Sports Science Discussion

    747

Journal of Sport & Exercise Psychology, 2007, 29, 747-762  © 2007 Human Kinetics, Inc.

Shields and Bredemeier are with the Center for Character and Citizenship, College of Education, University of Missouri–St. Louis; LaVoi is with the School of Kinesiology, University of Minnesota; and Power is with the Program of Liberal Studies, University of Notre Dame.

Predictors of Poor Sportspersonship in Youth Sports: Personal Attitudes

and Social Influences

David Light Shields,1 Nicole M. LaVoi,2 Brenda Light Bredemeier,1 and F. Clark Power3

1University of Missouri–St. Louis, 2University of Minnesota,  and 3University of Notre Dame

The present study examined personal and social correlates of poor sportspersonship among youth sport participants. Male and female athletes (n = 676) in the fifth through eighth grades from three geographic regions of the U.S. participated in the study. Young athletes involved in basketball, soccer, football, hockey, baseball/ softball, or lacrosse completed a questionnaire that tapped poor sportspersonship behaviors and attitudes, team sportspersonship norms, perceptions of the poor sportspersonship behaviors of coaches and spectators, and the sportspersonship norms of coaches and parents. Preliminary analyses revealed significant gender, grade, sport area, and location differences in self-reported unsportspersonlike behavior. The main analysis revealed that self-reported poor sport behaviors were best predicted by perceived coach and spectator behaviors, followed by team norms, sportspersonship attitudes, and the perceived norms of parents and coaches. Results are discussed in relation to the concept of moral atmosphere.

Key Words: sportspersonship, spectators, coaches, parents, moral atmosphere

Good sport behavior is probably the norm in most youth sport contexts. And yet, empirical evidence also indicates that aggression, cheating, and other poor sport behaviors occur with alarming frequency (Kavussanu, 2006; Shields, Bredemeier, LaVoi, & Power, 2005; Weinstein, Smith, & Wiesenthal, 1995). As far back as 1976, Martens inquired whether youth sports are “a den of iniquity or a land of promise” (p. 201). If poor sport behaviors are a problem in youth sports, what are the likely causes? The present study investigated potential predictors of poor sport behavior by participants in youth sports.

Whereas the term sportspersonship is often used more broadly (e.g., Valle- rand, Briere, Blanchard, & Provencher, 1997), we are using it here to refer to sport behaviors that carry moral connotations because of their connection to fundamental issues of fairness and respect. There are numerous ethically related behaviors that

748 Shields, LaVoi, Bredemeier, and Power

may be cause for concern among young sport participants, including cheating, fighting, retribution, arguing with officials, trash talking, teasing, bragging, and simply acting like a “bad sport” following a loss.

Researchers have identified a number of important social influences on sports- personlike play. Nicholls’s achievement goal theory has provided the theoretical foundation for many of these studies. According to Nicholls (1989), one reason that individuals engage in achievement contexts like sport is to develop and demonstrate competence. Competence, in turn, can be understood and evaluated in a manner that is primarily self-referenced (called a task orientation) or primarily other-referenced (called an ego orientation). Nicholls (1989) hypothesized that task orientation would relate positively to such moral values as fairness and respect, whereas those who are ego-involved would tend to subordinate moral concerns to the desire to demonstrate superiority. The research has generally supported these contentions (e.g., Duda, Olson, & Templin, 1991; Dunn & Causgrove Dunn, 1999; Lemyre, Roberts, & Ommundsen, 2002; Sage & Kavussanu, 2007; Sage, Kavussanu, & Duda, 2006). Additionally, the social goals of friendship, peer acceptance, and coach praise have been found to predict intention to engage in unsportspersonlike behavior (Stuntz & Weiss, 2003). Situational influences on task and ego motivation, referred to as motivational climate variables (Ames, 1992), have also been shown to predict sportspersonship attitudes, intentions, or behaviors (Miller, Roberts, & Ommundsen, 2005; Ommundsen, Roberts, Lemyre, & Treasure, 2003; Gano- Overway, Guivernau, Magyar, Waldron, & Ewing, 2005).

The present investigation is grounded theoretically in the constructivist psy- chological paradigm (e.g., Kohlberg, 1984, and Piaget, 1948). Constructivists hold that people possess agency and do not simply respond in predictable ways to environmental influences and contingencies. From this perspective, behavior invariably reflects an interaction between the interpreting, meaning-construing person and contextual influences.

The first major variable focusing on the person side of the person–context interaction to be investigated from a constructivist perspective on sportspersonship was moral reasoning stage (Bredemeier & Shields, 1984). From a constructivist perspective, knowing how an actor construes moral meaning in a situation is critical to understanding and predicting the actor’s behavior. Several subsequent studies have confirmed the importance of investigating moral reasoning stage in relation to sportspersonship attitudes, intentions, or behaviors (Bredemeier, 1985, 1994; Brede- meier & Shields, 1986; Bredemeier, Weiss, Shields, & Cooper, 1986, 1987).

Moral reasoning stages are “deep structures” that cannot be used to directly predict specific behavioral choices (Shields & Bredemeier, 1995). Even though moral stages appear to be important contributors to action choices, there are numer- ous mediating processes between moral stages and behaviors, such as a person’s specific attitudes and beliefs (Kohlberg & Candee, 1984; Rest, 1984). Although surprisingly little research has been conducted to determine whether sportsperson- ship attitudes are related to sportspersonship behavior, this appears likely (Stornes, 2001; Vallerand et al., 1997). In the present study, we focus on sportspersonship attitudes as potential predictors of sportspersonship behaviors.

