Hockey coaches' and players' perceptions of aggression and the aggressive behavior of players.

Loughead, Todd M. ; Leith, Larry M.

The general purpose of this study was to examine minor hockey league coaches' and players 'perceptions of the prevalence of aggression and the aggressive behavior of the players. More specifically, there were three main purposes in the present study. The first purpose was to examine the relationship between level of play and aggression. The second purpose was to examine the relationship between minor hockey league coaches' and players' perceptions of instrumental and hostile aggression, and the observed aggression on the ice. The final purpose was to explore the relationship between instrumental and hostile aggression for coaches, players, and observed aggression. Hockey players at the Atom (i.e., 10 to 11 yrs.), Peewee (i.e., 12 to 13 yrs.), and Bantam (i.e., 14 to 15 yrs.) levels and their coaches completed a modified version of the Bredemeier Athletic Aggression Inventory-Short Form (BAAGI-S). Observed aggression was operationally defined through penalties recorded on game sheets. Players on Atom teams wer e more approving of instrumental aggression, while Peewee/Bantam players favored the use of hostile aggression and received more hostile aggression penalties. On the other hand, both Atom and Peewee/Bantam coaches endorsed instrumental aggression more than hostile aggression. Regardless of the level of play, players' views were unrelated to coaches' views on aggression. In addition, players viewed instrumental aggression as more acceptable, but showed a tendency to behave otherwise as reflected in more hostile-type penalties being observed.

As Widmeyer and Birch (1984) have noted "the world of sport, like other aspects of society, is not without aggressive behavior" (P. 77). Since ice hockey is played at high speeds in a confined space and allows for body contact, it is not surprising that it is susceptible to aggressive behaviors (Widmeyer & McGuire, 1997). Numerous empirical studies have been conducted over the last 30 years on hockey violence (Bloom & Smith, 1996; Colburn, 1985; Cote, Trudel, Bernard, Boileau, & Marcotte, 1993; Faulkner, 1974; Smith, 1979a, 1979b, 1979c; Trudel, Guertin, Bernard, Boileau, & Marcotte, 1991; Widmeyer & McGuire, 1997).

Two kinds of aggression have been identified in the sport psychology literature: instrumental aggression and hostile aggression (Cox, 1994; Cratty, 1989). In both types of aggression, the intent is to harm a target person either physically and/or psychologically. Instrumental aggression serves as means to a particular goal. For example, in baseball a pitcher may throw a high inside fastball at a batter to insure he/she and subsequent batters stay well back from the plate. The injury that results is impersonal and simply designed to limit the effectiveness of the opponent (Russell, 1993).

On the other hand, the primary objective of hostile aggression is to inflict an injury on a target person. The intent is to make the person suffer, either physically and/or psychologically (Cox, 1994). If a baseball pitcher threw a high inside fastball at a batter simply to inflict an injury-possibly because a previous batter had hit a home run-the intent would be to injure and cause suffering.

In an attempt to provide insight into the factors associated with hockey violence, one general approach has been to identify the correlates of aggression. For instance, increases in aggression have been shown to be positively associated with increases in score differential (Cote et al., 1993), increases in the frequency of competition (Widmeyer & McGuire, 1997), and increased levels of masculinity (Weinstein, Smith, & Wiesenthal, 1995).

A second approach has been to focus on what is referred to as the reference other (Smith, 1979c). A person or group who directly influences an individual's perceptions or actions is a reference other (Morra & Smith, 1995). One of the most important reference other for young hockey players are coaches. It is suggested that perceptions of coach approval can influence athlete behavior (Smith, 1979b). As a consequence, coaches play a substantial role in athletes' views on the acceptability of behaviors including aggression. For example, Smith (1975) found the qualities a minor hockey league coach looked for in selecting a player included "being aggressive all the time", "possessing physical size and strength", and "having guts and courage". Similarly, Vaz (1982) concluded that players consider technical skills and the ability to be continually aggressive as fundamental qualities that coaches consider essential to the game of hockey.

