation for future research to explore this important phenomenon. Procedure IRB approval was obtained before data collection began. Both assistant and head coaches from selected sport teams were contacted via email to explain the study and request participation from the athletes on their teams. As a small incentive, coaches were informed that participation in the study would grant them access to a general summary of the study findings once the researcher has compiled and analyzed all data. If they chose to participate, coaches were asked to forward the Qualtrics survey link to all of the athletes on their team. After the initial introductory email, all assistant and head coaches were contacted every two weeks with follow-up emails requesting their athletes participate in the study. Towards the end of the data collection period, a small number of other university athletic department personnel (e.g., athletic director or sport psychology consultant) were contacted at some schools to try to increase the sample size and generate adequate statistical power. All participants who opened the survey link first saw an informative cover letter detailing the purpose of the study and an explanation of participant rights (i.e., confidentiality, right to AGGRESSIVENESS, ANGER, AND PROVOCATION IN SPORT 17 withdraw at any time). There was also an option to print a copy of the cover letter for personal records. If participants agreed to continue, they were then asked to enter basic demographic information. Participants were then presented with online versions of both the CAAS and the sport provocation questionnaire. Successful completion of the demographic information and two surveys was predicted to take approximately fifteen minutes. Finally, the participants received the primary investigator’s contact information so that participants could communicate with the researcher about any issues related to the collection procedures, their data, or any other study related concerns. Data collection began during the summer off-season period but continued into the competitive Fall 2017 season. The off-season was decided as the best time to sample since most athletes were not as busy with school and training obligations. Collecting data at this time may also have helped avoid any bias due to abnormally high or low frustration or anger being experienced by athletes in the middle of the competitive season (for example, after a particularly disappointing or successful game or overall season). Data Analysis A G*Power 3.1 analysis (Faul, Erdfelder, Lang, & Buchner, 2007) revealed approximately 120 total participants would be the minimum sample size required to see a significant gender effect at the 95% level. After an adequate sample was collected, data analysis included both descriptive (e.g., frequencies, correlations, measures of central tendency, and standard deviations) and inferential statistical tests (e.g., internal reliability analyses and ANOVAs) to investigate the study hypotheses. All collected data was cleaned in Microsoft Excel and entered into the IBM Statistical Package for Social Sciences version 24 (SPSS, 2016). AGGRESSIVENESS, ANGER, AND PROVOCATION IN SPORT 18 The independent variables in this study included gender and type of sport. These were assessed on the demographic page. There are three dependent variables in the study: 1) competitive anger, 2) aggressiveness, and 3) provocation. These variables were assessed using the CAAS and sport provocation online questionnaires. Frequencies, means, standard deviations, and other descriptive data were calculated for demographic information such as age, gender, race/ethnicity, school, division, and type of sport. Overall means and standard deviations were calculated for all CAAS and provocation subscales. Standard bivariate correlations were also calculated among gender, type of sport, and the competitive anger, aggression, and provocation data. There were two primary research questions in this study. The first was: “How does competitive anger, aggressiveness, and experience of provocation vary among male and female athletes?” The second primary research question was: “How does competitive anger, aggressiveness, and provocation vary among contact and collision sport athletes?” These primary research questions were examined using a general linear model (GLM) to run multiple one-way analysis of variance (ANOVAs). To investigate the possibility of an interaction between the independent variables of gender and type of sport, two-way ANOVAs was used. Two-way ANOVAs helped answer the question of how scores on the dependent variables of CAAS or sport provocation questionnaire scores differed by gender and type of sport. For correlations, anger and aggressiveness subscale scores from the CAAS and provocation subscale scores were run together in a correlational matrix to examine the strength of the relationship between the variables. This was completed for the male and female data, as well as type of sport, to compare correlation strengths among the different variables. AGGRESSIVENESS, ANGER, AND PROVOCATION IN SPORT 19 Additional ad hoc analyses were completed to assess other study variables such as NCAA