The Differences in Sport Aggression, Life Aggression, andLife Assertion Among Adult Male and Female Collision,

Contact, and Non-Contact Sport Athletes

LindaA.KeeierEast Tennessee State University

A relationship between contact level in sport and aggression has been found in previousliterature (Bredemeier, Weiss, Shields, & Copper, 1986; Silva, 1983; Tucker & Parks, 2001).Presently, a study has yet to compare adult aggression scores for both men and women in thesame contact sport types. This study was intended to compare both female and male colli-sion, contact, and non-contact sport athletes across sport hostile aggression, sport instru-mental aggression, life aggression, and life assertion. Results indicated that life aggression,life assertion, sport hostile aggression, and sport instrumental aggression did not varyamong different contact sport levels. Significant gender differences were found in life asser-tion and two subscales of life aggression. Relationships between aggression and assertionvariables were explored.

Address Correspondence To\ Linda A. Keeler, Box 70654, Johnson City, TN 37614, Phone:(423) 439-4382, E-mail: keelei@etsu.edu.

57

58/Journal of Sport Behavior, Vol. 30, No. I

Aggression is a negative personality trait that has been associated with sport participa-tion. Aggression is operationally defined as an intentional physically or psychologically harm-ful behavior that is directed at another living organism (Thirer, 1993). The frequency of aggres-sion in sports on all levels has led to a great deal of academic research. Even the InternationalSociety of Sport Psychology recently recognized that sport aggression has become a socialproblem both on and off the playing field and has recommended ways to curtail this behavior(Tenenbaum, Stewart, Singer, & Duda, 1997). The particulars of this position have recentlybeen challenged (Kerr, 1999), reafiFirmed (Tenenbaum, Sacks, Miller, Golden, & Doolin, 2000),and clarified (Kerr, 2002).

Aggression has been studied on many levels including behavior within the context ofsport and within daily life. A popular theory explaining aggression is the Revised Frustration-Aggression Theory (Berkowitz, 1965). This theory consists of aspects from Bandura's (1973)Social Learning Theory and Smith's (1972) theory on frustration and aggression. Berkowitzproposed that either frustration or another stimulus (e.g., threat) increases a person's arousaland anger levels, which increases one's readiness to aggress. However, aggression will onlyoccur if the person has learned the appropriateness of such behavior in that specific situation.In other words, aggression not only depends on the strength of the association between thesituation and aggressive behavior, but also the degree of readiness to aggress and the pres-ence of aggressive cues (Berkowitz). This easily generalizes to the sport socialization process.During an athletic contest, the potential for a frustrating situation is unlimited. Combine thatwith aggressive behavior that is rewarded by teammates, coaches, and parents, or vicariouslylearned from role models on television or during live contests, and the potential for aggressionin sport rises exponentially.

Two types of aggression have been defined in sport research, hostile and instrumental.Hostile, or reactive aggression is behavior performed with the sole intention of infiicting harmon a person (Silva, 1983). Instrumental aggression in sport is behavior that intentionally causesinjury or harm to an opponent in pursuit of another non-aggressive goal such as scoring orwinning (Bredemeier, 1975). Assertiveness is distinct from aggressiveness in that it is thenonhostile, noncoercive tendency to behave with intense and energetic behavior to accom-plish one's goal (Bredemeier, 1994; Silva, 1978). In the sport realm, tiiese types of behavior areoften within the rules of competition. It is hard to distinguish the relationship between aggres-sion and assertion because they have often been conceptually confused in the literature(Silva, 1978), and can usually only be differentiated by a person's intention, which remainsdependent on self-report. However, researchers have utilized various measurements to assessathlete aggression (AUawy, 1981; Bredemeier, 1994; Reid & Hay, 1979; Ryan, Williams, &Wimer, 1990; Silva, 1983; Wall &Gruber, 1986).

Differences in Sport Aggression.. . / 59

Recent trends in research link patterns of sport aggression to goal orientations (Duda etal., 1991;Dunn&Dumi, 1999), moralreasoning(Bredemeier& Shields, 1986;Bredemeier, 1994;Rosenberg, 2003), and gender (Silva, 1983). Aggression has also been examined in reference todifferent sport types (e.g., contact vs. non-contact and individual vs. team; Allawy, 1981;Bredemeier& Shields, 1986; Colley, Roberts, & Chipps, 1985; Mace & Baines, 1989; Silva,1983). It has been shown that sports with contact have positive associations with the amountof aggressiveness of their participants (Allawy; Bredemeier, 1994; Gardner & Janelle 2002;Reid&Hay, 1979; Silva, 1983; Tucker& Parks, 2001). Silva has differentiated contact sportsinto three distinct levels: collision (contact is necessary and integral to play), contact (contactis legal and occurs incidentally), and non-contact (contact between opponents is not allowed).Silva studied the relationships between the contact level of sport and legitimacy ratings of ruleviolations (i.e., aggressive behavior) in sport across women and men. Differences were foundbetween amount of contact and years of experience in legitimacy ratings for men and women.In a similar study. Tucker and Parks found that athletes in collision sports scored higher onlegitimacy ratings than those in contact and non contact sports. Further, a significant interac-tion effect was found in this study with greater gender differences in non contact sports thanin collision and contact sports with women scoring lower than men. More recently, Gardnerand Janelle did not fmd differences in legitimacy ratings across low and high contact, collegesport athletes, but did fmd males to perceive aggressive and assertive behavior as morelegitimate than females. In addition, behavior in sport situations was overall perceived as morelegitimate than in life situations.