Luxbacher (1986) found that coach's attitudes are related to the amount of aggression shown by soccer players. Specifically, players who perceived their coach to have a win-at-all-costs attitude expressed significantly higher levels of aggression, were more inclined to use illegal tactics, and displayed a more professional attitude towards competition.

Although a large body of research has been generated from a reference other perspective, individual behavior is not solely a product of situational exigencies. Thus, aggression should also be examined directly from both a coach's and a player's viewpoint. Also, while two types of aggression are theoretically present in sport (i.e., hostile and instrumental), little empirical research has examined the prevalence of these two types of aggression in players and coaches. The present study was undertaken to determine if coaches and players are similar in their perceptions of aggression, and if these perceptions are related to players' behavior on the playing surface.

The first of three main purposes was to examine the influence of level of play (i.e., reflected in increasing age and experience in hockey) on hostile and instrumental aggression of players and coaches. Research examining the relationship between level of play and measures of aggression such as fighting (Bloom & Smith, 1996) and total penalty minutes (Weinstein et al., 1995) have shown that as level of play increases (i.e., older players) acts of aggression also increase. Therefore, it was predicted that as age and experience increased (i.e., from Atom to Peewee to Bantam), approval of and engagement in hostile and instrumental aggression would also increase for both players and coaches.

The second main purpose was to examine the relationships between coaches' and players' perceptions toward hostile and instrumental aggression and observed aggression. Research by Luxbacher (1986) and by Stephens and Bredemeier (1996) have shown soccer players' attitudes toward aggression are related to those of their coach. Little empirical research examining this relationship in a hockey context has been conducted. Consequently, a positive relationship was predicted for the present study.

The final purpose was to explore the relationships between perceptions of (i.e., approval of) instrumental and hostile aggression in coaches and players, and the observed aggressive behavior of players. Given the general nature of previous research, a positive relationship between the perceptions of both players and coaches and the observed hostile and instrumental aggression of athletes was predicted.

A number of authors have pointed out that social phenomena can be examined at different levels-the individual group-member level, the aggregated group level, or the intact group level (e.g., Cota, Longman, Evans, & Dion, 1995; Paskevich, 1995). As Carron and Spink(l995) pointed out, the most appropriate level "depends upon the nature of the question" (p.91). The fundamental question in the present study concerned whether a reference other (cf. Smith, 1979c)-operationally defined as coaches in the present study-plays an important role in influencing the attitudes and behaviors of a target group-operationally defined as the players on their team. In short, the research question represented a group-level question and an aggregated group level approach was used in the analyses-the scores of individual players on a team were combined to reflect a "team perspective".

Method

Participants

Questionnaires were distributed to 30 competitive level minor male hockey teams at the Atom (10-11 years old), Peewee (12-13 years old), and Bantam (14-15 years old) levels. One major subdivision in minor hockey is house league versus competitive. The competitive level was targeted in the present study for two reasons. First, body checking is not permitted at any age in the house league system; a restriction that is not present in competitive hockey at Atom, Peewee, and Bantam. Second, the primary focus in house league is mass participation; in competitive hockey, players compete for positions on a team as well as ice time within teams.

Responses were obtained from 30 head coaches (range in age of 29-52 years, M = 42yrs., SD = 6.8), and 171 athletes (a rate of return of 38%). It was intended initially to statistically consider the three levels of experience (i.e., Atom, Peewee, and Bantam) as independent factors. However, given a relatively reduced number of responses at the Peewee and Bantam levels, a decision was made to group these two together. This produced 15 teams (a total of 74 athletes) at the Atom level and 15 teams at the Peewee and Bantam level (a total of 97 athletes).

Measures

Perceived aggression. An adapted version of the Bredemeier Athletic Aggression Inventory-Short Form (BAAGI-S) was used to assess perceived aggression for both coaches and players. The BAAGI-S distinguishes between two types of aggression: hostile (has the goal of inflicting injury on another person) and instrumental (has the goal of achieving a particular outcome such as winning, money, prestige, or recognition). The BAAGI-S contains 30 items (15 for each of hostile and instrumental). Responses are provided on a 4-point Likert scale containing the prompts strong agreement (1), agreement (2), disagreement (3), or strong disagreement (4). Thus, higher scores indicate lower levels of aggression.