Division level or year in school. For example, differences in anger, aggressiveness, and experience of provocation were examined between NCAA Division I and Division III athletes. No current research suggests there are anger or aggressiveness differences among NCAA Division levels, but it is possible that at higher competitive levels, more competitive anger and aggressive behavior is produced due to the rising stakes, pressure, and fanfare. AGGRESSIVENESS, ANGER, AND PROVOCATION IN SPORT 20 Results Demographics and Descriptive Statistics A total of 243 NCAA athletes from 18 universities participated in the current study. From the final data aggregate, any participants who completed less than 40% of the survey questionnaire items were excluded from the final data analyses. The mean participant age was 19.53 (SD = 1.36). Participants were composed of 124 males and 119 females. At the time of data collection, they were playing one of ten different male or female team sports that included basketball, soccer, lacrosse, hockey, football, and rugby. Approximately 38% of participants were freshmen, with 20% sophomores, 21% juniors, and 15% seniors. In addition, 5% of participants identified as fifth-year seniors or graduate students. Over one-fourth (26%) of participants were enrolled in NCAA Division I universities, with 29% attending Division II schools and the remaining 44% attending Division III schools. Just over half (53%) of all participants played contact sports, while the rest (47%) were involved in collision sports. This demographic information is summarized in Table A1. Subsequent to data collection, Cronbach’s alpha analyses of internal consistency were calculated for the four provocation and CAAS scales (see Table B1). Review of these analyses revealed that all four provocation scales had adequate internal consistency, and were moderate highly correlated to each other (Field, 2009). These Cronbach’s alphas coefficients provided evidence to support the interpretation that the scales are appropriately related to each other but addressing different parts of the experience of provocation in collegiate sport. Internal reliability analysis was also utilized for the six-item Anger and Aggressiveness subscales of the CAAS, which were found to have moderate-high correlations. AGGRESSIVENESS, ANGER, AND PROVOCATION IN SPORT 21 Data Cleaning and Assumption Testing The following adjustments were made on the data in SPSS to ensure the ensuing bivariate analyses were accurate. The first change was to code an additional “6” Likert-type response (i.e., N/A or have not experienced) as a discrete missing variable. This ensured that these responses were not included in the standard scaled data for the provocation items. The second adjustment was to multiply all of the CAAS scores by the proper mean intensity of each item (following the procedures outlined by Maxwell & Moores, 2007). The authors conducted this step during the development of the CAAS because individual items on the two subscales are not equally weighted, with some impacting the anger or aggressiveness score more than other items. Finally, a total score variable was created for both the CAAS and SPQ instruments. The CAAS total score was created by summing the six-item Anger and Aggressiveness subscales. The provocation total score was calculated as the average of the four provocation scales, using data from participants who filled out at least three out of four mean responses. This decision was made to ensure the anger level provocation subscale data was included despite having to code for the “6” missing data choice. Overall, this resulted in a total of 18 participants being removed from the final database before final analyses were conducted. Before conducting inferential statistical analyses, the data were examined using SPSS to ensure that it met the appropriate assumptions needed for valid two-way ANOVA results. The assumptions of a continuous dependent variable, independence of observations, and independent variables with categorical groups were all satisfied based on random sampling and the type of variables (e.g., continuous, ordinal) utilized in the study. Using box and whisker plots (Field, 2009), only one outlier was identified from the CAAS Anger subscale data and five from the CAAS Aggressiveness subscale data. These were not considered extreme outliers, (i.e., over AGGRESSIVENESS, ANGER, AND PROVOCATION IN SPORT 22 three times the Interquartile Range) so they were retained in the database. Up to three outliers were identified for the four different provocation subscales, but these were also included in the final data aggregate since they did not constitute extreme outliers that would likely have a significant detrimental effect on the analyses. The normality of the data was examined using kurtosis/skewness values as well as histograms and Q-Q plots. Overall, the data for nearly all of the subscales followed a normal distribution, with only slight departures from normality found on the CAAS Aggressiveness and negative response to provocation subscale data. These departures consisted of slight floor effects, meaning the subscales had a clear lower limit of possible