A limitation in the research on aggression is that studies in collision sports may begender biased. Most collision sports have traditionally only been available for men, excludingwomen from participation and therefore from research. Further, many traditional collision sportsfor men have modified rules to disallow certain types of contact in the women's version (e.g.ice hockey, lacrosse). This has created a gap in the literature regarding aggression and contactsport type utilizing male and female athletes from the same sports. The relationship betweenaggression and female collision sport athletes may be different from the relationship for maleathletes. The Revised Frustration-Aggression Theory would predict that through participa-tion in sport, females would leam to be as aggressive as their male counterparts because of theopportunity for observational learning and the presence of situational cues and reinforce-ments. More specifically, if women played the same collision sports as men they would showsimilar aggression levels because of the similar sport socialization processes. In fact, givensimilar situations, women have shown to be just as aggressive as men (White & Kowalski,1994). However, women and men have been shown to differ on their acceptance of sportaggression in relation to sport type participation (Bredemeier, 1994; Silva, 1983; Gardner &JaneUe,2002).

60 /Journal of Sport Behavior, Vol. 30, No. 1

The reason for aggression differences remains vague. It is unclear whether the differ-ences found across gender and sport types are due to different sport socialization processesor from an existing disparity in those who gravitate towards certain sports (Morgan, 1980).Exposure to contact in sports has been found to be previously related to men's traditionalideals of masculinity and negative attitudes towards women (Maier & Lavrakas, 1981), how-ever these trends may be changing (Smith & Stewart, 2003). On the other hand, women partici-pating in perceived low-feminine sports (i.e., collision and contact) held more liberal gender-role attitudes than high-feminine sport participants (Salisbury & Passer, 1982). Perhaps thegravitation of a woman towards a non-traditionally feminine sport reflects an upbringing thatencouraged individuality and non-conformity. Less stereotypical-feminine athletes may bedrawn to collision sports because ofthe innate non-feminine, traditionally masculine qualitiesof contact activities and the emancipation that participation in such sport brings to them. Infact, there are supporters ofthe notion that participation in contact sport maybe beneficial forwomen. It is believed that a woman uses her body in contact as a means to express and learnabout herself (Rail, 1992) and can actually help empower girls and women by teaching themabout their physical capabilities (Theberge, 2003). This could mean that women may experi-ence a different collision and contact sport socialization process than men. It is necessary toexamine sport aggression in women across sport groups to conclude ifthe differences alreadyfound between male athletes apply to female athletes. Once the relationship has been exam-ined across sport types and between genders on sport aggression, it is also important toexamine how the findings relate to personality traits foxmd in daily life. Research in this area isnon-conclusive and some researchers have suggested that behavior exhibited in sportingactivities may not refiect everyday life personality behavior and traits (Bredemeier, 1994;Bredemeieretal., 1986; Smith & Stewart; 2003 ;Thirer, 1993).

The direct comparison of sport aggression variables (i.e., hostile and instrumental) be-tween the same male and female collision, contact, and non-contact sports has not yet beenexplored. Comparisons of perceived legitimacy ratings in Silva (1983), Tucker and Parks (2001),and Gardner and Janelle (2002), integrated various sports per contact level, including bothteam and individual sports; a variable that may have confounded results. Further, these inves-tigations only utilized one type of women's collision sport out of five in that category for menand categorized wrestling as a contact sport rather than a collision sport. The present studywas conducted to investigate any differences between different domains of aggression andthe same interactive team sports for men and women across three levels of contact. Therelationship between sport type and daily life aggression was included since sport and dailylife variables have been shown to differ. Furthermore, since assertion is closely related concep-tually to aggression and is the socially desirable alternative behavior to aggression, its rela-

Differences in Sport Aggression.. . / 6 I

tionship to sport types was relevant to this investigation. Once differences in the personalityvariables across contact sport types and gender were established, the relationships betweenthe dependent measures were examined. The goal of this study was to investigate if adultwomen and men have the same pattern of self-perceived sport aggression, life aggression, andlife assertion across collision, contact, and non-contact sport types. In addition, the relation-ships between sport hostile aggression, sport instrumental aggression, life aggression, andlife assertion variables were explored. These comparisons were examined while controlling fordifferences in age, education, team affiliation, years of experience, and individual success.

