leadership, cohesion, and team norms regarding … · leadership, cohesion, and team norms...

13
Sociology of Sport Journal, 1995, 12, 324-336 O 1995 Human Kinetics Publishers. Inc. Leadership, Cohesion, and Team Norms Regarding Cheating and Aggression David Lyle Light Shields Brenda Jo Light Bredemeier John F. Kennedy Univ. Univ. of California, Berkeley Douglas E. Gardner Alan Bostrom John F. Kennedy Univ. Univ. of California, San Francisco The present study drewfrom the model of moral action proposed by Shields and Bredemeier (1995) according to which a sport team's collective norms influence behavior. The focus was on team cheating and aggression norms in relation to demographic variables, leader- ship style, and team cohesion. Participants were baseball and softball players (N=298) at the high school and community college level. It was found that age, year in school, and years playing ball all correlated positively with expectations of peer cheating and aggression, and with the belief that the coach would sanction cheating if necessary to win. MANOVA results indicated higher anticipations of cheating and aggression among males, college athletes, winning team members, and nonstarters. Significant relationships between leadership style variables and collective team norms, and between team cohesion variables and collective team norms, were also obtained. La pre'sente ttude emprunte le rnodgle d'action morale en sport propost par Shields et Bredemeier (1995) selon lequel l'atmosphgre moral entourant l'tquipe sportive est un facteur important influant sur les actions morales des athlgtes. L'ttude porte sur les relations entre d'une part les normes de I'tquipe en ce qui concerne la tricherie et I'agression et d'autre part les variables dtmographiques, le style de leadership et la cohtsion. Au total, 298 joueurs et joueuses de base-ball et softball de niveau secondaire et colle'gial ont comple'te' des questionnaires. Les re'sultats indiquent que l'cige, l'annte scolaire, et le nombre d'anne'es d'expe'rience en base-ball ou softball sont en corrtlation positive avec les attentes en terme d'agression et de tricherie des pairs et la croyance que I'entraineur approuverait la tricherie si elle e'tait ntcessaire ci la victoire. Les analyses MANOVA indiquent des attentes plus tleve'es chez les hommes, les athMtes colltgiaux et les membres d'e'quipes gagnantes. Des relations significatives existent tgalement entre le style de leadership et les normes de l'tquipe, et entre ces normes et la cohe'sion. Moral action in sport contexts is a theme of increasing interest to sport social scientists. In a recent model of the influences on moral action in physical D.L.L. Shields and D.E. Gardner are with the Graduate School of Professional Psychology at John F. Kennedy University, Orinda, CA 94563. B.J. Light Bredemeier is with the Dept. of Human Biodynamics at the U. of California, Berkeley, Berkeley, CA 94720. A. Bostrom is with the Center for AIDS Prevention Studies at the U. of California, San Francisco, San Francisco, CA 94143.

Upload: hadiep

Post on 31-Mar-2018

219 views

Category:

Documents


1 download

TRANSCRIPT

Sociology of Sport Journal, 1995, 12, 324-336 O 1995 Human Kinetics Publishers. Inc.

Leadership, Cohesion, and Team Norms Regarding Cheating and Aggression

David Lyle Light Shields Brenda Jo Light Bredemeier John F. Kennedy Univ. Univ. of California, Berkeley

Douglas E. Gardner Alan Bostrom John F. Kennedy Univ. Univ. of California, San Francisco

The present study drewfrom the model of moral action proposed by Shields and Bredemeier (1995) according to which a sport team's collective norms influence behavior. The focus was on team cheating and aggression norms in relation to demographic variables, leader- ship style, and team cohesion. Participants were baseball and softball players (N=298) at the high school and community college level. It was found that age, year in school, and years playing ball all correlated positively with expectations of peer cheating and aggression, and with the belief that the coach would sanction cheating i f necessary to win. MANOVA results indicated higher anticipations of cheating and aggression among males, college athletes, winning team members, and nonstarters. Significant relationships between leadership style variables and collective team norms, and between team cohesion variables and collective team norms, were also obtained.

La pre'sente ttude emprunte le rnodgle d'action morale en sport propost par Shields et Bredemeier (1995) selon lequel l'atmosphgre moral entourant l'tquipe sportive est un facteur important influant sur les actions morales des athlgtes. L'ttude porte sur les relations entre d'une part les normes de I'tquipe en ce qui concerne la tricherie et I'agression et d'autre part les variables dtmographiques, le style de leadership et la cohtsion. Au total, 298 joueurs et joueuses de base-ball et softball de niveau secondaire et colle'gial ont comple'te' des questionnaires. Les re'sultats indiquent que l'cige, l'annte scolaire, et le nombre d'anne'es d'expe'rience en base-ball ou softball sont en corrtlation positive avec les attentes en terme d'agression et de tricherie des pairs et la croyance que I'entraineur approuverait la tricherie si elle e'tait ntcessaire ci la victoire. Les analyses MANOVA indiquent des attentes plus tleve'es chez les hommes, les athMtes colltgiaux et les membres d'e'quipes gagnantes. Des relations significatives existent tgalement entre le style de leadership et les normes de l'tquipe, et entre ces normes et la cohe'sion.