participant’s responses. This caused a larger than usual number of scores to congregate near this limit. However, ANOVA is known to be particularly robust to violations of normality, so the analyses were carried out despite these aforementioned slight departures. Levene’s test, as well as assessing the data spread vs. Q-Q plots, was utilized to test the homogeneity of variances (Field, 2009). While the CAAS aggressiveness and negative response to provocation subscales varied more than expected in normally distributed data, the standard deviation spread was not large compared to the mean differences. This suggested that running the ANOVA using the data would not be overly problematic, and it would be unnecessary to run non-parametric tests of ANOVA. Bivariate Statistics The first primary research question in this study was: How does competitive anger, aggressiveness, and the experience of provocation vary among male and female athletes? To answer this question, a one-way between subjects ANOVA was conducted to compare scores on the CAAS among male and female athletes (see Table C1). For all ANOVAs, SPSS was used to calculate effect size, represented as partial eta square (η2). To compare magnitude of effect AGGRESSIVENESS, ANGER, AND PROVOCATION IN SPORT 23 sizes, rule of thumb values (small = .01, medium = .06, and large = 0.14) were used based on recommendations set forth by Vacha-Haase and Thompson (2004). There was a statistically significant effect of total CAAS score among gender [F(1, 237) = 23.68, p < 0.001, η2 = .091], with males (M = 55.79, SD = 19.41) scoring significantly higher than females (M = 45.52, SD = 12.12). Additional one-way between subjects ANOVAs were conducted to compare scores on the CAAS Anger and Aggressiveness subscales by gender. There was no statistically significant effect of CAAS Anger scores among gender [F(1, 241) = 1.35, p = 0.247, η2 = .006], although males (M = 26.43, SD = 8.59) scored non-significantly higher than females (M = 25.23, SD = 7.47) (see Table C2). There was, however, a significant effect for CAAS Aggressiveness subscale scores among gender [F(1, 237) = 46.27, p < 0.001, η2 = .163], with males (M = 29.36, SD = 12.60) scoring significantly higher than females (M = 20.29, SD = 7.02) (see Table C3). There was not a statistically significant effect of total SPQ score among gender [F(1, 217) = 2.49, p = 0.12, η2 = .011], although males (M = 2.38, SD = 0.68) scored slightly higher than females (M = 2.24, SD = 0.55) (see Table C4). However, when separated into the four provocation subscales, statistically significant differences were found among all four provocation subscales between male and female athletes. Specifically, males scored significantly higher than females on frequency of provocation experienced, frequency of negative response to provocation, and intensity of response to provocation. Females scored significantly higher than males on anger felt from provocation (see Table C5). The second primary research question in this study was: How does competitive anger, aggressiveness, and the experience of provocation vary among contact and collision sport athletes? To answer this question, a one-way between subjects ANOVA was conducted to AGGRESSIVENESS, ANGER, AND PROVOCATION IN SPORT 24 compare scores on the CAAS among type of sport (see Table C6). There was a statistically significant effect for total CAAS score among type of sport [F(1, 237) = 31.17, p < 0.001, η2 = .116], with collision sport athletes (M = 56.98, SD = 18.85) scoring significantly higher than contact sport athletes (M = 45.35, SD = 13.13). Additional one-way between subjects ANOVAs were conducted to compare scores on the CAAS Anger and Aggressiveness subscales by type of sport. There was a statistically significant effect for CAAS Anger subscale scores among type of sport [F(1, 241) = 4.04, p = .046, η2 = .016], with collision sport athletes (M = 26.95, SD = 8.35) scoring significantly higher than contact sport athletes (M = 24.88, SD = 7.72) (see Table C7). There was also a significant effect for CAAS Aggressiveness subscale scores among type of sport [F(1, 237) = 52.25, p < 0.001, η2 = .181], with collision sport athletes (M = 30.03, SD = 12.30) scoring significantly higher than contact sport athletes (M = 20.48, SD = 7.84) (see Table C8). There was also a statistically significant effect for total SPQ score among type of sport [F(1, 217) = 7.39, p = .007, η2 = .033], with collision sport athletes (M = 2.43, SD = 0.65) scoring significantly higher than contact sport athletes (M = 2.21, SD = 0.57) (see Table C9). To better understand what aspects of provocation were significantly different, one-way ANOVAs were run for each of the four subscales of the SPQ. The results of these analyses can be found in Table C10. Specifically, collision sport athletes scored significantly higher than contact sport athletes on the frequency of provocation experienced and frequency of negative response to provocation. A two-way ANOVA was conducted to examine the effect of gender and type of sport on total CAAS scores (see Table C11). The results of the two-way ANOVA provided evidence to indicate there was no statistically significant