Method

Participants and SamplingOne hundred and sixty-one club sport athletes {N= 92 females; N^69 males) from the

Mid-Atlantic region participated in this study. Participants ranged in age from 18 to 43 years(M= 28.76 females; M= 27.67 males) and ranged in sport experience from less than one year tothirty years (M= 10.05 females; M= 9.27 males). Participants were grouped according to theirsport type: collision, contact, non-contact, or a combination of sports. The sports chosen torepresent each sport type were rugby for collision (n = 59 females; n = 44 males), soccer forcontact (« = 19 females; « = 9 males), and volleyball for non-contact (n = 14 females; « = 16males). As an attempt to keep competitive levels similar, only competitive teams associatedwith organized leagues under a national governing body were approached. Further, only inter-active team sports were used to guard against the nature of sport effect previously foundbetween team and individual sport groups (Bredemeier & Shields, 1986). In terms of higherlevels of competition, individual success was categorized into none, local, regional, and na-tional levels. Participants were asked the highest level of education they had completed orwere currently seeking. Education was categorized into high school/GED/associate degree,bachelor degree, and masters' degree/doctorate degree/professional groups. The frequenciesand percents of individual success and education are presented in Table 1.

Data Collection ProcedureRepresentatives from each team were contacted by phone or email and were asked if

their team would be willing to participate in research on sport attitudes. An attempt was madeto collect all data before practice sessions (« = 27). However, due to the preference of teamrepresentatives, to ensure a larger attendance of players, and lack of other alternatives, ques-tionnaires were also administered after practice (n = 92), before games (n = 21), via ground mail(n = 8), and via electronic mail (n = 13). Those surveys given on game days were given at least

62 /Journal of Sport Behavior, Vol. 30, No. 1

Table 1. Frequencies and Percents of Athlete's Individual Success and Education by Gender(N = 161)

Descriptor

SuccessNoneLocalRegionalNational

EducationH.S./GED/A.A.B.A./B.S.M.A./PhD/Prof

Femalen

486299

75728

(n = 92)%

52.26.531.59.8

7.662.030.4

Male (nn

276305

74120

= 69)%

39.18.743.512

10.159.429

one half hour before competition. Since the scales used were trait rather than state measures,it was believed that overall results would not be significantly affected by varying collectiontimes. Regardless, a preliminary analysis was conducted to ascertain any differences in re-sponses across collection times. One logistical hardship of using adult competitive leagueswas that some teams did not practice and/or did not have formal warm-up periods prior togames. Men's soccer, in particular, fell under this category, which resulted in a small samplesize for this group.

InstrumentationLife aggression. The Buss-Durkee Hostility Inventory (BDHI) assesses global hostility

through seven individual sub-scales and one scale measuring guilt (Buss & Durkee, 1957).Participants respond either true or false to statements to indicate whether the individual actionor thought described is self-descriptive. Preliminary concurrent validity has been shown forthe items in the scale (Buss, 1961). Three of the sub-scales that loaded on a "motor" factor ofaggression (Buss & Durkee) were used in this study to comprise a composite measurement oflife aggression: assault (10 items), indirect hostility (9 items), and verbal hostility (13 items).During the initial construction of the BDHI, the highest correlation found between the three

Differences in Sport Aggression... / 63

sub-scales was .40 indicating that the scales measure relatively independent components(Buss & Durkee). These individual sub-scales have been used in many studies (Bredemeier,1975; Buss, 1961; Wann, Fahl, Erdmann, & Littleton, 1999). Atest/retest with a five-weekseparation showed moderate stability with the assault (r = .78), indirect (r = .72), and verbalhostility (r = .72) subscales (Buss, 1961).

Life assertion. The Rathus Assertiveness Schedule (RAS) is a 30-item inventory thatmeasures perceived assertiveness (Rathus, 1973). It measures assertive behavior along a six-point scale ranging from 3 = "very characteristic of me, extremely descriptive" to -3 = "veryuncharacteristic of me, extremely non-descriptive." One item was amended in the presentstudy to prevent sexual orientation bias (item 11 was changed from "I often don't know whatto say to attractive members of the opposite sex" to "I often don't know what to say toattractive persons"). Evidence for test-retest reliability (r = .78), split-half reliability (r = .77),and concurrent validity (r = .77) has been provided (Rathus).

Sport aggression. The short form, the B AAGI-S (Bredemeier, 1975), has 15 hostile and 15instrumental items with the highest factor loadings, representing the elements of anger, hostil-ity, and frustration (Stephens, 1998; Wall & Gruber, 1986). Items are answered on a 4-pointLikert scale ranging from 1 = "strong agreement" to 4 = "strong disagreement." Lower scoresrepresent higher levels of aggression for each subscale. However, during the present investi-gation, scores were reversed so that higher scores would indicate higher levels of aggression.Bredemeier's testing of the original scale on female athletes showed high alpha reliabilitycoefficients for both hostile (.90) and instrumental (.86) subscales. The two subscales werehighly and negatively correlated (-.69). Evidence was provided for concurrent and predictivevalidity (Bredemeier) and intra-class coefficients were significant (Wall & Gruber). Significantcorrelations with the Crown Marlowe Social Desirability Scale suggested that hostile aggres-sion was socially undesirable and instrumental aggression was more socially desirable(Bredemeier).