The BAAGI-S has been administered primarily to older elite level athletes. Consequently, an adapted version of the BAAGI-S was required for both minor hockey coaches and players. A panel of three minor hockey coaches and a linguistics expert made the modifications to the BAAGI-S. For example, a typical question in the original version is "I am usually calm and poised before participating in an athletic event". When revised for the coach's version, the question became: "I am usually calm and poised before coaching an athletic event." When the same original question was revised for minor hockey players, it read: "I am usually relaxed before a game".

Observed aggression. A measure of actual aggressive behavior was obtained using penalties from game summary sheets (i.e., all penalties awarded by the referee are specifically listed). A typology advanced by McCarthy and Kelly's (1978) was used to categorize penalties as either hostile or instrumental in nature. Hostile aggressive penalties included elbowing, slashing, boarding, roughing, cross-checking, high-sticking, butt-ending, fighting, spearing, kneeing, checking from behind, charging, and face masking. Instrumental aggressive penalties included tripping, hooking, holding, and interference. Penalties considered neither instrumental nor hostile (and, therefore, not used in the analyses) included delay of game, too many men on the ice, bench penalty, playing with a broken stick, and unsporting conduct.

Within the rules of hockey, the punishment (in terms of time spent in the penalty box) for different infractions varies from two minutes to five and ten minutes. As a consequence, total penalty minutes would represent an inappropriate (i.e. misleading) dependent variable (McCarthy & Kelly, 1978). Therefore, using the approach of McCarthy and Kelly (1978) the total number of penalties within each of the two categories (hostile, instrumental) was used.

Since referees are rotated randomly through the various leagues and age levels, it was assumed that errors in observation and interpretation were distributed evenly for all teams (Widmeyer & McGuire, 1997). In fact, Audette, Trudel, and Bernard (1994) reported that the change in the number of different types of penalties assessed by National Hockey League (NHL) referees from the 1978 season until the end of the 1988 season remained unchanged. In addition, Katorji and Cahoon (1992) found that only 4.8% of aggressive acts seen by two independent observers went unnoticed by referees in Junior B hockey games. Therefore, it was assumed that the fact that some aggressive acts go unnoticed would have only a minor impact on the examination of aggressive behavior as assessed through the number of penalties (Widmeyer & McGuire, 1997).

Procedure

Hockey teams were contacted near the end of their competitive season to obtain permission to test the players and coaches, and to obtain game summary sheets. The researcher and research assistants met with the participating teams. Coaches completed the questionnaire on-site and submitted their game summary sheets. Each player received a package to take home. It included a questionnaire, a parent consent form, an athlete assent form and a stamped envelope with the senior researcher's address printed on the front of the envelope. Players were instructed to complete their questionnaires independently and return them to the researcher. All participants were assured that their participation was voluntary and that responses would be kept confidential.

Results and Discussion

Level of Play and Aggression

Descriptive statistics for the variables of interest are presented in Table 1 for the total sample and in Table 2 for the subsamples divided on the basis of level of play. Prior to testing the specific predictions, a one-way MANOVA was computed using the 6 manifestations of aggression as dependent variables (coach's perception of instrumental and hostile aggression, player's perception of instrumental and hostile and aggression, and player's observed instrumental and hostile aggression). As indicated above, the player data were aggregated; as a consequence the number of cases was 30. The results from the MANOVA showed a significant effect for aggression (F (6, 24) = 3070.04, p [less than].05).

To determine if there were differences in aggression on the basis of level of play, independent t tests were conducted on each of the aggression types. Since multiple tests were computed, a criterion level of .05 with Bonferroni correction was used. The Bonferroni correction controls the overall error rate by setting the error rate for each test to the experimentwise error rate divided by the total number of tests. Hence, the observed significance level is adjusted for the fact that multiple comparisons are made (Cone & Foster, 1993).