Moral action in sport contexts is a theme of increasing interest to sport social scientists. In a recent model of the influences on moral action in physical

D.L.L. Shields and D.E. Gardner are with the Graduate School of Professional Psychology at John F. Kennedy University, Orinda, CA 94563. B.J. Light Bredemeier is with the Dept. of Human Biodynamics at the U. of California, Berkeley, Berkeley, CA 94720. A. Bostrom is with the Center for AIDS Prevention Studies at the U. of California, San Francisco, San Francisco, CA 94143.

Cheating and Aggression Norms 325

activity settings, Shields and Bredemeier (1995) propose that the "moral atmo- sphere" characterizing a sport setting will play a significant role in mediating moral action. Central to the concept of moral atmosphere is the idea that any group, over time, develops collective norms about appropriate behavior on the part of group members. In this study, collective team norms regarding cheating and aggression were investigated in relation to coaches' leadership style, team cohesion, and player and coach demographics.

A nearly endless stream of media reports detail instances of rule violation and aggression by athletes. While there are no reliable statistics to indicate whether the problems of cheating and aggression are any more frequent now than in times past, there is certainly a perception on the part of many that unfair and harmful behaviors are widely endorsed, openly or surreptitiously, in the sport world (Coakley, 1994; Kohn, 1986; Orlick, 1978; Smith, 1983).

Two lines of research have helped shed light on moral issues in the context of sport. The f is t goes back to a seminal study by Webb (1969), who provided evidence that a socialization process occurs in sport that leads athletes to increas- ingly prioritize victory over fair play. According to the main findings of this literature on professionalization, an early focus on fair play and fun is increasingly supplanted by a value orientation centered on winning as athletes mature, spend more time in sport, or enter more elite realms of sport (see Knoppers, 1985, and Shields & Bredemeier, 1995, for reviews and critiques).

In a second relevant line of research, Bredemeier and Shields and their colleagues have investigated moral action in sport using the tools of structural developmental and social cognitive psychology. Key findings include the follow- ing: (a) There is an inverse relationship between athletic aggression and moral reasoning maturity (Bredemeier, 1994; Bredemeier & Shields, 1984b); (b) there is a negative relationship between moral reasoning maturity and athletes' legitimacy judgments regarding athletic aggression (Bredemeier, 1985; Bredemeier, Weiss, Shields, & Cooper, 1987); (c) children's participation in high contact sports is associated with less mature moral reasoning andgreater self-reported tendencies to aggress (Bredemeier, Weiss, Shields, & Cooper, 1986); and (d) both athletes and nonathletes use patterns of moral reasoning that appear to diverge structurally when making judgments about sport versus nonsport moral issues (Bredemeier, in press; Bredemeier & Shields, 1984a).

Shields and Bredemeier (1995) have proposed a model of moral action in sport that highlights 12 sets of influences on moral behavior. Of particular interest to the present research, they identify "moral atmosphere" as a significant contextual mediator of moral action. They draw from the work of Kohlberg to operationally define moral atmosphere as a set of collective norms regarding morally relevant action on the part-of group members (Power, Higgins, & Kohlb- erg, 1989). A collective norm, in turn, is "a complex of specific behavioral expectations that share a common value" (Power et al., 1989, p. 115).

Though terminology varies, the construct of collective norms has been discussed by various sociologists of sport. Donnelly (Donnelly & Young, 1985; Williams & Donnelly, 1985), for example, discussed collective norms in his analysis of the reproduction and transformation of cultural forms of sport, and Fine (1987) used the construct to characterize moral aspects of the ideoculture of Little League baseball teams. Nixon (1992, 1994) employed the construct to highlight sport norms that encourage coaches and athletes to accept risk, pain,

326 Shields, Bredemeier, Gardner, and Bostrom

and injury, while Duquin (1984) found that women more than men embrace collective norms that mitigate against tactics that might lead to pain or injury.

As used in the present study, collective norms are one component of moral atmosphere. One research strategy for assessing the existence and strength of moral collective norms, particularly those dealing with moral transgression, is to ask study participants about how many of their peers would likely behave in a particular way. When someone indicates a high percentage of peers would act in a transgressive or prosocial fashion, it infers the presence of a collective norm endorsing that mode of behavior (Power et al., 1989).

