Athlete-Perceived Coaching Behaviors: Relating Two Measurement ...

28 downloads 32858 Views 93KB Size Report
For more than two decades, the behavioral categories of the Leadership Scale for Sports (LSS) and the Coaching Behavior Assessment System (CBAS) have ...
BRIEF REPORT JOURNAL OF SPORT & EXERCISE PSYCHOLOGY, 2006, 28, 205-213 © 2006 Human Kinetics, Inc.

Athlete-Perceived Coaching Behaviors: Relating Two Measurement Traditions Sean P. Cumming, Ronald E. Smith, and Frank L. Smoll University of Washington For more than two decades, the behavioral categories of the Leadership Scale for Sports (LSS) and the Coaching Behavior Assessment System (CBAS) have been used by a wide range of researchers to measure coaching behaviors, yet little is known about how the behavioral categories in the two models relate statistically to one another. Male and female athletes on 63 high school teams (N N = 645) completed the LSS and the athlete-perception version of the CBAS (CBAS-PBS) following the sport season, and they evaluated their coaches. Several of Chelladurai’s (1993) hypotheses regarding relations among behavioral categories of the two models were strongly supported. However, many significant and overlapping correlations between LSS subscales and CBASPBS behavioral categories cast doubt upon the specificity of relations between the two instruments. The LSS and the CBAS-PBS accounted for similar and notable amounts of variance in athletes’ liking for their coach and evaluations of their knowledge and teaching ability. Key Words: coach, leadership, youth, Leadership Scale for Sports, Coaching Behavior Assessment System

The study of coaching behaviors, their antecedents, and their effects on athletes has long occupied the attention of sport science researchers. Over the past 25 years, two theoretical models and measurement traditions have contributed to the understanding of leadership behaviors in sport. One of them, Chelladurai’s multidimensional leadership model (Chelladurai, 1993; Chelladurai & Saleh, 1980), focuses on important leadership dimensions identified in previous research in nonsport settings and applies them to the study of sport leadership. To measure these leadership dimensions, Chelladurai and Saleh (1980) developed the Leadership Scale for Sports (LSS), a 40-item measure with subscales for training and instruction, democratic behavior, autocratic behavior, social support, and positive feedback. The scale has been used to measure athletes’ perceptions of their coaches’ behavior, their preferred leadership behaviors, and coaches’ perceptions of their own behavior (for reviews of LSS research, see Chelladurai, 1993; Chelladurai & Riemer, 1998). There is no question that the LSS is an important and useful research tool. The authors are with the Department of Psychology, University of Washington, Seattle, WA 98195-1525. 205

206 / Cumming, Smith, and Smoll

At about the same time Chelladurai’s model appeared, Smoll, Smith, and Hunt (1978; Smoll & Smith, 1989) advanced a cognitive-behavioral model derived from social learning theory (Bandura, 1969; Mischel, 1973). They labeled their model a mediational model because, like Chelladurai’s, it emphasized the role of athlete perceptions as causally mediating relations between overt coaching behaviors and athletes’ reactions to their athletic experiences. Using the methods of behavioral assessment, they developed the Coaching Behavior Assessment System, a coding system for observing and recording coaching behaviors during practices and games (CBAS; Smith, Smoll, & Hunt, 1977a). The original CBAS contained 12 behavioral categories subdivided into reactive and spontaneous categories. The reactive behaviors were responses to identifiable situations. They included reinforcement and nonreinforcement as responses to positive athlete behaviors or effort; mistakecontingent technical instruction, mistake-contingent encouragement, punishment, punitive technical instruction, and ignoring mistakes as responses to mistakes and errors; and keeping control as a response to misbehaviors. Spontaneous behavioral categories included general technical instruction, general encouragement, organization, and general communication. Because the mediational model also required the measurement of athleteperceived behaviors, Smith, Smoll, and Curtis (1978) developed a parallel measure of athlete-perceived coaching behaviors, the CBAS Perceived Behavior Scale (CBAS-PBS). They used a definitional approach in which a narrative description of each CBAS category was derived from the descriptors and coding criteria in the CBAS observer training manual (Smith, Smoll, & Hunt, 1977b). For each of the 12 definitional items, the athlete indicates the frequency with which the coach behaved in that manner. Later, an additional reinforcement + instruction category introduced by Horn (1985) was added. The CBAS-PBS can be used to measure athlete perceptions of coaches’ behaviors, and a slightly modified version can be used to assess coaches’ perceptions of their own behavior. Although the LSS and the CBAS emerged from different research paradigms, the constructs defined by the two measures appear to have some notable commonalities. Indeed, Chelladurai (1993) advanced specific hypotheses about how three of the five LSS dimensions might relate to the CBAS categories. Specifically, he predicted that the CBAS behaviors of general technical instruction, punitive technical instruction, keeping control, and organization would fall within the training and instruction dimension of the LSS. Second, Chelladurai predicted that reinforcement and general encouragement would correlate positively with the LSS’s positive feedback scale, whereas nonreinforcement would correlate negatively with this scale. Finally, Chelladurai expected the CBAS category of general communication to correlate positively with the social support scale of the LSS. This latter prediction is at variance with the factor analytic findings of Smith and Smoll (1990), who found reinforcement and mistake-contingent encouragement to be markers of a supportiveness dimension. Chelladurai (1993) made no CBAS-related predictions concerning the LSS dimensions of democratic and autocratic leadership, so their relations with the CBAS categories remain an empirical question of considerable interest, given the divergent athlete preference and satisfaction findings for these LSS dimensions (Dwyer & Fischer, 1988; Reimer & Chelladurai, 1995). For more than two decades, the multidimensional and mediational models, while informing and stimulating significant amounts of research on coaching behaviors, have proceeded along parallel but largely independent lines. Although the two