The results showed that the Atom and Peewee/Bantam teams differed in observed hostile (t(28) = -4.76,p[less than].008) and instrumental aggression (t(28) = -3. 14,p[less than].008). Peewee/Bantam players received more penalties, both instrumental and hostile, than their Atom counterparts. For players' perceived aggression, a statistically significant difference was present for both hostile (t(28) = 3.00,p[less than].008) and instrumental aggression (t(28) = -4.80, p [less than].008) in relation to level of play. In terms of hostile aggression, the older Peewee/Bantam players were more approving than Atom players. On the other hand, Atom players' were more approving of instrumental aggression than Peewee/Bantam players. However, no differences were present between Atom and Peewee/Bantam coaches' in either perceived hostile aggression (t(28) = -.98, p = ns) or perceived instrumental aggression (t(28) = .65, p = ns).

Congruent with previous research and the predictions, both perceived and observed aggression in players increased with advancing levels of play. The Peewee/Bantam players viewed hostile aggression as more acceptable than the younger Atom players. However, the finding of higher approval for perceived instrumental aggression at the Atom level was unexpected. As players proceed through minor hockey, they realize what is required to become upwardly mobile. Players quickly learn what is required for them to continue in hockey (Smith, 1979b). At the Atom level, the focus is on teaching and development of skills. Players at this level are learning the fundamentals of the game. It would appear that Atom players use instrumental aggression more than Peewee/Bantam players to compensate for their lack of skill. Nonetheless, as players' progress through minor hockey, they quickly learn what it takes to advance through the minor hockey system (Smith, 1979b). If young athletes are to succeed in hockey, they are expected t o demonstrate potential no later than Peewee or Bantam. It is at this level that the criteria for player evaluation changes (Vaz, 1976). Players are expected to be continuously aggressive, and a greater emphasis is placed on winning.

As level of play increased, higher amounts of approval of hostile and instrumental aggression from the coaches were predicted. However, regardless of level of play, coaches tend to exhibit the same levels of perceived aggression for both hostile and instrumental aggression types. Nearly 75% of minor hockey coaches are former players (Spallanzani, 1988), and have progressed through a minor hockey league system characterized by an occupational subculture based on a theme of violence (Smith, 1979a). As former minor hockey league players, coaches bring the occupational standard with them, and begin looking for players who can withstand and produce illegal physical coercion (Smith, 1979b, 1979c).

Hostile and Instrumental Aggression for Coaches, Players, and Observed Aggression

Intercorrelations among the variables are presented in Table 3. As these results show, players' perceived hostile aggression and observed hostile aggression were positively related (r = .56, p [less than].003). Neither the relationship between coaches' and players perceived hostile aggression (r = -.19, p = ns) nor coaches' perceived and players observed hostile aggression (r = -.l4, p = ns) was statistically significant.

Players' perceived instrumental aggression and observed instrumental aggression were negatively related (r = -.69, p [less than].003). However, no significant relationship was present for the perceived instrumental aggression scores of coaches and players (r = -.15, p = ns). Similarly, the relationship between coaches' instrumental aggression scores and observed instrumental aggression scores was not significant (r = -.16, p = ns).

A positive relationship among coach perceived, player perceived, and player observed hostile and instrumental aggression was predicted. The results revealed that only players' perceived and observed aggression were significantly correlated for both instrumental and hostile aggression. For hostile aggression, the more players approved of hostile aggression, the more hostile type penalties they received. This result may not be surprising. Smith (1979a) found that hockey players who fought more and received more major penalties displayed a more proviolence value-attitude pattern than nonviolent players and nonplayers. Smith interpreted his results as supportive of a violent subculture hypothesis in hockey (i.e., violent individuals put significantly greater importance than nonviolent individuals on values and attitudes supportive of violence).

Interestingly, a contrary pattern of results was obtained for instrumental aggression. That is, players who indicated lesser amounts of approval for instrumental aggression received more instrumental type penalties. This is best explained by what is known as the instrumental orientation of players (Colburn, 1985). Previous research by Faulkner (1974) and Smith (1979a), demonstrated that illegal infractions in hockey sometimes reflect an occupationally directed and controlled method of achieving an approved end (e.g., winning the game, career advancement). As Colburn (1985) noted, violence in hockey can be referred to as instrumental for the reason that it is engaged in by a player not as an end in itself, but rather as a means to the realization of some other end. Furthermore, Colburn suggested that players engage in such activity because "what is at stake in their conduct are such social goods as self-esteem, honor and respect which are intrinsically, if not extrinsically, rewarding and valuable" (p.154). A s a result, players do not perceive instrumental behaviors as aggressive since this type of aggression is associated with values that are socially desirable.