This strategy was used recently by Stephens and Bredemeier (1995) in a study of girls' youth soccer. They found that the best predictor of players' self- described likelihood to engage in unfair game tactics (operationally defined as cheating, lying to an official, and hurting an opponent) was their belief about the likelihood that teammates would play unfairly. Stephens and Bredemeier also found that the players' perception of their coaches' motivational orientation (task or ego) was a stronger predictor of their likelihood to engage in unfair game tactics than even their own motivational orientation. These findings support the contention that moral atmosphere plays a significant role in mediating moral action in sport, and that peers and the coach all play a prominent role in shaping the team's collective norms.

The present study takes both the coach and teammates into account by investigating the relationship of collective team norms with the coach's leadership style and team cohesion. Most recent studies of sport leadership have been based on the Multidimensional Model of Leadership (Chelladurai & Carron, 1978), which gave rise to the Leadership Scale for Sports (LSS) (Chelladurai & Saleh, 1980). The LSS has been used for many purposes, such as to determine the relationship between preferred leader behaviors and athlete maturity (Chella- durai & Carron, 1983), to assess the behavior preferences of elite athletes (Terry, 1984), and to characterize the leadership styles of coaches in different sports or contexts (Dwyer & Fischer, 1988; Serpa, Pataco, & Santos, 1991).

More closely paralleling the present study, the LSS has also been used to investigate the relationship between leadership style and an outcome variable, such as performance (Gordon, 1986; Home & Carron, 1985; Serpa, Pataco, & Santos, 1991; Weiss & Friedrichs, 1986), satisfaction (Chelladurai, 1984; Chella- durai, Imamura, Yamaguchi, Oinuma, & Miyauchi, 1988; Dwyer & Fischer, 1990; Home & Carron, 1985; McMillin, 1990; Schliesman, 1987; Weiss & Friedrichs, 1986), or team cohesion (Gardner, Shields, Bredemeier, & Bostrom, 1995; Pease & Kozub, 1993; Shields, Gardner, Bostrom, & Bredemeier, 1995; Westre & Weiss, 1991). One outcome of specific coaching styles, previously unstudied, may be a tendency for team members to construct shared social understandings regarding appropriate and inappropriate game behavior.

In addition to leadership variables, the extent of team cohesion may influ- ence the norms shared collectively. Since few people want to perceive themselves as moral outcasts (Haan, Aerts, & Cooper, 1985), it may be that for collective norms to sanction such behaviors as cheating and aggression, a fairly high level of team cohesion must exist. A high level of team cohesion might facilitate individuals' beliefs that others are united in the prioritizing of outcome over fair play. It also makes sense theoretically that as teams become more cohesive, the

Cheating and Aggression Norms 327

players will increasingly share common understandings as to the appropriateness of such behaviors as cheating and aggression.

The purpose of the present study was to explore leadership, cohesion, and demographic variables in relation to team norms about cheating and aggression. Due to a lack of previous research on leadership and collective team norms, the present study was considered exploratory and no hypotheses were advanced. For similar reasons, no hypotheses were formulated with regard to the relationship between team cohesion and the extent to which team members would expect peers to cheat or aggress. In keeping with previous research, however, it was hypothesized that higher expectations of cheating and aggression would be associ- ated with (a) males more than females, (b) older athletes more than younger athletes, and (c) athletes with more extensive sport experience than those with less experience.

Method

Study Participants

The participants in this study were baseball (n = 182) and softball (n = 116) players representing six community colleges and six high schools from large urban areas of California. All baseball players were male, all softball players were female. Players from 23 teams, 12 softball and 11 baseball, participated in the study.

High school athletes (n = 106) were drawn from five high school varsity baseball teams (52) and six varsity softball teams (54). They ranged in age from 13 to 19 (M = 16.3, SD = 1.07). Among the high school athletes, 77 described themselves as starters, 27 as nonstarters, and 2 did not respond to the question. All high school teams were from inner-city public schools; study participants described themselves as members of the following raciallethnic groups: African- American (64), HispanicILatino (19), European-American (1 I), Asian-American (6), Filipino (5), and Native American (1).

The six baseball (130) and six softball (62) community college teams were all from the same conference in Northern California. The community college athletes ranged in age from 18 to 33 (M = 19.7, SD = 1.89). Of the community college athletes, 125 described themselves as starters, 64 as nonstarters, and 3 did not respond. Community college athletes (n = 192) reflected the following racial-ethnic composition: European-American (1 lo), HispanicJLatino (39), Afri- can-American (26), Asian-American (a), Filipino (2), Native American (2), and Hawaiian (2); three respondents did not report their raciallethnic identity.