Moral behaviors, such as sportspersonship, are also influenced by contextual factors. Kohlberg and his colleagues identified the moral atmosphere as a critical influence on people’s moral behavior (Higgins, Power, & Kohlberg, 1984; Power & Reimer, 1978). Moral atmosphere is a broad term that encompasses a number of

Predictors of Youth Sportspersonship    749

distinct aspects of a group or organization’s moral climate or ethos (Power, Higgins, & Kohlberg, 1989). Perhaps the most studied components of the moral atmosphere are the collective norms of the group. A collective norm is “a complex of specific behavioral expectations that share a common value” (Power et al., 1989, p. 115). In sports, shared team moral norms tend to develop over time through interactions between coach and team members.

Shields, Bredemeier, Gardner, and Bostrom (1995) operationalized the con- struct of collective norms by asking sport participants to estimate the number of their teammates likely to engage in targeted behaviors if doing so would help their team win. If, for example, most members of the team said that most of their team- mates would cheat to win, then, it was concluded, there was no strong collective norm on that team against cheating. Conversely, if no one thought any of their teammates would cheat, then the researchers concluded that there was a strong collective norm against cheating. The researchers also asked respondents whether their coaches would expect them to cheat or aggress.

In a study of adolescent soccer players, Guivernau and Duda (2002) used a variation of the approach employed by Shields et al. (1995) and found that the ath- letes’ perceptions of their team’s aggression norms emerged as the most consistent predictor of their self-described likelihood to aggress. Similarly, in a series of stud- ies, Stephens and her colleagues found that collective norms are good predictors of self-described likelihood to aggress among female soccer players (Stephens, 2000; Stephens & Bredemeier, 1996), coed soccer players (Stephens, 2000), youth bas- ketball players (Stephens, 2001), and male adolescent ice hockey players (Stephens & Kavanagh, 2003). Comparing the influence of a team’s collective norms and the perceived performance climate, Kavussanu, Roberts, and Ntoumanis (2002) found that college basketball players’ perceptions of their team’s collective norms had a significant effect on moral functioning, whereas the effect of perceived performance climate was not significant. In contrast, Kavussanu and Spray (2006), in a study of adolescent male soccer players, found both collective norms and perception of a performance climate to correlate with moral functioning variables.

Whereas a team’s collective norms are important influences on how an individ- ual perceives moral expectations within a sport event, a young athlete is also likely to be influenced by other perceived expectations. The coach and other members of the team, such as a player’s best friend, may exercise moral influence apart from just how they influence the team’s collective norms (Guivernau & Duda, 2002). Parents may also exercise moral influence in the sport realm (Guivernau & Duda, 2002; Stuart & Ebbeck, 1995). Guivernau and Duda (2002) examined the potential influence of a number of significant others, including best friends, most popular players, team captains, coaches, and parents. Of these influences, they found the coach to be the most important.

The purpose of the present study was to examine potential predictors of poor sportspersonship behavior, including the respondents’ own sportspersonship attitudes, the perceived collective norms of the team, and other likely sources of social influence. In the third category, we included perceptions of the behavior of coaches and spectators, and perceptions of the sportspersonship norms of coaches and parents. We also sought to expand on previous studies by utilizing a large geographically diverse sample representing multiple sport areas.

For this investigation, we sought to include a diverse set of sports in which possibilities for aggression, cheating, and other forms of poor sport behavior might

750 Shields, LaVoi, Bredemeier, and Power

occur with sufficient frequency. Ideally, the decision of which sports to include would be based on empirical data about the relative frequency with which targeted behaviors occur in the various sports. Lacking such data, however, we chose to focus on medium- and high-contact team sports. Thus, from an original sample (N = 899) that included participants in more than a dozen sports, we focused on those young athletes who participated in the high-contact sports of football and hockey and the medium-contact sports of basketball, soccer, baseball or softball, and lacrosse. In terms of numbers, the largest groups to be excluded from these analyses were those involved in swimming, volleyball, tennis, and track.

Method

Participants

Our investigation focused on young athletes in the fifth through eighth grades—a span selected because it is often the time when school-sponsored competitive sports begin in earnest. A total of 676 young athletes participated in the present investiga- tion, ranging in age from 9 to 15 (M = 12.1 years, SD = 1.14). There were 126 fifth graders, 261 sixth graders, 139 seventh graders, and 148 eighth graders, with two participants neglecting to provide their grade. The athletes were drawn from six sports: basketball (n = 290), soccer (n = 191), baseball/softball (n = 109), football (n = 58), hockey (n = 18), and lacrosse (n = 10). The sample included 381 males and 285 females, with 10 individuals not providing their gender. Approximately 70% of the sample was White/European American, 11% African American, 8% Asian American, 5% Hispanic, and 7% other. For specific analyses, the sample size varied owing to missing data.