Athlete-Perceived Coaching Behaviors / 207

models have constructs that appear to have considerable conceptual overlap, few attempts have been made to assess empirical relations among their constructs and between the measurement tools they employ. Likewise, Chelladurai’s hypotheses are deserving of empirical testing. Thus an examination of LSS and CBAS-PBS relations seems long overdue on both conceptual and methodological grounds. A previous study by Amorose and Horn (2000) suggested linkages between LSS and CBAS-PBS behaviors, marking a useful beginning to the study of relations between the two measures. However, Amorose and Horn collapsed 8 of the CBAS-PBS categories into broader “feedback” indices, and 5 other CBAS-PBS categories were not utilized. Therefore the need remains for a more basic and fine-grained examination of LSS and CBAS-PBS links using all of the CBAS-PBS categories. In comparing the two measures, it is important to know not only how the LSS and CBAS-PBS categories relate to one another but also how they are associated with other meaningful variables. One of the most important aspects of the sport experience is the relationship formed between athletes and coaches. Both positive and negative coach-athlete relations have important psychosocial, motivational, emotional, and performance consequences, and they are related to athletes’ decisions to continue or discontinue sport participation (Smoll & Smith, 2002). Both the LSS and CBAS-PBS categories have been shown to be related to athletes’ enjoyment of and liking for the coach, but so far there has been no comparative study of the LSS and the CBAS-PBS relations to athletes’ evaluative reactions to their coach. Liking for the coach and evaluations of the coach’s knowledge and teaching ability have served as major outcome variables in previous studies of coaching effectiveness and coach interventions (Smith, Smoll, & Christensen, 1996). In the present study, we relate LSS and CBAS-PBS scores to these outcome measures and compare the amounts of attitude variance captured by the two athlete-perception measures.

Method Participants The participants were 388 male and 257 female varsity athletes who played on 63 varsity teams at 13 high schools in the western United States. The sample consisted of 101 male basketball players, 146 male football players, 41 male cross-country runners, 100 male wrestlers, 108 female basketball players, 47 female gymnasts, 22 female cross-country runners, and 79 female soccer players. The athletes ranged in age from 13 to 18 years (M M = 16.2, SD = 1.0). The majority of the participants were Caucasian (89.7%), with only a small number of African Americans (3.7%), Asians (4.0%), Hispanics (1.0%), and Pacific Islanders (1.6%).