Measures

Four instruments were used in the present study: a demographic question- naire, three versions of the Leadership Scale for Sports (LSS), the Group Environ- ment Questionnaire (GEQ), and the Team Norm Questionnaire.

Demographic Questionnaire. The demographic questionnaire elicited in- formation about participants' age, gender, racelethnicity, school level, starting status, and playing history. For each team, the win-loss record and the sex of the head coach was also obtained.

328 Shields, Bredemeier, Gardner, and Bostrom

The Leadership Scale for Sports. The LSS contains 40 items that assess five dimensions of coaching behavior: training and instruction (TI), democratic behavior (DB), autocratic behavior (AB), social support (SS), and positive feed- back (PF). These five dimensions consist of two decision making styles (autocratic and democratic behavior), two motivational tendencies (social support and posi- tive feedback), and one instructional behavior (training and instruction). The primary distinction between social support and positive feedback is that the former is given regardless of performance whereas the latter is contingent on performance.

The LSS was administered in each of its three possible forms: (a) a perceived version that asks athletes to describe their coaches' actual behavior, (b) a preferred version that asks athletes to describe the coaching behaviors they would prefer, and (c) a coach's version that asks the coach to describe his or her own leadership behaviors. Item responses were recorded on a 5-point Likert scale ranging from never to always, and scores for each LSS scale were produced by summing the item responses and dividing by the number of items in that category.

The Group Environment Questionnaire. Based on the multidimensional model of group cohesion developed by Carron, Widmeyer, and Brawley (1985), the GEQ measures four aspects of team cohesiveness. Specifically, individual attraction to the grouptask (ATG-T) is a measure of individual team members' feelings about their personal involvement with the group task. Individual attraction to the groupsocial (ATG-S) is a measure of individual team members' feelings about personal involvement and social interaction with the group. These first two subscales assess personal attraction to the group's task and social activities. In contrast, the next two subscales assess the individual's perceptions of the group as a whole in terms of its coherence around task and social activities. Specifically, group integration-task (GI-T) is a measure of the individual team members' feelings about the similarity, closeness, and bonding within the team as a whole around the group's task, and group integration-social (GI-S) measures the individual team members' feelings about the similarity, closeness, and bonding within the team, but for the group as a social unit. The GEQ consists of 18 nine- point Likert scale items ranging from strongly disagree to strongly agree. Scores for each scale were computed by summing the relevant item values and dividing by the number of items in that category.

The Team Norm Questionnaire. The TNQ, an experimental measure con- structed specifically for this study, comprised six items designed to assess four components of collective team norms. Two questions asked athletes to estimate how many of their teammates would violate a rule if it would help their team win (Peer Cheat), and two questions asked athletes to estimate how many of their teammates would deliberately hurt an opponent if it would help their team win (Peer Aggress). The following six response options, together with their point coding, were offered: none (0), a few (I), several (2), about half (3), most (4), and everyone (5). Thus the scale for these questions ranged from 0 to 5. Finally, two questions pertained to the coach and queried as to whether, in the athlete's opinion, the coach would want the athlete to cheat (Coach Cheat) or injure an opponent (Coach Aggress) if it would help the team win. Responses were provided using the following categories and point values: no (0), probably not (1), probably (2), and yes (3). Thus the scale for these two questions ranged from 0 to 3.

Cheating and Aggression Norms 329

Procedures

At the high school level, permission was obtained from district commission- ers, athletic directors, and team coaches. All six high schools gave consent' for administering the questionnaires. One baseball team was not available due to conflicting schedules, which left us with an odd number of high school teams, 11. For the community college teams, permission was obtained through each school's athletic director and head coach. Questionnaires were administered during the last third of the regular season. Players completed the demographic questionnaire, followed by the two versions of the LSS, the GEQ, and finally the TNQ. The measures were administered by one of us; the coach was not present during administration. The coach completed the LSS coach's version separately from the athletes, but at the same time.

Results

Measure Statistics

Internal reliability scores using Cronbach's alpha coefficient for four of the LSS scales were reasonably strong: .88 (training & instruction), .83 (democratic behavior), .8 1 (social support), .85 (positive feedback). The alpha reliability score for the Autocratic Behavior scale was rather low, .65, thus one should be cautious about drawing firm conclusions with regard to this scale.

Internal reliability for the four GEQ subscales was also somewhat low: .60 (ATG-T), .61 (ATG-S), .68 (GI-T), and .60 (GI-S). Since these reliabilities were not robust, we combined the two task subscales into a single composite Task Cohesion scale, and the two social subscales into a single composite Social Cohesion scale. When the two task and social dimensions were combined, the reliability scores improved: .72 (task) and .7 1 (social). Only the composite scales were used in further analyses.