Demographic information. Backgroimd information including age, primary sport, yearsof experience, education, individual success (i.e., highest level achieved), and perceived con-tact level of sport was collected in the questionnaire packet. This information was collectedbecause these variables have been shown to have confounding effects in previous aggres-sion research (e.g., Bredemeier, 1994; Howe, 1973; Silva, 1983; Widemeyer, 1984).Statistical Analysis

Four general linear models were applied to the data to study the main effects of genderand primary sport type and their interaction on the dependent measures of life aggression, lifeassertion, hostile sport aggression, and instrumental sport aggression. The analyses wereconducted while controlling for age, years of experience, individual success, education level.

64 / Journal of Sport Behavior, Vol. 30, No. 1

and team dififerences. Years of experience, individual success, and education levels were en-tered into the model as fixed factors and age was entered as a covariate. The team effect wastreated as a random factor and as a nested variable within gender and primary sport type. Thisensured that the calculations of the F statistics for the main effects of gender and sport type,as well as their interaction, would utilize the sums of squares of the nested team factor ratherthan the overall residual sums of squares. Three additional general linear models were per-formed as further exploration into the independent variables effect on the life aggressioncomponents of assault, indirect, and verbal aggressions. Least Significance Difference posthoc analyses were employed to find the differences between levels of significant main effectsor interactions when applicable. Pearson correlation coefficients were calculated to examinethe relationships between life aggression, life assertion, and sport aggression variables.

Results

Data Collection ComparisonPost hoc analysis of the individual subscales was done on an exploratory basis. Data

analysis comparing the different data collection procedures suggested that there was littleeffect on the results due to differences in collection methodology. There was one differencefound on the life assault subscale between "before practice" collection (Af = 4.380) and "afterpractice" collection (M= 2.963) with the after practice participants scoring higher on assault.However, due to the number of comparisons executed in this analjrais, these results may havebeen due to a Type I error.

Gender and Sport Type DifferencesThe means, number of cases, and standard deviations for males' and females' sport

hostile aggression, sport instrumental aggression, life assertion, and life aggression are pre-sented in Table 2. A test for the homogeneity of variances was conducted with no significantviolations found. There were no significant sport type main effects for any of these dependentvariables, indicating that there were no differences among collision, contact, and non-contactsport groups in life aggression, hfe assertion, sport hostile aggression, and sport instrumentalaggression. A significant gender main effect was found, F(l , 35.863) = 4.959,;? < .05,7- = . 121,power = .582, showing men (M= 12.21, SD = 22.78) to have significantly higher assertionscores than women (M= 4.69, SD = 24.69). No other significant gender effects were found.Furthermore, a significant sport type by gender interaction was not observed for any of thedependent variables.

Differences in Sport Aggression... / 65

Table 2. Means, Number of Cases, and Standard Deviations of Sport and Life Variables byGender and Sport T^pe

Sport Type

FemaleCollision (« = 59)

MSD

Contact («= 19)MSD

Non-contact (n =MSD

Total (A =̂ 92)MSD

MaleCollision (M = 44)

MSD

Contact (« = 9)MSD

Sport hostileaggression

17.176.06

17.316.4714)19.576.58

17.576.21

22.586.26

13.443.43

Non-contact (« = 16)MSD

Total (Af= 68)MSD

18.915.11

20.486.47

Sport instrumentalaggression

24.994.71

25.173.02

25.045.24

25.034.46

24.735.91

29.673.16

24.816.19

25.415.87

Lifeassertion

7.0925.53

-2.7827.58

4.3614.80

4.6924.69

13.9122.63

21.0017.63

3.2523.84

122122.78

Lifeaggression

15.804.89

15.374.89

14.935.09

15.584.87

16.484.82

14.784.06

15.806.07

16.104.99

66 / Journal of Sport Behavior, Vol. 30, No. I

The means, number of cases and standard deviations for the three life aggression subscalesare presented in Table 3. Analyses did not show any differences across sport type for assault,indirect hostility, or verbal hostility, suggesting that these variables do not differ across colli-sion, contact, and non-contact sport athletes. Gender main effects were found for the assaultand indirect hostility scales. Men scored significantly higher on the assault scale F(l , 20.094)=4.435,;? < .05, ;7̂ = . 181, power=.518, while women scored significantly higher on the indirecthostility scale F(l , 63.033) = 4.792,p < .05, rf = .071, power = .578. No interactions betweengender and sport type were found in any ofthe three scales. Men (M= 4.63, SD = 2.38) scoredsignificantly higher on the assault scale than women (M= 3.18,5D = 2.22), while women (M=5.67, SD=l .86) scored significantly higher on the indirect hostility scale than men {M= 4.59,5£)=1.75).