The convenience sample was drawn from three geographic regions: the Philadelphia area (n = 99), the South Bend (Indiana) area (n = 323), and the San Francisco–East Bay area (n = 254). All participants were recruited through their schools, with notes sent home inviting participation. All students who participated in a school-sponsored or community-sponsored sport program were invited to participate, and each participant was asked to report on his or her most recent sport experience. The largest group of athletes (n = 341) represented school-sponsored sport teams; 312 represented community-sponsored teams, and 23 did not indicate whether their team was school or community affiliated. Years of involvement in their current sport ranged from <1 to 12 (M = 4.5, SD = 2.6).

Measures

The survey-style assessment questionnaire, available on request from the first author, was developed specifically for this investigation and contained seven sections. The first requested the participants’ age, gender, race/ethnicity, most recent sport played, and years of involvement in that sport. The other sections assessed athlete sportspersonship behaviors (the investigation’s criterion variable), along with the following predictor variables: athletes’ sportspersonship attitudes, their percep- tions of the likely behavior of their teammates (team norms), their perceptions of coach and spectator behavior, and their perceptions of how much their coaches and parents embraced specific moral norms. All items on the survey were designed to

Predictors of Youth Sportspersonship    751

have face validity, and six readers who work professionally with students in the sample’s age range confirmed their face validity.

A number of core themes (cheating, hurting, arguing, and teasing) cut across most or all of the measures. However, we did not seek complete uniformity in item structure or content across the scales, instead preferring to focus on salient issues for each scale. For example, in the measure of sportspersonship attitudes, we asked respondents whether they thought it was okay for fans to boo the other team. However, there was no corresponding item in the self-reported sportspersonship behavior measure simply because we asked the respondents about their behavior as athletes, not their behavior as fans. Thus, measures had different numbers of items with some variation of content. Each specific scale is described below.

Self-Reported Poor Sportspersonship Behavior (Poor Sport Behavior). The survey contained seven items designed to tap self-reported poor sport behavior. The items asked respondents whether they had, in the current season, cheated to help their team win; tried to hurt an opponent to help their team win; tried to get back at an opponent who played dirty; argued with a ref or sport official follow- ing a bad call; said something to anger, hurt, or upset an opponent; made fun of a teammate who was less skilled; or acted like a “bad sport” when their team lost. All behaviors are ones with which young athletes of this age are familiar (cf. Fine, 1987; Stuart, 2003; Stuart & Ebbeck, 1995). Respondents first responded “yes” or “no” and then, if yes, indicated whether they had done so “once or twice,” “a few times,” or “often.” Each item was then given a score on a 4-point Likert scale, with 1 indicating a response of no and 4 indicating a response of often. This athlete behavior scale functioned as the outcome variable in the present investigation.

Athlete Attitudes About Poor Sportspersonship (Poor Sport Attitude). The survey contained nine items designed to tap the athletes’ attitudes about poor sportspersonship behaviors. Respondents were asked, “How much do you agree or disagree with the following statements.” Respondents then rated each statement on a 4-point scale, ranging from strongly disagree (1) to strongly agree (4). The items dealt with trash-talking, rule breaking, arguing with officials by players or coaches; trying to hurt or get back at an opponent; fans booing; making fun of a less skilled teammate; flashy and public celebration; and faking an injury. For example, the statement for rule breaking read, “It is OK to break rules if you can get away with it.”

Predicted Teammate Poor Sportspersonship Behavior (Team Norm). To assess shared norms regarding poor sport behavior, we asked respondents how likely their teammates were to engage in seven unsportspersonlike behaviors (cf. Shields et al., 1995). Specifically, respondents were asked how many of their team- mates would cheat or hurt an opponent if it would help their team win; get back at an opponent who plays dirty; argue with a sport official following a bad call; say things to hurt, anger, or upset an opponent; make fun of a less skilled teammate; or act like bad sports when the team lost. Response options on the 4-point scale ranged from none (1) to most (4).

Perceived Poor Sportspersonship Behavior of the Coach (Coach’s Poor Sport Behavior). The poor sportspersonship behavior of coaches was assessed via 10 items. The items asked the respondent whether their coach, in the most recently

752 Shields, LaVoi, Bredemeier, and Power

competed season, encouraged cheating or aggression and whether their coach had angrily argued with officials, said bad things about opponents, made fun of a member of the team, acted like a bad sport when the team lost, hit or kicked someone on the team, or angrily yelled at a player for making a mistake. Responses to each item ranged from never (1) to often (4).

Perceived Poor Sportspersonship of Spectators (Spectators’ Poor Sport Behavior). Eight items were used to assess poor spectator behavior. Six items asked respondents to indicate whether they had observed spectators engaging in particular problematic behaviors during the last season and, if so, how often. Response options on the 4-point scale ranged from no (1) to often (4). Specifi- cally, respondents were asked whether they had personally been teased or yelled at by someone watching, or had been hit or kicked by a fan. They were also asked whether they had seen any of the following, and, if so, how often: a fan tease or yell at a teammate, a fan angrily yell at a ref or sport official, a fan angrily yell at his or her coach, or a fan hit another adult during or following an athletic event. Finally, respondents were asked whether they had been embarrassed by the behav- ior of a fan, or been scared by the behavior of a fan. These last two items, despite not measuring specific behaviors, were included because they shed light on how respondents felt about the moral behavior of spectators.