interaction effect between gender and type of sport. AGGRESSIVENESS, ANGER, AND PROVOCATION IN SPORT 25 This suggests that any total CAAS score differences between contact and collision sport athletes were not dependent upon the gender identified with by the participants and that any total CAAS score differences between females and males were not dependent upon which type of sport they played. A separate two-way ANOVA was conducted to examine the effect of gender and type of sport on total SPQ scores (see Table C12). The results of the two-way ANOVA provided evidence to support the notion that there was no significant interaction effect between gender and type of sport. This indicates that any total SPQ score differences between contact and collision sport athletes were not dependent on the gender identified with by the participants, and any total SPQ score differences between male and female athletes were not dependent upon which type of sport they played. Statistical analyses regarding the division level of participants were not part of the researcher’s original research questions. However, since data was collected from athletes in all three NCAA division levels, additional statistical analyses were conducted to explore possible differences among the dependent variables. There was a statistically significant effect of total CAAS score among NCAA Division level [F(2, 236) = 4.96, p = .008, η2 = .040] (see Table D1). Post hoc comparisons using the Tukey HSD test provided evidence to suggest that the mean total CAAS score for Division I athletes (M = 54.74, SD = 19.21) was significantly higher than Division III athletes (M = 47.11, SD = 14.36), but not Division II (M = 53.13, SD = 17.94) athletes (see Table D2). One-way, between subjects ANOVAs were conducted to compare scores on the CAAS Anger and Aggressiveness subscales among NCAA Division level. There was a statistically significant effect of CAAS Anger subscale scores among NCAA Division level [F(2, 240) = AGGRESSIVENESS, ANGER, AND PROVOCATION IN SPORT 26 5.48, p = .005, η2 = .044] (see Table D3). Post hoc comparisons were conducted using Tukey’s HSD test (Field, 2009). The results of this analysis provided support for the interpretation that the mean CAAS Anger subscale score for Division I athletes (M = 27.68, SD = 8.66) was significantly higher than Division III athletes (M = 23.98, SD = 6.94), but not statistically different from Division II (M = 27.01, SD = 8.60) athletes. Division II athletes were also found to score significantly higher than Division III athletes on CAAS Anger subscale scores (see Table D4). There was also a statistically significant effect for CAAS Aggressiveness subscale scores among NCAA Division level [F(2, 236) = 3.12, p = .046, η2 = .026] (see Table D5). Post hoc comparisons using Fisher’s LSD test (Field, 2009) indicated that Division I athletes (M = 27.16, SD = 12.50) scored significantly higher than Division III athletes (M = 23.05, SD = 9.79), but not Division II (M = 26.05, SD = 11.80) athletes (see Table D6). There were no statistically significant differences in total SPQ scores between NCAA division levels [F(1, 216) = .574, p = .564, η2 = .005] (see Table D7). One-way between subjects ANOVAs were conducted for all four provocation subscales to look for differences by NCAA division level. As displayed in Table D8, no significant differences were found by division level among any of the four provocation subscales. AGGRESSIVENESS, ANGER, AND PROVOCATION IN SPORT 27 Discussion The primary goal of this study was to investigate and better understand competitive aggressiveness, anger, and the experience of provocation in collegiate sport athletes. In particular, it was important to understand if these variables differed among male and female contact and collision sport athletes, and if any further differences occurred among NCAA division level. Both male and female collegiate athletes participating in contact and collision sports were surveyed electronically using questionnaires that assessed the variables of competitive aggressiveness and anger, and the experience of provocation. The two hypotheses proposed at the beginning of the study were that: 1) male athletes would score higher on the CAAS anger and aggressiveness subscales, and would report experiencing (and responding negatively to) more provocation than female athletes; and 2) collision sport athletes would score higher on the CAAS anger and aggressiveness subscales, and would report experiencing more provocation and more frequently responding to provocation than contact athletes. Competitive Aggressiveness and Anger Gender comparison. Based on the results of the current study, the first hypothesis was partially supported. Male collegiate athletes scored significantly higher overall on the CAAS than female athletes, and they were significantly more likely to tolerate aggressiveness and be inclined to aggress in an athletic setting. However, while male athletes scored slightly higher on competitive anger, this difference was not statistically significant. These findings point to the notion that the amount of anger felt during competitive sports was not significantly different based on