The TNQ consists only of the six experimental questions. Intercorrelations among the TNQ items were fairly high. The Peer Cheat scale correlated at .68 with Peer Aggress (p < .0001), .57 with Coach Cheat (p < .0001), and .36 with Coach Aggress ( p < .0001). The correlation between Peer Aggress and Coach Cheat was .47 (p < .0001), and the correlation between Peer Aggress and Coach Aggress was .49 (p < .0001). Finally, the two coach items correlated at .55 (p < .0001). These findings suggest that athletes tended to perceive their peers and their coach similarly with regard to cheating and aggression norms, and those athletes who thought their peers and coach generally approved of cheating also tended to think their peers and coach approved of aggression.

Demographics and Team Norms

Two types of analyses were conducted to assess the relationships among demographic variables and the constructs assessed through the TNQ. First, corre- lation analyses were conducted, the results of which are presented in Table 1.

As can be seen, age, year in school, and years playing baseball or softball all correlated positively with the same three TNQ constructs-all except Coach Aggress. A positive correlation was also found between teams with higher winning

330 Shields, Bredemeier, Gardner, and Bostrom

Table 1 Correlations Between Team Norms and Demographic Variables

Peer cheat Peer aggress Coach cheat Coach aggress

Age .13" .14' .22" ns School year .2ga .26" .32" ns Years in baseball/ .29" .22a .3 la ns

softball Seasons on team ns ns ns ns Win percentage .20a . lgb .13' .18b

percentages and higher TNQ scores. In contrast, the number of seasons the athlete had been involved with the current team did not correlate with any of the TNQ constructs.

In order to determine whether team norms differed significantly by gender, school level, team success, and starting status, we conducted two multivariate analyses of variance (MANOVA). Gender, school level, and team success were combined into a single analysis since these refer to between-team variables. In contrast, playing status (starter vs. nonstarter) is a within-team variable and therefore was analyzed separately.

First, a three-factor (2 x 2 x 2) MANOVA using gender, school level, and team success as independent variables and team norms as dependent variables was conducted. For team success, teams were divided into those with winning records (i.e., those that won more than 50% of games, 13 teams and 173 players) and those with losing records (10 teams and 125 players). As indicated in Table 2, all three main effects were significant, as was the gender-by-school-level interaction. The univariate analyses, summarized in Table 3, indicated the follow- ing significant differences: (a) males were higher than females for all four TNQ variables; (b) college players were higher than high school players for all TNQ variables except Coach Aggress; and (c) winning teams were higher than losing teams (significant only for Peer Aggress).

The interaction plots for the gender-by-school-level interactions revealed that for the Peer Cheat variable, males increased dramatically from high school (M = 3.04) to college (M = 4.33), while females declined slightly (from M = 2.42 to 2.21). Both genders increased from high school to college with regard to Peer Aggress, though the increase for males was much sharper (from 2.47 to 3.57) than for females (from 1.72 to 2.08). For the coaching variables the patterns displayed a consistent increase from high school to college. With regard to Coach Cheat, males rose from a mean of .56 in high school to a mean of 1.62 in college; the increase for females was from .44 to .73. For the Coach Aggress variable, the increase for males was from .57 to 30 , and for females it went from .25 to .44.

A one-factor MANOVA using starting status as the independent variable and the four TNQ constructs as the dependent variables revealed a significant

Cheating and Aggression Norms 331

Table 2 Multivariate Relationships Between Demographic Variables and Team Norms

Source Lambda F(4, 278) P

Gender School level Success Gender . school level Gender . success School level . success Gender . school level . success

Starter status

Table 3 Univariate Relationships Between Demographic Variables and Team Norms

Peer cheat Peer aggress Coach cheat Coach aggress

F ( l , 281) F ( l , 281) F ( l , 281) F ( l , 281) Gender 68.08" 41.26" 18.22" 8.72b School level 10.47b 17.49' 32.84' 3.39 Success 1.22 9.98' 0.47 1.84 Gender . school level 20.72" 4.6@ 10.76' 0.03

F ( l , 291) F ( l , 291) F ( l , 291) F(1, 291) Starter status 8.05b 14.54" 2.05 2.55

effect (see Tables 2 and 3). The univariate analyses revealed that nonstarters were higher than starters on the two peer variables, but not on the coach variables.

To follow up on the gender difference revealed in the first MANOVA, we ran one other analysis. Since some female athletes were coached by male coaches and others by female coaches, we ran t tests comparing female athletes of male coaches and female athletes of female coaches on each of the four TNQ variables. Female athletes of male coaches were higher than female athletes of female coaches on Peer Cheat, but the difference only approached significance (t = 1.77, df = 114, p < .08). For Peer Aggress, the difference between the two groups was significant ( t = 3.38, df = 114,p < .001). The coach variables were also significant, with females athletes of male coaches again higher than those of female coaches (for Coach Cheat, t = 2.16, df = 114, p < .05; and for Coach Aggress, t = 2.76,

332 Shields, Bredemeier, Gardner, and Bostrom

df = 114, p < .01). Similar analyses could not be run for the males since they all had male coaches.