Perceived Attitude of Coach Toward Players’ Poor Sportspersonship (Coach Norm). To assess how the youth perceived their coaches’ norms regarding poor sportspersonship, we asked respondents how disappointed their coach would be if someone on the team cheated, tried to hurt an opponent, argued with a sport official, said mean or hurtful things to an opponent, made fun of another team member, or acted like a bad sport. Responses on the 4-point scale ranged from not at all (1) to very (4). This approach was based on a strategy employed by Power et al. (1989) in their assessment of school moral norms. The strategy reflects the assumption that when there is an expectation about a specific behavior—that is, when there is a commitment to a moral norm—those who share the norm will be disappointed when it is violated.

Perceived Attitude of Parents Toward Poor Sportspersonship (Parent Norm). The parental norm items paralleled those of the coach with one difference. Rather than ask how disappointed parents would be if “someone on the team” exhibited the target behaviors, the respondents were asked how disappointed their parents would be if they acted in the specified way. This shift in wording reflects the fact that whereas coaches are responsible for the whole team, parents are focused most on their own children. Responses again ranged from not at all (1) to very (4).

Procedure

Permission to conduct the study was obtained from the Committee for the Protection of Human Subjects at the University of Notre Dame. The researchers then asked for assistance from colleagues at two other universities to assist with regional data collections. At each of three sites, researchers contacted appropriate personnel within local school districts, sharing with them the goals and instruments of the study. Once permission to conduct the investigation was obtained, the researchers

Predictors of Youth Sportspersonship    753

followed the lead of contacts within the target schools to disseminate information to the target audiences and schedule data collections. Prior to collection of data from the students, parental informed consent forms were completed, and each student signed an assent form. Participation in the study was voluntary and open to all students in the fifth through eighth grade who participated in any organized sport program. Data collection times were established and the questionnaires were administered to participants by researchers.

Data Analysis

Preliminary analyses were conducted to determine whether the scales were unidi- mensional and internally consistent. Exploratory analyses were then conducted to determine whether gender, grade, sport type, or sample location were significantly related to the criterion variable.

The main analysis consisted of a hierarchical multiple regression. Athlete behavior served as the criterion variable. Athletes’ gender, grade, sport type, and location were entered in the first block. The second block entered athletes’ attitudes toward sportspersonship. The third and final block entered the social influence variables: predicted teammate behavior (shared norms), perceived coach and spectator behavior, and perceived coach and parental norms. The rationale for this three-block procedure was that we wanted to investigate the role of different types of environmental influence on athlete behavior, after accounting for the roles of demographic variables and the athletes’ own attitude toward sportspersonship. Finally, we checked the variance inflation factor (VIF) to determine whether multi- collinearity biased the regression model.

Results Psychometric Properties of Scales

To test whether the individual items could be combined into the scales described above, we first tested the unidimensionality of each scale using principal component analysis; all scales were found to be unidimensional. For each of the scales, the first principal component accounted for 37% to 63% of the variance in item responses, and all items had substantial positive loadings on the first unrotated principal component (M = .68, range: .45 to .84). We then examined the internal consistency of the scales and found that Cronbach alpha coefficients were adequate, ranging from a low of .75 for the Spectator Behavior Scale to .88 for the Coach Norm and Parent Norm Scales. Table 1 offers summary statistics for the eight scales used in the study, along with a matrix of their correlations.

Preliminary Analyses

In order to determine whether type of sport significantly related to athletes’ self- reports of sportspersonship behavior, a one-way ANOVA was conducted. The univariate F(5, 645) was 6.87, p < .001, η2 = .05. Scheffe post hoc comparisons of the six specific sports indicated that self-reported unsportspersonlike behavior was higher in hockey (M = 1.82) and football (M = 1.51) than in the other four

754

Ta b

le 1

M

e a n

s , S

ta n

d a rd

D e v ia

ti o

n s

, A lp

h a s , a

n d

C o

rr e la

ti o

n s B

e tw

e e n

V a ri

a b

le s

V ar

ia b

le n

M S

D α

1 2

3 4

5 6

7

1. P

oo r

sp or

t be

ha vi

or 65

1 1.

32 .4

7 .8

0 —

.3 6*

* .5

1* *

.5 9*

* .5

7* *

− .1

4* *

− .2

6* *

2. P

oo r

sp or

t at

ti tu

de 67

5 1.

75 .6

1 .8

3 —

.3 4*

* .3

1* *

.1 6*

* −

.3 0*

* −

.3 9*

* 3.

T ea

m n

or m

65 9

1. 77

.6 6

.8 6

— .5

3* *

.4 0*

* −

.2 5*

* −

.2 6*

* 4.

C oa

ch ’s

p oo

r sp

or t

be ha

vi or

64 5

1. 31

.4 1

.7 9

— .5

4* *

− .3

2* *

− .2

4* *

5. S

pe ct

at or

s’ p

oo r

sp or

t be

ha vi

or 59

7 1.

50 .5

0 .7

5 —

− .0

5 −

.1 0*

6. C

oa ch

n or

m 66

8 3.

43 .6

7 .8

8 —

.6 3*

* 7.

P ar

en t

no rm

66 5

3. 47

.6 6

.8 8

—

*p <

.0 5,

* *p

< .0

1.