gender. Some researchers have suggested that males are more prone to anger than females (e.g., Maxwell & Moores, 2007; Maxwell, Visek, & Moores, 2009). However, the results of the current study suggest that the female athletes in these types of sports may simply be AGGRESSIVENESS, ANGER, AND PROVOCATION IN SPORT 28 less likely to act on anger than male athletes. For example, female athletes were found to have significantly higher anger levels from provocation by opponents, but males had more negative and intense responses to the same types of provocative behavior. This finding seems to be in disagreement with some previous research on high school tennis athletes who reported that male and females tended to cope with anger in similar ways (Bolgar, Janelle, & Giacobbi, 2008). The notion that male athletes are, in general, more inclined to aggress than female athletes has been suggested in previous literature (e.g., Bredemeier, 1978; Coulomb-Cabagno & Rascle, 2005; Maxwell, 2004). Even at the middle school level, some male athletes displayed significantly more aggression off the field than female athletes, with participation in sport being a risk factor for antisocial behavior (Burton & Marshall, 2005). Further, during the development of the CAAS questionnaire, the authors found that male CAAS scores were higher than female scores on both the Anger and Aggressiveness subscales (Maxwell & Moores, 2007). This result was consistent for both contact and non-contact sport athletes. On the other hand, not all researchers have found evidence to support the classic aggressive male athlete stereotype. For example, Keeler (2007) found that males and females did not differ in hostile or instrumental sport aggression, and this finding was consistent among non-contact, contact, and collision sport athletes. However, males did score higher on questions assessing life assertion and assault aggression. Overall, the finding that male athletes (regardless of sport type) displayed more competitive aggressiveness is in agreement with most of the previous research on gender differences in aggressive behavior. The results of the current study parallel a number of findings from previous research studies (Burton & Marshall, 2005; Coulomb-Cabagno & Rascle, 2006; Gardner & Janelle, 2002). One example is males reporting that they were more inclined to aggress than female athletes, AGGRESSIVENESS, ANGER, AND PROVOCATION IN SPORT 29 regardless of what type of sport they played. However, unlike research conducted in the creation of the CAAS (Maxwell & Moores, 2007), the current findings support the idea that male and female college athletes are not significantly different in the amount of anger experienced in competitive environments. It should be noted that Maxwell and Moores (2007) also utilized participants from non-contact sports. Therefore, the current study is unique in that it is the first known study to explore competitive aggressiveness and anger differences between only contact and collision sport athletes. It is important to differentiate these types of sports from non-contact ones, since contact and collision sport athletes compete as members of teams, in environments that produce consistent physical contact with opponents. The finding that males participating in contact and collision sports reported higher competitive aggressiveness could have important implications for educational programs, such as ones that focus on anger management skills. Aggression has been shown numerous times to be a possible byproduct of anger, especially when anger is unable to be controlled (e.g., Feindler & Ecton, 1994; Maxwell, Visek, & Moores, 2009; Robazza et al., 2006). While the male and female athletes in the current study experienced similar levels of anger during competition, males said that they would be more inclined to justify aggressive behavior as a means of dealing with that anger. It appears that male athletes might benefit more from resources intended to teach athletes other ways to deal with competitive anger, especially under the stressful conditions seen in sport competition. Sport type comparison. The second hypothesis was supported based on the findings of the current study. Collision sport athletes were found to have significantly higher competitive aggressiveness and anger than contact sport athletes. This finding was consistent irrespective of gender. What seems to be one of the clearest predictors of competitive aggressiveness and anger AGGRESSIVENESS, ANGER, AND PROVOCATION IN SPORT 30 is the type of sport an athlete plays. Although all sports involved in the current study featured contact between opponents, the key difference between collision and contact sports is that the latter allows in-game contact, but extreme contact or direct collisions are not implicit (or required) by the written rules of the sport (Keeler, 2007; Silva, 1983). In contrast, collision sports generally involve high-speed collisions as an integral part of the sport. Simply put, they are necessary to achieve appropriate goals needed for success. A few previous studies have produced findings that are relevant to the sport type differences found in the current study. In part