It was also predicted that players' endorsement of aggression would be related to that of their coach. The results did not support this hypothesis. This result was unexpected since previous findings have shown that an athlete's perception of their coach's ego orientation was a predictor of the likelihood to aggress (Stephens & Bredemeier, 1996).

Finally, it was predicted that coaches' perceived aggression would be associated with player observed aggression. Again, the results did not support this hypothesis. The results reflect the fact that coaches-although highly influential--are only one of many influences on players' perceptions and behaviors of aggression. Besides coaches, Smith (1979c) and Morra and Smith (1995) found that parents, teammates, spectators, and the media help perpetuate that aggression and violence are acceptable and necessary elements of the game of hockey.

Differences Between Manifestations of Aggression

A number of paired t tests were conducted to examine for differences in instrumental versus hostile aggression insofar as coaches' perceptions, players' perceptions', and players' behavior is concerned. Results showed that coaches (t(29) = 11.50, p[less than].017) and players (t(29) = 4.30, p[less than].017) showed greater approval for instrumental aggression. However, players displayed a greater amount of observed hostile aggression (t(29) = 8.10, p[less than].017).

The results provide support for the suggestion that minor hockey league coaches and players are less inclined to endorse aggression for its own sake than as a means to an end. It is possible that there is an overemphasis placed on winning within the culture of competitive hockey. As a player progresses through the minor hockey system, the value of success, (i.e., winning), takes precedence (Vaz, 1976). It could be argued that players do not view some of their perceptions as instrumental aggression since it is associated with values that are socially desirable (i.e., respect, honor)(Colburn, 1985).

It seems likely that coaches show greater approval for instrumental aggression because they have learned that it is advantageous to do so (Bandura, 1973). Smith (1979b) noted that hockey coaches encourage the use of aggressive play for what it symbolizes (character development), and for its utility in winning hockey games. In addition, Trudel, Dionne, and Bernard (1992) found Bantam hockey coaches approved of penalties if they helped prevent goals. The results from Smith (I 979b) and Trudel et al. (1992) could be considered forms of instrumental aggression, as they are perceived to be necessary elements in winning hockey games.

Although coaches and players indicated a higher approval for instrumental aggression, the observed on-ice behavior told a different story. There were almost twice as many hostile penalties received as instrumental penalties. This result is difficult to explain, since players were less approving of hostile aggression. It would appear that a player's perception is one thing, but what actually occurs on the ice is another. It would appear that for many young hockey players, the importance of earning the respect of fellow players is crucial (Colburn, 1985; Faulkner, 1974; Smith, I 979b; Vaz, 1976). Courage, toughness, and a willingness to fight are examples of ways in which players earn respect. Respect is lost if a player backs down, shows lack of heart, "no guts", and/or "chickens out" (Morra & Smith, 1995). Players who show a reputation for being tough, regardless of how they feel about aggression, are the ones who demonstrate the characteristics that the game of hockey reinforces.

The results of the present study provide support for the suggestion that minor hockey league coaches, regardless of level of play, are similar in their perceptions towards both types of aggression. Furthermore, as players advance through the minor hockey system, they develop a more approving perception towards hostile aggression, which in turn leads to more hostile types of penalties. Finally, both coaches and players viewed instrumental aggression as a more acceptable behavior than hostile aggression. However, the behavior on the ice is far different with players receiving more hostile aggression penalties than instrumental aggression penalties. It would be fruitful to examine higher levels of hockey (i.e.,junior and professional ranks) to determine if players and coaches at these levels display similar perceptions and behaviors towards instrumental and hostile aggression. If the results of the present study are any indication then it would be expected that players and coaches in senior levels of play would endorse both instrumental and hostile aggression more than their younger, less experience counterparts.