Leadership and Team Norms

To determine whether there was any relationship between leadership behav- iors as assessed through the three versions of the LSS, and the team norm variables as assessed by the TNQ, we conducted three multivariate canonical correlation analyses.

The overall test of a multivariate relationship between perceived leader behaviors and team norms was significant, Wilks' lambda = .815, F(20, 959) = 3.05, p = .0001. The canonical correlation between the two sets of variables was r, = .41. Labeling variables with loadings greater than .30 as significant (Pedhazur, 1982), we found that two of the five LSS scales contributed significantly to the relationship, with AB(.65) most important, followed by SS(.46). Three of the four TNQ constructs contributed significantly to the multivariate relationship: Peer Cheat (.85), Coach Cheat (.85), and Peer Aggress (.52).

The overall test of a multivariate relationship between preferred leader behaviors and team norms was also significant, Wilks' lambda = 375, F(20, 959) = 1 . 9 8 , ~ = .0065. The canonical correlation between the two sets of variables was rc = .32. Of the five LSS scales, the same two again contributed significantly to the relationship: SS(.77) and AB(.40). As with the previous analysis, Peer Cheat (.98), Peer Aggress (.65), and Coach Cheat (.40) were the contributing items from the TNQ.

Finally, the overall test of a multivariate relationship between the coach's version of the LSS and team norms was also significant, Wilks' lambda = .702, F(20, 959) = 5.40, p = .0001. The canonical correlation between the two sets of variables was rc = .50. Four of the LSS scales contributed significantly to the relationship: SS(.90), TI(.%), PF(.53), and AB(.52). All four of the TNQ con- structs contributed significantly to the relationship: Peer Cheat (.91), Coach Cheat (.El), Peer Aggress (.51), and Coach Aggress (.35).

Team Cohesion and Team Norms

To determine whether there was a relationship between task and social team cohesion as assessed by the GEQ, and team norm variables as assessed by the TNQ, we conducted a canonical correlation analysis. The overall test of a multivariate relationship between the two GEQ scales and the four team norms was significant, Wilks' lambda = .929, F(8,584) = 2.76, p = .0054. The canonical correlation between the two sets of variables was r, = .25. Task Cohesion (-.72) contributed significantly to the relationship, but Social Cohesion (.24) appeared to play only a marginal role. Three of the four TNQ constructs contributed significantly: Peer Cheat (.74), Peer Aggress (.72), and Coach Cheat (.52).

To determine whether there was a relationship between task or social cohesion and relative consensus regarding team norms, we conducted a simple correlation analysis between task and social cohesion and the standard deviation for each of the four TNQ scales. None of the correlations reached significance.

Cheating and Aggression Norms 333

Discussion The various demographic analyses demonstrated that cheating and aggres-

sion are, in general, more expected at the college level as opposed to the high school level, and by males, older athletes, and those more experienced in the sport. These findings are generally consistent with the literature on professionalization (Knoppers, 1985; Shields & Bredemeier, 1995) and other empirical findings suggesting a socialization process in sport that is supportive of an increased acceptance of aggression (Bredemeier, 1985; Silva, 1983; Smith, 1976). On the other hand, the finding that community college females expected less cheating from peers than did high school females deviated from the expected pattern and would seem to be an anomaly in the results.

The finding of a positive relationship between playing on a winning team and higher expectations regarding peer cheating and aggression were unantici- pated. Perhaps those who subordinate the value of fair play to that of victory do in fact win more often. This conclusion, however, is rendered less likely by another unanticipated finding, namely that nonstarters were higher than starters in the expectation of peer cheating and aggression. Perhaps a nonstarter, who can observe game behavior but cannot know starting players' motivations and thoughts, tends to overestimate the starting teammates' willingness to engage in cheating and aggression if necessary to win. Similarly, being a nonstarter may encourage an attribution process in which starting peers are thought to have obtained their position through a greater willingness to value victory over all other values. Perhaps, alternately, starting players are simply less willing to characterize teammates in unsportspersonlike terms.

It is interesting to note that female athletes who had a female coach believed their teammates were less likely to cheat or aggress, and their coaches less accepting of such behaviors, than female athletes who had a male coach. This may suggest that males who coach female teams continue to promote the same priorities as their male colleagues coaching male sport teams, which may create value conflicts for many female athletes (Blinde, 1989). Particularly for those who critique the traditional male model of sport and do not wish female sports to be modeled on it, this finding may be disturbing. Clearly, this issue needs additional study.