Life and Sport RelationshipsCorrelations for the dependent variables of sport hostile aggression, sport instrumental

aggression, life aggression, and life assertion are presented in Table 4. Sport hostile aggres-sion and sport instrumental aggression scores were negatively related for both women {r = -.437, p < .01) and men (r = -.252,p < .05). The sport aggression variables were each significantlyrelated to total life aggression across gender. The correlations between total life aggressionand sport hostile aggression scores were significant for women (r=.460,/><.01) and men (r=.451,/? < .01). Total life and sport instrumental aggression scores were significantly negativelycorrelated for women (r=-.360,/? < .01) and men (r=-.256,p < .05). These results indicate thatathletes with high levels of life aggression are more likely to have high levels of sport hostileaggression and low levels of sport instrumental aggression. Life assertion was positivelyrelated to life aggression but was significant for women only (r = .295,p< .01). Life assertionwas not significantly correlated to sport hostile or instrumental aggressions for either womenor men. Although not a primary purpose of this study, more specific analyses ofthe correla-tions between sport aggression, life aggression, and life assertion variables with the life ag-gression subscales were conducted and reported in Table 5. The subscales themselves wereall significantly positively correlated to one another for men. However, only the assault andverbal hostility scales were significantly correlated for women.

Differences in Sport Aggression.. ./67

Table 3. Means, Number of Cases, and Standard Deviations of Life Aggression Subscales byGender and Sport Type

Sport Type

CollisionMSDn

ContactMSDn

Non-contactMSDn

TotalMSDn

Assaultaggression

3.1522359

3.681.8619

2.642.6214

3.1822292

FemaleIndirecthostility

6.021.7459

4.892.05

19

5.291.86

14

5.671.86

92

Verbalhostility

6.632.8559

6.7932419

7.002.0014

6.722.8092

Assaultaggression

5.052.1244

4.222.739

3.672.72

15

4.632.38

68

MaleIndirecthostility

4.551.7644

3.891.699

5.131.6815

4.591.7568

Veibalhostility

6.892.3244

6.671.739

7.002.5615

6.8822868

68 / Journal of Sport Behavior, Vol. 30, No. 1

Table 4. Pearson Correlation Coefficients of Sport and Life Variables by Gender

Measure

1. Sport hostile aggression

2. Sport instrumental aggression

3. Life aggression

4. Life assertion

Measure

1. Sport hostile aggression

2. Sport instrumental aggression

3. Life aggression

4. Life assertion

1

-.437**

-.460**

0.97

1

-.252*

-.451**

-.149

Female2

-

-.360**

.122

Male2

-

-.256**

.097

3 4

-

.295** -

3 i

-

.121

Differences in Sport Aggression... / 69

Table 5. Pearson Correlation Coefficients of Sport and Life Variables with Individual LifeAggression Subscales by Gender

Measure

Female

Sport hostile aggression

Sport instrumental aggression

Life assertion

Life aggression

MaleSport hostile aggression

Sport instrumental aggression

Life assertion

Life aggression

Assaultaggression

.376**

-.196

.252*

.754**

.473**

-.088

.184

.800**

Indirecthostility

.263*

-.370**

-.076

.534**

.434**

-.356**

-.250*

.698**

Verbalhostility

.328**

-.224*

.365**

.789**

.124

-.185

.260*

.817**

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

70 / Journal of Sport Behavior, Vol. 30, No. I

Discussion

The purpose of this study was to identify the differences between sport hostile aggres-sion, sport instrumental aggression, life aggression, and life assertion across collision, con-tact, and non-contact sport groups for both female and male adult athletes. By studying thedifferences of both sport aggression and life aggression across all sport types for both menand women using the same sports for each, the relationship between contact in sport andaggression can be more accurately reported. This accurate reporting could then be useful inany future attempts to predict and control aggressive behavior.

Sport Hostile and Instrumental AggressionThe results indicated that sport hostile and instrumental aggression did not differ across

sport type or gender. The similarities in reported hostile aggression across gender and be-tween men's sport groups are contrary to previous research (e.g., Bredemeier et al., 1986;Kemler, 1988; Reid & Hay, 1979;Silva, 1983; Tucker&Parks, 2001). In previous work, perceivedlegitimacy of rule violating behavior in college athletes (Silva; Tucker & Parks) and perceivedphysical aggression in children (Bredemeier et al.), increased as a function of contact level forboth male and female athletes. These results support Gardner and Janelle's (2002) findings thatperceived legitimacy ratings of aggressive behavior in sport are not a function ofthe level ofcontact in sport. They suggested the differences in aggression levels across contact levelsmay not be present in higher levels of sport. The present analysis held personal level of sportsuccess consistent and still did not yield differences. Perhaps these differences are not or nolonger found in college and adult athletes. In both cases, male groups demonstrated higherlevels of aggression. The lack of significant differences in instrumental aggression betweensport types or gender is consistent with Bredemeier's (1994) work where no gender differenceswere reported when studying children's sport assertion. The results imply that women andmen do not differ in instrumental aggression. However, when interpreting these results, oneshould consider that Stephens (1998) has questioned the construct validity of certain items ofthe BAAGI-S suggesting that they may measure constructs other than sport instrumentalaggression.