Procedures and Measures Parental consent and athlete assent were obtained prior to the study. Athletes completed the measures in a group setting at the end of the sport season. The questionnaires included the LSS and CBAS-PBS measures of athlete-perceived coaching behaviors, together with a questionnaire measuring evaluative reactions toward the coach, peers, and the sport experience. The 40-item LSS (Chelladurai & Saleh, 1980) has been used in many sport psychology studies and its psychometric integrity is well documented (Chelladurai, 1993). Athletes indicate on a 5-point Likert scale how often their coach engaged in

208 / Cumming, Smith, and Smoll

a specific behavior. The response categories range from 0 to 4, i.e., always (100% of the time) (0); often (75%) (1); occasionally (50%) (2); seldom (25%) (3); and never (0%) (4). The five scales of the LSS contain 5 to 13 items. In our sample, Cronbach’s alphas were .92 for training; .88 for democratic; .76 for autocratic; .89 for support; and .89 for positive feedback. For the purpose of data analysis and interpretation in relation to the CBAS-PBS scores, item responses were reverse coded (e.g., 0 = 4, 1 = 3) so that higher values reflected more frequently observed leadership behaviors. Athletes also completed the 13-item CBAS-PBS, the definitional items of which consist of behavioral descriptions derived from the CBAS observer training manual (Smith et al., 1977b). A sample definitional item (mistake-contingent encouragement), derived from the CBAS training manual, is: “Sometimes players goof and make mistakes. Some coaches give their players support and encouragement after they make a mistake. For example they may say, ‘That’s OK, don’t worry about it; you’ll get ‘em next time.’ Other coaches don’t give much encouragement after mistakes. Circle how often your coach encouraged you after you made mistakes.” Athletes indicated how frequently their coaches engaged in each class of behavior on a Likert scale ranging from 1 (never) to 7 (almost always). Three attitudinal items used in previous research on coach-athlete relations (Smith et al., 1978; Smith & Smoll, 1990) and as outcome measures in coaching intervention outcome studies (Smith, Smoll, & Curtis, 1979; Smoll, Smith, Barnett, & Everett, 1993) were used to assess specific evaluative reactions of the athletes to their coach. On 7-point rating scales, the athletes responded to the questions, “How much did you like playing for your coach?” “How much does your coach know about your sport?” and “How good is your coach at teaching your sport?” These items have been shown to be highly sensitive to intervention effects in studies designed to enhance coach-athlete relations and to reduce sport attrition through coach training (e.g., Barnett, Smoll, & Smith, 1992; Smith et al., 1979; Smoll et al., 1993).

Results and Discussion Relations Between LSS and CBAS-PBS Perceived Behaviors All analyses were first conducted for males and females separately. The results of the analyses showed little variation by gender. Therefore males and females were combined for the analyses reported in this article. Pearson product moment correlations were computed so as to examine the predicted associations between athlete perceptions of coaching behaviors, as measured by the CBAS-PBS, and coach leadership styles, as measured by the LSS. The results are presented in Table 1. Table 1 also shows intercorrelations among the five LSS scales. As in previous research (Amorose & Horn, 2000; Chelladurai & Reimer, 1998), many of the LSS scales were correlated with one another. The positively toned training, democratic, social support, and positive feedback scales exhibited positive intercorrelations ranging from .32 to .61. The autocratic scale was closest to a statistically independent scale, showing low and nonsignificant negative correlations with the training, democratic, and social support scales. Only in its negative correlation of –.30 with positive feedback did the autocratic scale exhibit significant shared variance with another LSS scale.

Athlete-Perceived Coaching Behaviors / 209 Table 1 Intercorrelations Among LSS Scales and LSS Correlations With the CBAS-PBS Scales LSS Scales

LSS Scales

Training

Training Democratic Autocratic Social support

Democratic

Autocratic

Social support

Positive feedback

.32**

–.17* –.02

.61** .49** –.17*

.51** .35** –.30** .60**

.25** –.19** .27** .34** .21** –.25** –.19** .01 –.10** .10** .20** .10** .22**

–.28** .33** –.20** –.21** –.19** .31** .36** .20** .07* –.19** –.29** –.24** –.35**

.47** –.38** .45** .40** .43** –.13** –.17** –.24** .09* .41** .37** .29** .34**

.57** –.48** .44** .38** .37** –.23** –.24** –.23** .07* .33** .46** .34** .40**

CBAS-PBS Scales Reinforcement .37** Nonreinforcement –.32** Reinforcement + instruction .44** Mistake-cont. encouragement .30** Mistake-cont. technical instr. .47** Punishment .06 Punitive technical instruction –.02 Ignoring mistakes –.36** Keeping control .16** General technical instruction .50** General encouragement .31** Organization .37** General communication .27** * < .05; ** *p **p < .01; N = 645