Predictors of Youth Sportspersonship    755

sports (Ms = 1.19 to 1.31), and that the mean score for hockey was significantly (p < .05) greater than that for each of the other sports except football. Consequently, the six initial sports were combined into two categories for further analysis: high- contact (hockey, football) and medium-contact (basketball, soccer, baseball/softball, lacrosse) sports.

A 2 (sport type) × 3 (location) × 4 (grade) × 2 (gender) ANOVA of self-reported poor sportspersonship behavior yielded significant main effects for sport type, F(1, 616) = 13.84, p < .001, partial η2 = .02; location, F(2, 616) = 4.51, p < .02, partial η2 = .01; grade, F(3, 616) = 5.63, p < .02, partial η2 = .03; and gender, F(1, 616) = 5.51, p < .02, partial η2 = .01. As expected, given the results of the previously reported one-way ANOVA, unsportspersonlike behavior was higher in high-contact (M = 1.58) than in medium-contact sports (M = 1.28). Unsportspersonlike behavior also varied by location (Ms = 1.24, 1.33, and 1.44 for California, South Bend, and Philadelphia, respectively), and increased by grade (Ms = 1.19, 1.28, 1.39, and 1.42 for Grades 5 through 8, respectively). Finally, males self-reported more unsport- spersonlike behavior (M = 1.36) than did females (M = 1.27). Although none of the interactions among these factors was statistically significant, the Sport Type × Grade interaction approached significance, F(3, 616) = 2.51, p = .058, partial η2 = .01. Inspection of the cell means indicated that unsportspersonlike behavior increased with grade for medium-contact sports but decreased by grade for high- contact sports. As a result of these preliminary analyses, sport type, location, grade, gender, and the Sport Type × Grade interaction were included as control variables in the outcome analyses.

Predictors of Athletes’ Self-Reported Poor Sportspersonship Behavior

As indicated above, we conducted a hierarchical multiple regression to assess predictors of athletes’ self-reported unsportspersonlike conduct. Before interpret- ing results, we examined the VIF statistics. All VIFs were between 1.0 and 1.7, indicating that multicollinearity was not a concern. For the control variables entered in the first step of the regression, dummy codes were used for sport type (reference category = medium contact), location (reference category = California), and gender (reference category = female), and grade was rescaled as a simple linear contrast (with Grade 5 = 0). The Sport Type × Grade interaction effect was represented by their product term.

Details of each step in the regression are presented in Tables 2 and 3. The results show that the increase in R2 was significant for each block of variables and that the final regression equation accounted for almost 55% of the variance in athlete sportspersonship scores, F(12, 548) = 55.54, p < .001. The first block of control variables (gender, grade, sport type, location, and the Sport Type × Grade interaction) collectively accounted for about 8% of the variance. Adding the ath- letes’ own attitudes toward sportspersonship accounted for an additional 10% of the variance. Entering the social influence variables as a block accounted for an additional 37% of the variance.

The part correlations, which look at the correlation of each predictor variable with the outcome variable after controlling for the variance accounted for by the other predictor variables, are probably the best indicators of the importance of the

756 Shields, LaVoi, Bredemeier, and Power

Table 2 Regression Model Summaries

Model R R 2 R 2 F F p <

1 0.288 0.083 n/a 8.36 n/a 0.001

2 0.422 0.178 0.095 17.11 63.93 0.001

3 0.741 0.549 0.371 55.54 90.04 0.001

Note. Model predictors:

Model 1: (constant), location, gender, sport type, grade, Sport Type × Grade.

Model 2: adds athlete attitude.

Model 3: adds spectator behavior, coach behavior, team norm, parent norm, coach norm.

Table 3 Multiple Regression of Athlete Sportspersonship Behavior

Variable Ba

95% confidence interval

β t p ≤ Part

correlationLower Upper

Block 1 Location 1 .017 −.041 .076 .019 .578 ns .017 Location 2 .060 −.019 .139 .049 1.491 ns .043 Male −.049 −.106 .009 −.053 −1.661 ns −.048 Grade −.004 −.032 .024 −.009 −.020 ns −.008 Sport type .108 −.059 .276 .072 1.273 ns .030 Sport Type × Grade −.058 −.143 .028 −.074 −1.320 ns .037 Block 2 Poor sport attitude .125 .071 .178 .155 4.593 .001 .132 Block 3 Coach’s poor sport behavior .373 .283 .462 .325 8.170 .001 .234 Spectators’ poor sport behavior .245 .181 .310 .266 7.474 .001 .214 Team norm .145 .095 .195 .205 5.705 .001 .164 Parent norm −.129 −.183 −.074 −.180 −4.614 .001 −.132 Coach norm .118 .063 .173 .164 4.219 .001 .121

aUnstandardized.

various predictor variables. As can be seen in Table 3, none of the part correlations for the demographic variables were very strong. After controlling for the influence of the other variables, coach and spectator behavior had the strongest relationship with athlete behavior, followed by team norm, personal attitude, and parent and coach norms.

Discussion In this investigation, we sought to advance our understanding of theoretically relevant potential influences on self-reported poor sportspersonship behaviors among youth sport participants. On the whole, it is worth noting that the athletes’

Predictors of Youth Sportspersonship    757

self-reported frequency of poor sport behavior was relatively low. In investigat- ing that behavior, we were interested in looking at the predictive importance of variables reflecting theoretically relevant social influences, along with the athletes’ own attitudes toward sportspersonship.