Perhaps the expansion in the last thirty years of women's sport participants since theadoption of Title IX has resulted in similar sport socialization processes for both women andmen. With more participants and more women's sports turning corporate, the level of pressureand competition could now be more similar to men's sports. These similarities might haveinfluenced women's sport aggression in a way that they are now on par with men's sportaggression. The high demands from sport on an athlete in addition to greater chances in

Differences in Sport Aggression... /71

vicariously learning through an increase in exposure to women in sport perhaps in the media,could have modified social norms regarding sport aggression for female athletes. These chancesfor learning and opportunities for frustrating situations are consistent with the Revised Frus-tration-Aggression Hypothesis (Berkowitz, 1965). Again, the present study only containedthe sports of rugby, soccer, and volleyball; thus, it is unclear whether the results are indicativeof actual sport contact types or these specific sports.

Life Aggression and Life AssertionThe mean life aggression scores ofthe present study, given in Table 2, are comparable to

those from Buss and Durkee's (1957) original testing ofthe BDHI on college men and women.The results showed that total life aggression and life assertion did not differ between sportcontact types. The absence of differences across men's sport groups is contrary to previousresearch with children (Bredemeier, 1994; Bredemeier et al., 1986), where higher levels of con-tact sport participation and interest was related to higher life aggression. Interestingly,Bredemeier and colleagues foimd that sport interest was a better predictor of three moralityvariables than sport participation and had strong predictive qualities for physical and non-physical aggression. Perhaps sport interest, rather than participation, would have yieldeddifferent results in this study. Furthermore, physical size was recently found to be related to lifeaggression in male contact athletes, non-contact athletes, and non-athletes (Lemieux, McKelvie,& Stout, 2002). Although contact athletes reported more life aggression than non-contactathletes, there was no difference between physically bigger (in terms of height and weight)non-athletes. The researchers also propose the possibility that previous findings related todifferences in aggression across contact sports might have been related to the tendency forbigger individuals to participate in contact sports than non-contact sports.

The absence of differences between men and women in total life aggression contradictsprevious work (Bredemeier, 1994; Buss & Durkee, 1957; Gardner & Janelle, 2002). Men didscore higher on life assertion and assault aggression, but lower on the indirect hostility scalesthan women, indicating that there may be gender socialization differences in the expression ofaggression. The gender difference in assertion found is consistent with previous findingsfrom Nevid and Rathus (cited in Rathus, 1981) and Gardner and Janelle (2002), yet contrary toBredemeier (1994), suggesting that males report higher levels of assertion than females. Theassault differences are consistent with prior findings where boys reported more physicalaggression than girls (Bredemeier, 1994; Bredemeier et al., 1986). Perhaps this is indicative ofsocializing women to express their aggression and assertion but only in more circuitous meansthan men, tending to be more passive aggressive and relational aggressive. However, consid-eration should be given to the large variance in assertion scores before any conclusions are

72 / Journal of Sport Behavior, Vol. 30, No. 1

drawn based on this gender difference. It should also be noted that DeGiovanni and Epstein(1978) suggested that existing assessment tools of assertion failed to conceptually differenti-ate aggression from assertion. Among other scales, they stated that the RAS appeared toconfound aggression and assertion with the positive correlations of 13 items to ratings ofaggressiveness.

Life and Sport RelationshipsThe results support the significant positive relationship between both men's and women's

sport hostile aggression scores and their total life aggression scores. Additionally, for bothwomen and men, life aggression scores were significantly negatively related to their sportinstrumental aggression scores. The results indicate that athletes with high levels of lifeaggression are more likely to have high levels of sport hostile aggression but low levels ofsport instrumental aggression. These results are consistent with the original testing of theBAAGI long form (Bredemeier, 1975). This suggests that trait aggression may be consistentacross sport and life domains for both men and women and refutes the theory that sportpersonalities do not refiect those of everyday life (e.g. Bredemeier, 1994; Bredemeier et al.,1986; Thirer, 1993). Different conclusions are drawn for assertiveness if sport instrumentalaggression can be used as a valid measure of sport assertion. The findings indicate that sportinstrumental aggression for both men and women is not related to assertion in daily life,implying that assertion may not be consistent across sport and life domains. However, withoutan instrument validated to measure sport assertiveness, conclusions are merely tentative.Interestingly, life assertion was positively related to life aggression but was only significantfor women. Perhaps this suggests that women can be categorized as either forward (i.e., highin aggressive and assertive behavior) or not forward (i.e., low in aggressive and assertivebehavior).