With regard to the relationship between leadership and team norms, it appears that team norms sanctioning cheating and aggression may be facilitated both by a leadership style (as perceived by athletes) that combines high levels of autocratic behavior and social support, and also by a preference for that same style of leadership. It is not surprising that autocratic behavior is related to the team norms under consideration, since autocratic behavior tends to reduce the likelihood of independent, critical thinking on the part of players, subordinating it to strategic considerations.

The social support finding, however, is surprising. Perhaps the explanation is that social support behaviors are independent of any performance criteria. Thus, social support may communicate a nonevaluative stance on the part of the coach, which may be perceived to extend to unfair game tactics. Thus, if a coach autocratically points to the importance of winning while simultaneously being supportive of her or his athletes regardless of the tactics they use to win, team norms may be more likely to condone cheating and aggression. This speculative conjecture needs further substantiation.

334 Shields, Bredemeier, Gardner, and Bostrom

With the coach's version of the LSS, both training and instruction and positive feedback were also related to the team norms. Apparently, the coach's belief that he or she was offering training and instruction as well as performance- contingent positive feedback (but not the athletes' perceptions of such) was related to acceptance of cheating and aggression. Once again, further investigation of these relationships is warranted.

Team cohesion, particularly task cohesion, was positively related to expecta- tions that peers would cheat and aggress, and that the coach would condone cheating. Task cohesion may be facilitated by a coherent and widely shared set of team priorities focused on obtaining victory, and such a value set lends itself to embracing unfair game tactics. Nonetheless, the present study did not support the contention that close agreement on team norms pertaining to cheating and aggression would itself be related to team cohesion. Higher levels of task cohesion may facilitate a subordination of fair play to victory, but discrepant viewpoints on the issue remain.

Future studies are needed to replicate and refine the findings of this study before any conclusion is accepted as established. In particular, the Team Norm Questionnaire needs to be further developed and its validity and reliability as- sessed. Nonetheless, the findings presented here are suggestive of important directions for future research into the influences on critical dimensions of team moral atmosphere. Should future research substantiate the important role of the coach in relation to transgressive moral norms, it will become incumbent upon sport scientists to help show how coaches can deter the development of norms that support cheating and aggression.

References

Blinde, E.M. (1989). Participation in a male sport model and the value alienation of female intercollegiate athletes. Sociology of Sport Journal, 6, 36-49.

Bredemeier, B.J. (1985). Moral reasoning and the perceived legitimacy of intentionally injurious sport acts. Journal of Sport Psychology, 7 , 110-124.

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

Bredemeier, B.J. (in press). Divergence in children's moral reasoning about sport and everyday life. International Journal of Sport Psychology.

Bredemeier, B.J., & Shields, D.L. (1984a). Divergence in moral reasoning about sport and life. Sociology of Sport Journal, 1, 348-357.

Bredemeier, B.J., & Shields, D.L. (1984b). The utility of moral stage analysis in the investigation of athletic aggression. Sociology of Sport Journal, 1, 138-149.

Bredemeier, B.J., Weiss, M.R., Shields, D.L., & Cooper, B. (1986). The relationship of sport involvement with children's moral reasoning and aggression tendencies. Jour- nal of Sport Psychology, 8, 304-318.

Bredemeier, B.J., Weiss, M.R., Shields, D.L., & Cooper, B. (1987). The relationship between children's legitimacy judgments and their moral reasoning, aggression tendencies, and sport involvement. Sociology of Sport Journal, 4 , 48-60.

Carron, A.V., Widmeyer, N., & Brawley, L. (1985). The development of an instrument to assess cohesion in sport teams: The Group Environment Questionnaire. Journal of Sport & Exercise Psychology, 10, 244-266.

Cheating and Aggression Norms 335

Chelladurai, P. (1984). Discrepancy scores between preferences and perceptions of leader- ship behavior and satisfaction of athletes in varying sports. Journal of Sport Psychol- ogy, 6, 27-41.

Chelladurai, P., & Carron, A.V. (1978). Leadership CAHPER. Sociology of Sport Mono- graph Series A. Calgary, AB: University of Calgary.

Chelladurai, P., & Carron, A.V. (1983). Athletic maturity and preferred leadership. Journal of Sport Psychology, 5, 371-380.

Chelladurai, P., Imamura, H., Yamaguchi, Y., Oinuma, Y., & Miyauchi, T. (1988). Sport leadership in a cross-national setting: The case of Japanese and Canadian university athletes. Journal of Sport & Exercise Psychology, 10, 374-389.

Chelladurai, P., & Saleh, S.D. (1980). Dimensions of leader behavior in sports: Develop- ment of a leadership scale. Journal of Sport Psychology, 2, 34-45.