Preliminary analyses revealed a number of demographic differences. Consistent with previous findings, males self-reported more unsportspersonship behavior than did females (e.g., Allison, 1982; Silva, 1983; Duda et al., 1991). Also consistent with the literature, there was a trend for increasingly poor behavior by grade, which may support the notion of an in-sport socialization process (Allison, 1982; Kavussanu, Seal, & Phillips, 2006; McIntosh, 1979; Silva, 1983; Smith, 1983; Vaz, 1982). We also found that poor sport behaviors tend to occur more frequently in sports such as football and hockey than in such sports as basketball and soccer. When there is extensive physical contact within a sport, it may be more challenging to maintain good sport behaviors (see Bredemeier et al., 1986). The differences by geographic location may have reflected differing regional norms, the different proportions of sport participants between the Philadelphia and San Francisco–East Bay samples, or idiosyncrasies arising from the use of convenience samples. These various demographic findings are suggestive, but it is important to note that all effect sizes were quite small.

Previous moral atmosphere research has focused almost exclusively on the salient role of coaches and/or teammates (Guivernau & Duda, 2002; Kavussanu et al., 2002; Smith, 2003; Stephens, 2000, 2001; Stephens & Bredemeier, 1996; Stephens, Bredemeier, & Shields, 1997; Stephens & Kavanagh, 2003; Stornes, 2001; Stuart & Ebbeck, 1995). The moral influence of parents has also been stud- ied occasionally (Smith, 1979; Stuart & Ebbeck, 1995; Guivernau & Duda, 2002). The findings of the present study are consistent with the view that coaches, team- mates, and parents are important influences, while also suggesting that spectators be considered as well.

Consistent with previous research (Stephens, 2000, 2001; Stephens & Brede- meier, 1996; Stephens, Bredemeier, & Shields, 1997; Stornes, 2001; Stuart & Ebbeck, 1995), perceptions of the coach were found to be associated with whether youth sport participants will behave in appropriate ways. Concurring with the cliché “actions speak louder than words,” the results from the present study suggest that it may be the coaches’ behavior, rather than their expressed attitude, that matters most to athletes. If the coach wants to have a positive influence, he or she may need to uphold high standards of professional and ethical behavior.

We were somewhat surprised by the association between perceived spectator behavior and young people’s reports of their own poor sport behavior. Spectator misbehavior may provide a cue to athletes that poor sport behavior will be tolerated or endorsed. It is worth noting that in youth sports, most spectators are parents of the participants. As such, they are likely to be highly identified with their sport team. Previous research has demonstrated that such identification is correlated with aggressive spectator behavior (Wann, Carlson, & Schrader, 1999; Wann, Schrader, & Carlson, 2000). From the child’s side, both one’s own parents and those of friends and teammates are likely esteemed sources of information about what matters within a sport context. When spectator behavior deviates from positive norms, it may communicate to the child that significant others do not place a high value on good sportspersonship.

758 Shields, LaVoi, Bredemeier, and Power

Self-reported poor sport behavior was also significantly associated with what participants believed about the likelihood of teammates engaging in similar behavior, a finding that is congruent with previous research on team moral atmosphere among both college athletes (Kavussanu et al., 2002; Shields et al., 1995) and youth athletes (Guivernau & Duda, 2002; Kavussanu & Spray, 2006; Stephens, 2001; Stephens & Bredemeier, 1996; Stephens & Kavanagh, 2003). This finding, in conjunction with relevant theory, suggests that if coaches spend more time developing posi- tive, shared norms regarding good sportspersonship, it may reduce the incidence of poor sport behavior. Future research is needed, however, to confirm the causal links postulated by the theory.

Beyond perceived behaviors and team norms, the perceived norms of parents and coaches also contributed significantly, although very modestly, to the prediction of self-reported poor sport behavior. It is important to emphasize that these predic- tors remained significant after controlling for the influence of perceived spectator and coaching behavior. Based on the simple correlations, one might theorize that youth care what their coaches and parents think and do not wish to disappoint them. Future researchers may want to separate the influences of fathers and moth- ers, which were conflated in the present study. Future researchers may also want to examine more closely the independent contributions of each of the predictor variables. In the present investigation, we found that the coach norm variable cor- related negatively with self-reported poor sport behavior, as anticipated, but in the regression analysis the direction of association was positive. It is unclear how this finding should be interpreted.

Finally, the attitudes that the youths themselves held toward unsportspersonlike behaviors were also significant predictors of that behavior. That the perceived behav- iors of coaches and spectators appear to be better predictors than the athletes’ own attitudes may seem to suggest that environmental factors are more influential than personal factors when it comes to unsportspersonlike behavior. We would caution against such an interpretation, however. The dynamic interactions between various personal and contextual influences are likely to lead to fluctuating patterns of causal- ity and influence. Moreover, important personal attributes, such as moral reasoning or achievement orientation, were not included in the present investigation.

The present findings are congruent with constructivist theory, which posits that individual attitudes and social influences are both implicated in the regulation of behavior. Additional research is needed to probe the complex relations among personal and social influences, also taking account of potential developmental changes in how these relationships operate. For example, younger children may be more likely than older to conform to adult expectations (Damon, 1983).