ConclusionThis study revealed that there were no differences in life aggression, life assertion, sport

hostile aggression, and sport instrumental aggression between different contact sport typesor interactions between sport types and gender when controlling for individual's age, yearsexperience, education, team, and success level. The only differences that were indicated werethat male and female athletes have varying levels of reported life assertion, assault, and indi-rect hostility. This study completes a gap in the existent research in comparing the same non-contact, contact, and collision sports between male and female adult athletes. The contribu-tion of these results is that they counter previous suggestions that a greater degree of contactsport participation is related to aggression and may be detrimental to moral and character

Differences in Sport Aggression... /73

development more so than sports with less contact. Differences found in children's sport andlife aggression across varying levels of contact sports had led to conclusions that contactsport participation is counterproductive to moral advancement for pre-adolescents (Bredemeier,1988; Bredemeier, et al., 1986). Although the present examination was confined to an adultsample, it included athletes that had been participating in their sport since childhood. Thetransition between youth participation and adult participation is unclear. However, previousdifferences found in children's aggression were not present in this adult sample. Correlationsbetween life and sport variables generally supported previous fmdings (Bredemeier, 1975;Buss & Durkee, 1957).

It is suggested that fiiture research expand the type (e.g. coactive team, individual) andnumber of sports per category to clarify whether the current results represent contact types orare specific to the sports of rugby, soccer, and volleyball. Regardless, the absence of genderdifferences in hostile aggression for same sport participants is interesting and may be a prod-uct of increased participation in sports by women. Results should also be interpreted withcaution as even though the differences found between men and women had moderate (indirecthostility) and large effect sizes (assertion and assault aggression), the observed powers werelow. A larger number of participants would be necessary in any follow up studies to observe ahigher power.

Silva (1983) suggested that as women's involvement in highly competitive organizedsports increased, aggression differences found between male and female athletes might change.In addition, in 1972, Smith predicted that as the expansion of women in contact sports in-creased, and as winning became more important, women's sports would become more vio-lently aggressive. Thirty years later, we have seen a dramatic increase in women's participationin all types of sports including physical contact activities. It is evident from the media andcollege campuses that women's contact and collision sport participation is increasing. In thepresent investigation, the absence of gender differences in sport aggressions and total lifeaggression lend support to Silva and Smith's predictions. Perhaps the increasing extrinsicrewards in sport, the promotion of a "win at all costs" attitude, and the reinforcement ofharmful acts in sport, have forced athlete's aggression to transcend sport type as well asgender.

74 / Journal of Sport Behavior, Vol. 30, No. 1

References

Allawy, M.H. (1981). Differences in athletic aggression among Egyptian female athletes.Medicine and Sport, 75,63-66.

Bandura, A. (1973). Aggression: a social learning analysis. New Jersey: Prentice-Hall, Inc.Berkowitz, L. (1965). The concept of aggressive drive: Some additional considerations. In L.

Berkowitz (Ed.), Advances in experimental social psychology (pp.301-329). NewYork:Academic Press.

Bredemeier, B.J. (1975). The assessment of reactive and instrumental athletic aggression. InD.M. Landers (Ed.), Psychology of sport and motor behavior II (pp.7l-S4). Penn StateHPER Series No. 10. The Pennsylvania State University.

Bredemeier, B.J. (1988). The moral of the youth sport story. In E.W. Brown & C.F. Branta(Eds.), Competitive sports for children and youth: An overview of research andissues (pp. 285-296). Champaign, IL: Human Kinetics.

Bredemeier, B.J. (1994). Children's moral reasoning and their assertive, aggressive, andsubmissive tendencies in sport and daily life. Journal of Sport and Exercise Psychol-ogy, 16,1-14.

Bredemeier, B.J., & Shields, D.L. (1986). Moral growth among athletes and nonathletes: Acomparative analysis. The Journal of Genetic Psychology, 147(1), 7-18.

Bredemeier, B.J., Weiss, M.R., Shields, D.L., & Copper, B.A.B. (1986). The relationship ofsport involvement with children's moral reasoning and aggression tendencies.Journal of Sport Psychology. 8(4), 304-318.

Buss, A.H., & Durkee, A. (1957). An inventory for assessing different kinds of hostility.Journal of Consulting Psychology, 4,343-349.

Buss, A.H. (1961). Psychology of aggression. NewYork: John Wiley and Sons, Inc.Colley, A., Roberts, N., & Chipps, A. (1985). Sex role identity, personality and participation in

team and individual sports by males and females. International Journal of SportPsychology, 16(2), 103-112.

DeGiovanni, I.S., & Epstein, N. (1978). Unbinding assertion and aggression in research andclinical practice. 5e/iaviorMb<ii/icarion, 2(2), 173-192.

Duda, J.L., Olson, L.K., & Templin, T.J. (1991). The relationship of task and ego orientationto sportsmanship attitudes and the perceived legitimacy of injurious acts. ResearchQuarterly for Exercise and Sport, 62(1), 79-87.

Dunn, J.G., & Dunn, J.C. (1999). Goal orientations, perceptions of aggression, andsportspersonship in elite male youth ice hockey players. The Sport Psychologist, 13,183-200.

Differences in Sport Aggression.. ./75

Howe, B. (1973). Aggressive responses of rugby players. New Zealand Journal Of Health,Physical Education, and Recreation, April.

Kemler, D.S. (1988). Level of athletic, instrumental, and reactive aggression betweencontact and non-contact, male and female high school athletes under pre- and post-testing conditions. Unpublished master's thesis. Southern Connecticut State Univer-sity.