Future research should attempt to determine the impact of leagues where specific rules pertaining to fair play have been implemented. For example, teams receive points based not only on the outcome of the game but also on their behavior. Each team begins the game with six behavior points, and each minor and major penalty results in the loss of one point until the team exceeds the limit determined by league administrators (Gilbert, Trudel, & Bloom, 1995). It could be argued that fewer penalties, both instrumental and hostile, would be observed in fair play leagues, and coaches' and players' perceptions towards aggression would differ from those in the more traditional leagues.

Finally, future research is required to determine the contrast between players positive perception towards instrumental aggression while engaging in more hostile type of behaviors on the playing surface. One approach would be to conduct in-depth interviews with players to determine their intentions during certain game situations. It could be argued that players had instrumental aggression intentions however received a hostile penalty as called by the referee.

Authors' Acknowledgement

The authors would like to thank Bernard Legare and Jennifer Robertson for their assistance with data collection. The authors would also like to express their gratitude to Albert V. Carron for his helpful suggestions and comments.

References

Audette, S. M., Trudel, P., & Bernard, D. (1994). L'utilisation d'etudes d'archives pour l'analyse d'actes deviants au hockey sur glace. CAHPER/ACSEPL Research Supplement, I, 129-139.

Bandura, A. (1973). Aggression: A social learning analysis. Englewood Cliffs, NJ: Prentice-Hall.

Bloom, G. A., & Smith, M. D. (1996). Hockey violence: A test of cultural spillover theory. Sociology of Sport Journal, 13 (1), 65-77.

Carron, A. V., & Spink, K. S. (1995). The group size-cohesion relationship in minimal groups. Small Group Research, 26, 86-105.

Colburn, K. (1985). Honor, ritual and violence in hockey. Canadian Journal of Sociology 10, 153-171.

Cone, J. D., & Foster, S. L. (1993). Dissertations and theses: From start to finish. Washington, DC: American Psychological Association.

Cota, A. A., Longman, R. S., Evans, C. R., & Dion, K. L. (1995). Using and misuing factor analysis to explore group cohesion. Journal of Clinical Psychology, 51, 308-316.

Cote, J., Trudel, P., Bernard, D., Boileau, R., & Marcotte, G. (1993). Observation of coach behaviors during different game score differentials. In C. R. Castaldi, P. J. Bishop, & E. F. Homer (Eds.), Safety in ice hockey: Second volume (pp. 78-87). Philadelphia: American Society for Testing and Materials.

Cox, R. H. (1994). Sport psychology: Concepts and applications. (3rd ed.) Madison, WI: Brown and Benchmark.

Cratty, B. J. (1989). Psychology in contemporary sport. Englewood Cliffs, NJ: Prentice-Hall.

Faulkner, R. (1974). Making violence by doing work. Sociology of Work and Occupations, 1, 288-312.

Gilbert, W. D., Trudel, P., & Bloom, G. A. (1995). Intramural ice hockey officiating: A case study. Avante, 1, 63 - 76.

Katorji, K., & Cahoon, M. A. (1992). The relationship between aggression and injury in Junior "B" hockey, Unpublished manuscript, University of Waterloo, Waterloo, Ontario, Canada.

Luxbacher, J. (1986). Violence in sport: An examination of the theories of aggression, and how the coach can influence the degree of violence in sport. Coaching Review 9, 14-17.

McCarthy, J. F., & Kelly, B. R. (1978). Aggressive behavior and its effect on performance over time in ice hockey athletes: An archival study. International Journal of Sport Psychology, 9, 90-96.

Morra, N., & Smith, M. D. (1995). Interpersonal sources of violence in hockey: The influence of the media, parents, coaches and game officials. In R. E. Smith & F. L. Smoll (Eds.), Children and youth sport: A biopsychosocial perspective (pp. 142-155). Madison, WI: Brown & Benchmark.

Paskeich, D. M. (1995). Conceptual And Measurement Factors of Collective Efficacy in its Relationship to Cohesion and Performance Outcome. Unpublished doctoral dissertation, University of Waterloo, Waterloo, Ontario.