Limitations of the Present Study and Directions for Future Research

The implications of the present findings for moral atmosphere research suggest that several influences on the moral atmosphere need to be isolated and evaluated, and that moral atmosphere researchers need to examine other contextual influences on

Predictors of Youth Sportspersonship    759

behavior as well. We define the moral atmosphere of a sport team as the collective moral norms that team members experience as proscribing or prescribing specific types of behavior based on shared team values. No doubt coaches and players are the primary influences on team moral atmosphere, but distal influences, such as those of parents and spectators, play a role as well. Moreover, although parents and spectators may play an indirect role in shaping the moral atmosphere of a sport team, they may also exercise an independent role in influencing behavior in specific situations and with specific individuals. These theoretical speculations extend beyond what can be demonstrated from the present investigation, however, and point to future directions for research.

The present study utilized a cross-sectional design, and, consequently, it cannot establish cause–effect relationships. Future investigators may wish to employ a longitudinal research strategy to clarify patterns of causality. Additionally, the present study focused on a relatively small age range and results should not be generalized to other populations. Shifts that often occur with development in the role of parents and peers, in sources of competence information, in social perceptions, and in moral reasoning, to name only a few, may result in predictable age-related changes in the correlates and predictors of poor sport behavior.

The questionnaire utilized in the present study contained items tailored to each specific reference group (self, teammates, coaches, spectators, parents). This was both a strength and a limitation. On the positive side, it enabled us to tap items of specific relevance to each group. However, it also resulted in each scale having unique items. Thus, even though the present study captured a broad and salient range of sportspersonship variables, future researchers might consider focusing on a narrower list of shared items that could apply to all groups.

Finally, intervention studies designed to build collective norms for good sport behavior could be used to address the question of whether a deliberately created moral atmosphere can eliminate or minimize any negative influences from outside the team, such as misbehaving spectators. We would hypothesize that shared moral norms become increasingly potent in their influence over individual attitudes and behaviors as the norms become more explicit and clear within the team, but that is an hypothesis that must be tested in future studies.

In conclusion, the present investigation provides strong evidence for the importance of both personal attitudes and external moral influences, including the moral atmosphere of the team. Based on the current findings, perceptions of the behaviors of coaches and spectators, and perceptions of the sportsper- sonship norms of coaches, teammates, and parents should be considered likely sources of social influence on poor sport behavior.

Acknowledgments

This research was commissioned by the Citizenship through Sports Alliance and received in-kind support from the Office of Research of the National Collegiate Athletic Asso- ciation (NCAA). Thanks are extended to Ed Hastings, Steve Miller, and Carl Edwards for their valuable assistance in data collection.

760 Shields, LaVoi, Bredemeier, and Power

References

Allison, M.T. (1982). Sportsmanship: Variation based on sex and degree of competitive expe- rience. In A.O. Dunleavy, A.W. Miracle, & C.R. Rees (Eds.), Studies in the sociology of sport (pp.153-165). Fort Worth, TX: Texas Christian University Press.

Ames, C. (1992). Achievement goals, motivational climate, and motivational processes. In G.C. Roberts (Ed.), Motivation in sport and exercise (pp. 161-176). Champaign, IL: Human Kinetics.

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 Psychology, 16, 1-14.

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

Bredemeier, B., & Shields, D. (1986). Athletic aggression: An issue of contextual morality. Sociology of Sport Journal, 3, 15-28.

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

Bredemeier, B., Weiss, M., Shields, D., & 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.

Damon, W. (1983). Social and personality development: Infancy through adolescence. New York: W.W. Norton.

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

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

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

Gano-Overway, L., Guivernau, M., Magyar, T.M., Waldron, J., & Ewing, M. (2005). Achieve- ment goal perspectives, perceptions of the motivational climate, and sportspersonship: Individual and team effects. Psychology of Sport and Exercise, 6, 215-232.

Guivernau, M., & Duda, J. (2002). Moral atmosphere and athletic aggressive tendencies in young soccer players. Journal of Moral Education, 31, 67-85.

Higgins, A., Power, C., & Kohlberg, L. (1984). The relationship of moral atmosphere to judg- ments of responsibility. In W. Kurtines & J. Gewirtz (Eds.), Morality, moral behavior, and moral development (pp. 74-106). New York: Wiley.

Kavussanu, M. (2006). Motivational predictors of prosocial and antisocial behaviour in football. Journal of Sport Sciences, 24, 575-588.

Kavussanu, M., Roberts, G., & Ntoumanis, N. (2002). Contextual influences on moral functioning of college basketball players. The Sport Psychologist, 16, 347-367.

Kavussanu, M., Seal, A., & Phillips, D. (2006). Observed prosocial and antisocial behaviors in male soccer teams: Age differences across adolescence and the role of motivational variables. Journal of Applied Sport Psychology, 18, 326-344.

Kavussanu, M., & Spray, C.M. (2006). Contextual influences on moral functioning of male youth footballers. The Sport Psychologist, 20, 1-23.

Kohlberg, L. (1984). Essays on Moral Development: Vol. 2: The Psychology of Moral Development. San Francisco: Harper & Row.

Predictors of Youth Sportspersonship    761

Kohlberg, L., & Candee, D. (1984). The relationship of moral judgment to moral action. In L. Kohlberg, Essays on moral development, Vol. 2: The psychology of moral develop- ment (pp. 498-581). San Francisco: Harper & Row.