Kerr, J.H. (1999). The role of aggression and violence in sport: A rejoinder to the ISSPposition stand. Sport Psychologist, 13(1), 83-88.

Kerr, J.H. (2002). Issues in aggression and violence in sport: The ISSP position standrevisited. Sport Psychologist, 16(1), 68-78.

Lemieux, P., McKelvie, S.J., & Stout, D. (2002). Self-reported hostile aggression in contactathletes, no contact athletes and non-athletes. Athletic Insight, 4(3). RetrievedSeptember 15,2005, fromhttp://www.athleticinsight.comA'bl4Iss3/SelfReportedAggression.htm.

Mace, R., & Baines, C. (1989). Personality characteristics and mood states of females inspoTi. Physical Education Review, 12(2), 158-162.

Maier, R.A., & Lavrakas, P.J. (1981). Some personality correlates of attitudes about sports.IntemationalJoumal of Sport Psychology, 12(1), 19-22.

Morgan, W.P. (1980). The Trait Psychology Controversy. Research Quarterly for Exerciseand Sport, 51(1), 50-15.

Rail, G. (1992). Physical contact in women's basketball: Aphenomenological constructionand contextualization. International Review for the Sociology of Sport, 27 (1), 1-25.

Rathus, S.A. (1973). A 30-item schedule for assessing assertive behavior. Behavior Therapy,4,398-406.

Rathus, S.A. (1981). Psychology. New York: CBS College Publishing.Reid, R., & Hay, D. (1979). Some behavioral characteristics of rugby and association

ioothaWers. IntemationalJoumal of Sport Psychology, 10(4), 239-251.Rosenberg, D. (2003). The banality of violence in sport and the McSorley afiFair. Avante,

9(2),20A2.Ryan, M.K., Williams, J.M., & Wimer, B. (1990). Athletic aggression: Perceived legitimacy

and behavioral intentions in girl's high school basketball. Journal of Sport andExercise Psychology, 72,48-55.

Salisbury, J., & Passer, M. (1982). Gender-role attitudes and participation in competitiveactivities of varying stereotypic femininity. Personality and Social Psychology, 8(3),486493.

76 /Journal of Sport Behavior, Vol. 30, No. 1

Silva, J.M. (1978). Assertive and aggressive behavior in sport: A definitional clarification. InNaedau, C.H. et al. (Eds.), Psychology of motor behavior and sport (pp. 199-208).Champaign, 111: Human Kinetics.

Silva, J.M. (1983). The perceived legitimacy of rule violating behavior in sport. Journal ofSport Psychology, 5,438-448.

Smith, M.D. (1972). Aggression and the female athlete. In D.V. Harris (Ed.), Women andsport: A national research conference (pp. 91 -114). University Park, PA: Penn StateUniversity.

Smith, D., & Stewart, S. (2003). Sexual aggression and sports participation. Journal of SportBehavior, 26,384-396.

Stephens, D.E. (1998). Aggression. In J.L. Duda (Ed.), Advances in sport and exercisepsychology measurement (pp.277-292). Morgantown, WV: Fitness InformationTechnology, Inc.

Tenenbaum, G., Stewart, E., Singer, R.N., & Duda, J. (1997). Aggression and violence in sport:An ISSP position stand. Sport Psychologist, 11, 1-7.

Tenenbaum, G, Sacks, D.N., Miller, J.W., Golden, A.S., Doolin, N. (2000). Aggression andviolence in sport: A reply to Kerr's rejoinder. Sport Psychologist, 14(4), 315-326.

Theberge, N. (2003). "No fear comes": Adolescent girls, ice hockey and the embodiment ofgender. Youth & Society, 34 (4), 497-516

Thirer, J. (1993). Aggression. In R.N. Singer, M. Murphey, & L.K. Tennant (Eds.), Handbookof research on sport psychology (pp. 365-378). New York: MacMillian PublishingCompany.

Tucker, L. W., & Parks, J.B. (2001). Effects of gender and sport type on intercollegiateathlete's perceptions of the legitimacy of aggressive behaviors in sport. Sociology ofSport Journal, 18(4),A0'i-A\2,.

Wall, B.R., & Gruber, J.J. (1986). Relevancy of athletic aggression inventory for use inwomen's intercollegiate basketball: A pilot investigation. International Journal ofSport Psychology, 77,23-33.

Wann, D.L., Fahl, C.L., Erdmann, J.B., & Littleton, J.D. (1999). Relationship between identifi-cation with the role of sport fan and trait aggression. Perceptual and Motor Skills,88,1296-1298.

White, J., & Kowalski, R. (1994). Deconstructing the myth of the nonaggressive woman: Afeminist analysis. Special issue: Transformations: Reconceptualizing theory andresearch with women. Psychology of Women Quarterly, 18(4), 487-508.

Widemeyer, W.N. (1984). Aggression - performance relationships in sport. In J.M. Silva andR.S. Weinberg (Eds.), Psychological foundations of sport (pp. 274-286). Champaign,IL: Human Kinetics Publishers, Inc.