Coakley, J. J. (1994). Sport in society: Issues and controversies (5th ed.). St. Louis: Mosby.

Donnelly, P., & Young, K.M. (1985). Reproduction and transformation of cultural forms in sport: A contextual analysis of rugby. International Review for the Sociology of Sport, 20, 19-38.

Duquin, M.E. (1984). Power and authority: Moral consensus and conformity in sport. International Review of Sociology of Sport, 19, 295-303.

Dwyer, J.J.M., & Fischer, D.G. (1988). Psychometric properties of the coach's version of the Leadership Scale for Sports. Perceptual and Motor Skills, 67, 795-798.

Dwyer, J.M., & Fischer, D.G. (1990). Wrestlers' perceptions of coaches' leadership as predictors of satisfaction with leadership. Perceptual and Motor Skills, 71, 51 1- 517.

Fine, G.A. (1987). With the boys: Little League baseball and preadolescent culture. Chicago: University of Chicago Press.

Gardner,D.E., Shields, D.L.L., Bredemeier, B.J.L., & Bostrom, A. (1995). The relationship between perceived coaching behaviors and team cohesion among baseball and sofball players. Manuscript submitted for publication.

Gordon, A.M.D. (1986). Behavioral correlates of coaching effectiveness. Unpublished doctoral dissertation, University of Alberta.

Haan, N., Aerts, E. & Cooper, B.B. (1985). On moral grounds: The search for apractical morality. New York: New York University.

Home, T., & Carron, A.V. (1985). Compatibility in coach-athlete relationships. Journal of Sport Psychology, 7, 137-149.

Knoppers, A. (1985). Professionalization of attitudes: A review and critique. Quest, 37, 92-102.

Kohn, A. (1986). No contest: The case against competition. Boston: Houghton Mifflin. McMillin, C.J. (1990). The relationship of athlete self-perceptions and athlete perceptions

of leader behaviors to athlete satisfaction. Unpublished doctoral dissertation, Uni- versity of Virginia.

Nixon, H.L. (1992). A social network analysis of influences on athletes to play with pain and injuries. Journal of Sport and Social Issues, 16, 127-135.

Nixon, H.L. (1994). Coaches' views of risk, pain, and injury in sport, with special reference to gender differences. Sociology of Sport Journal, 11, 79-87.

Orlick, T. (1978). Winning through cooperation: Competitive insanity, cooperative alterna- tives. Washington, DC: Acropolis Press.

Pedhazur, E.J. (1982). Multiple regression and behavioral research. New York: Holt, Rinehart, & Winston.

336 Shields, Bredemeier, Gardner, and Bostrorn

Power, F.C., Higgins, A., & Kohlberg, L. (1989). Lawrence Kohlberg's approach to moral education. New York: Columbia University Press.

Schliesman, E. (1987). Relationship between the congruence of preferred and actual leader behavior and subordinate satisfaction with leadership. Journal of Sport Behavior, 10, 157-166.

Serpa, S., Pataco, V., & Santos, F. (1991). Leadership patterns in handball international competition. International Journal of Sport Psychology, 22, 78-89.

Shields, D.L.L., & Bredemeier, B.J.L. (1995). Character development andphysical activ- ity. Champaign, IL: Human Kinetics.

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

Smith, M.D. (1976). The legitimation of violence: Hockey players' perceptions of their reference groups' sanctions for assault. Canadian Review of Sociology and Anthro- pology, 12, 72-80.

Smith, M.D. (1983). Violence and sport. Toronto: Buttenvorths. Stephens, D., & Bredemeier, B.J.L. (1995). Toward an understanding of moral behavior

in sport: The relationships among motivational variables and judgments about unfair play in girls' soccer. Manuscript submitted for publication.

Terry, P. (1984). The coaching preferences of elite athletes competing at Universiade '83. Canadian Journal of Applied Sport Sciences, 9, 201-208.

Webb, H. (1969). Professionalization of attitudes toward play among adolescents. In G. Kenyon (Ed.), Aspects of contemporary sport sociology (pp. 161-179). Chicago: The Athletic Institute.

Weiss, M.R., & Friedrichs, W.D. (1986). The influence of leader behaviors, coach attri- butes, and institutional variables on performance and satisfaction of collegiate basketball teams. Journal of Sport Psychology, 8, 332-346.

Westre, K., & Weiss, M. (1991). The relationship between perceived coaching behaviors and group cohesion in high school football teams. The Sport Psychologist, 5 , 41- 54.

Williams, T., & Donnelly, P. (1985). Subcultural production, reproduction and transforma- tion in climbing. International Review for the Sociology of Sport, 20, 3-16.