Russell, G. W. (1993). The social psychology of sport. New York: Springer-Verlag.

Smith, M. D. (1975). The legitimation of violence: Hockey players' perceptions of their reference groups' sanctions for assaults. Canadian Review of Sociology and Anthropology, 12, 72-80.

Smith, M. D. (1979a). Hockey violence: A test of the violent subculture hypothesis. Social Problems, 27, 235-247.

Smith, M. D. (1979b). Social determinants of violence in hockey. Canadian Journal of Applied Sport Sciences, 4 (1), 76-82.

Smith, M. D. (1979c). Towards an explanation of hockey violence: A reference other approach. Canadian Journal of Sociology, 4, 105-124.

Spallanzani, C. (1988). Profil d'entraineurs en hockey mineur et motifs de participation et de demission. Canadian Journal of Sport Science, 13 (2), 157-165.

Stephens, D. E., & Bredemeier, B. J. (1996). Moral atmosphere and judgments about aggression in girls' soccer: Relationships among moral and motivational variables. Journal of Sport and Exercise Psychology, 18, 158-173.

Trudel, P., Guertin, D., Bernard, D., Boileau, R., & Marcotte, G. (1991). Analyse des comportements de l'entraineur par rapport a la violence au hockey mineur. Canadian Journal of Sport Sciences, 16(2), 103-109.

Trudel, P., Dionne, J-P., & Bernard, D. (1992). Etude qualitative de la violence au hockey: Perceptions d'entraineurs et des joueurs. Canadian Journal of Sport Science, 17, 320-332.

Vaz, E. (1976). The culture of young hockey players: Some initial observations. In A. Yianmakis, T. D. McIntyre, M. J. Melnick, & D. P. Hart (Eds.), Sport sociology: contemporary themes (pp. 211-215). Dubuque, IA: Kendall-Hunt.

Vaz, E. (1982). The professionalization of young hockey players. Lincoln, NE: University of Nebraska Press.

Weinstein, M., Smith, M. D., & Wiesenthal, D. (1995). Masculinity and hockey violence. Sex Roles, 33, 831-847.

Widmeyer, W. N., & Birch, J. S. (1984). Aggression in professional ice hockey: A strategy for success or a reaction to failure? The Journal of Psychology, 117, 77 - 84.

Widmeyer, W. N., & McGuire, E.J. (1997). Frequency of competition and aggression in professional ice hockey. International Journal of Sport Psychology, 28, 57 - 66. Table 1 Mean Scores and Standard Deviations for All Aggression Variables Variable M SD Coaches' Perceived Hostile 48.40 5.24 Aggression Instrumental Aggression 31.13 5.04 Players' Perceived Hostile 39.10 4.33 Aggression Instrumental Aggression 32.90 4.09 Observed Hostile Aggression 2.14 .90 Instrumental Aggression 1.13 .37 Note. The higher the mean score, the lower the aggression for Coaches and Players. Observed aggression mean scores represent the average number of penalties per game. Table 2 Mean Scores for Hostile and Instrumental Aggression and Level of Play Type of Aggression Level of Play Atom Peewee/Bantam Hostile Coaches' Perceived 47.47 49.33 Players' Perceived 41.20 37.00 Observed 1.55 2.73 Instrumental Coaches' Perceived 31.73 30.53 Players' Perceived 30.20 35.60 Observed 0.94 1.31 Note. The higher the mean score, the lower the aggression for Coaches' Perceived and Players Perceived. Observed aggression mean scores represent the average number of penalties per game. Table 3 Intercorrelations Between Subscales for Hostile and Instrumental Aggression Subscale 1 2 3 Hostile 1. Coaches' Perceived -- -.19 -.14 2. Players' Perceived -- .56 (**) 3. Observed -- Instrumental 1. Coaches' Perceived -- -.15 -.16 2. Players' Perceived -- -.69 (**) 3. Observed -- Note. (**)p[less than].01 with Bonferroni correction. Subscales 1, 2, and 3 refer to coaches' perceived, players' perceived, and observed aggression respectively.