Lemyre, P.N., Roberts, G.C., & Ommundsen, Y. (2002). Achievement goal orientations, perceived ability, and sportspersonship in youth soccer. Journal of Applied Sport Psychology, 14, 120-136.

Martens, R. (1976). Kid sports: A den of iniquity or land of promise. In R. Magill, M. Ash, & F. Smoll (Eds.), Children in sport: A contemporary anthology (pp. 201-216). Champaign, IL: Human Kinetics.

McIntosh, P.C. (1979). Fair play: Ethics in sport and education. London: Heinemann. Miller, B., Roberts, G.C., & Ommundsen, Y. (2005). Effect of perceived motivational cli-

mate on moral functioning, team moral atmosphere perceptions, and the legitimacy of intentionally injurious acts among competitive youth football players. Psychology of Sport and Exercise, 6, 461-477.

Nicholls, J. (1989). The competitive ethos and democratic education. Cambridge, MA: Harvard University Press.

Ommundsen, Y., Roberts, G.C., Lemyre, P.N., & Treasure, D. (2003). Perceived motivational climate in male youth soccer: Relations to social-moral functioning, sportspersonship and team norm perceptions. Psychology of Sport and Exercise, 4, 397-413.

Piaget, J. (1948/1973). To understand is to invent. New York: Grossman. (First published in Prospects, UNESCO Quarterly Review of Education.)

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

Power, F.C., & Reimer, J. (1978). Moral atmosphere: An educational bridge between moral judgment and action. In W. Damon (Ed.), New directions for child development: Moral development 2 (pp. 105-116). San Francisco: Jossey-Bass.

Rest, J. (1984). The major components of morality. In W. Kurtines & J. Gewirtz (Eds.), Morality, moral behavior, and moral development (pp. 356-429). New York: Wiley.

Sage, L., & Kavussanu, M. (2007). The effects of goal involvement on moral behavior in an experimentally manipulated competitive setting. Journal of Sport & Exercise Psychology, 29, 190-207.

Sage, L., Kavussanu, M., & Duda, J.L. (2006). Moral identity and motivational predictors of prosocial and antisocial functioning in association football. Journal of Sports Sci- ences, 24, 455-466.

Shields, D., & Bredemeier, B. (1995). Character development and physical activity. Cham- paign, IL: Human Kinetics.

Shields, D., Bredemeier, B., Gardner, D., & Bostrom, A. (1995). Leadership, cohesion and team norms regarding cheating and aggression. Sociology of Sport Journal, 12, 324-336.

Shields, D., Bredemeier, B., LaVoi, N., & Power, F.C. (2005). The behavior of youth, par- ents, and coaches: The good, the bad, and the ugly. Journal of Research on Character Education, 3, 43-59.

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

Smith, A.L. (2003). Peer relationships in physical activity contexts: A road less traveled in youth sport and exercise psychology research. Psychology of Sport and Exercise, 4, 25-39.

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

Smith, M.D. (1983). Violence and sport. Toronto: Butterworths. Stephens, D. (2000). Predictors of likelihood to aggress in youth soccer: An examination of

coed and all-girls teams. Journal of Sport Behavior, 23, 311-325.

762 Shields, LaVoi, Bredemeier, and Power

Stephens, D. (2001). Predictors of aggressive tendencies in girls’ basketball: An examination of beginning and advanced participants in a summer skills camp. Research Quarterly for Exercise and Sport, 72, 257-266.

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

Stephens, D., Bredemeier, B., & Shields, D. (1997). Construction of a measure designed to assess players’ descriptions and prescriptions for moral behavior in youth sport soccer. International Journal of Sport Psychology, 28, 370-390.

Stephens, D., & Kavanagh, B. (2003). Aggression in Canadian youth ice hockey: The role of moral atmosphere. International Sports Journal, 7, 109-119.

Stornes, T. (2001). Sportspersonship in elite sports: On the effects of personal and environ- mental factors on the display of sportspersonship among elite male handball players. European Physical Education Review, 7, 283-304.

Stuart, M.E. (2003). Moral issues in sport: The child’s perspective. Research Quarterly for Exercise and Sport, 74, 445-454.

Stuart, M., & Ebbeck, V. (1995). The influence of perceived social approval on moral devel- opment in youth sport. Pediatric Exercise Science, 7, 270-280.

Stuntz, C.P., & Weiss, M.R. (2003). Influence of social goal orientations and peers on unsportsmanlike play. Research Quarterly for Sport and Exercise, 74, 421-435.

Vallerand, R.J., Briere, N.M., Blanchard, C., & Provencher, P. (1997). Development and validation of the multidimensional sportspersonship orientations scale. Journal of Sport & Exercise Psychology, 19, 197-206.

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

Wann, D., Carlson, J., & Schrader, M. (1999). The impact of team identification on the hostile and instrumental verbal aggression of sport spectators. Journal of Social Behavior and Personality, 14, 279-286.

Wann, D., Schrader, M.P., & Carlson, J.D. (2000). The verbal aggression of sport specta- tors: A comparison of hostile and instrumental motives. International Sports Journal, 4, 56-63.

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

Manuscript submitted: October 18, 2005 Revision accepted: June 18, 2007