The CBAS-PBS categories were less highly correlated with one another, exhibiting a mean correlation of ±.29. Only 4 of the 78 CBAS-PBS intercorrelations reached .50. The highest correlation (rr = .64) occurred between punishment and punitive technical instruction, two conceptually and behaviorally related categories. Given these results, the CBAS-PBS may be considered to have relatively distinct categories suitable for evaluating relations with the LSS scales. As predicted by Chelladurai (1993), the LSS training dimension was positively associated with general technical instruction, mistake-contingent technical instruction, keeping control, and organization, as well as with reinforcement + instruction. Negative correlations with nonreinforcement and ignoring mistakes, though not predicted by Chelladurai, are also consistent with a training/instructional orientation and with results reported by Amorose and Horn (2000), and these results indicate convergent validity (Campbell & Fiske, 1959). Contrary to Chelladurai’s hypothesis, the training dimension was unrelated to punitive technical instruction, and it correlated only weakly with keeping control. As expected, the LSS positive feedback dimension correlated positively and strongly with reinforcement and reinforcement + instruction and negatively with

210 / Cumming, Smith, and Smoll

nonreinforcement, the most conceptually similar CBAS-PBS categories. Modest negative correlations occurred with punishment and punitive technical instruction as well, which also seems consistent with a positive coaching style. However, LSS positive feedback also correlated substantially with both encouragement categories, both technical instruction categories, organization, and general communication. The relations involving the latter three CBAS-PBS categories are not consistent with high discriminant validity. Chelladurai’s prediction that the social support dimension would be positively correlated with general communication was also supported. Even more strongly correlated with social support, however, were the CBAS-PBS categories of reinforcement and mistake-contingent encouragement, which are the marker variables identified through factor analysis for Smith and Smoll’s (1990) supportiveness dimension on the observed behavior CBAS. On the other hand, discriminant validity was not evident in positive correlations of LSS social support with the two technical instruction categories, keeping control, and organization. Although Chelladurai did not predict any associations between democratic and autocratic leadership and the CBAS-PBS behaviors, a number of significant correlations were observed. The democratic dimension was positively correlated with most of the positively toned CBAS-PBS behavior categories and negatively with the two punitive categories and with nonreinforcement. The autocratic dimension showed the opposite pattern of correlations, being positively correlated with nonreinforcement and both punishment categories, and negatively correlated with the positively valenced reinforcement, encouragement, technical instruction, and general communication categories.

Coaching Behaviors and Evaluative Reactions to the Coach LSS Relations. Pearson product moment correlations were calculated to examine the associations between the LSS dimensions, the CBAS-PBS scores, and the athletes’ evaluative reactions toward their coaches. The results of the analysis are presented in Table 2. The athletes’ degree of enjoyment in playing for the coach and ratings of the coach’s knowledge of the sport and teaching ability were all positively correlated with the training, democratic, social support, and positive feedback dimensions of the LSS and negatively correlated with the autocratic dimension. Training was positively and substantially correlated with all three attitude measures. Support and positive feedback correlated more highly with the liking and teaching ability variables than they did with coach’s perceived knowledge about the sport. Although the democratic and autocratic subscales were also significantly correlated (positively and negatively, respectively) with all three coach evaluation measures, they accounted for substantially less variance than did the other three LSS subscales. It appears that the behaviors referenced by the autocratic scale have a negative valence for athletes and the other four LSS scales are positively toned. This assumption is supported by the pattern of relations with the coach evaluation measures, where the training, social support, and positive feedback all correlated highly with liking for the coach. CBAS-PBS Relations. The athlete evaluations of the coach were also significantly correlated with most of the perceived coaching behaviors in the CBAS-PBS. On the liking scale, the strongest positive correlations occurred for the reinforcement, the two encouragement and instructional categories, for organization, and

Athlete-Perceived Coaching Behaviors / 211 Table 2 Coach

LSS and CBAS-PBS Coaching Behaviors and Athletes’ Evaluations of the

Coaching behaviors

Liked playing for

Coach knowledge

Teaching ability

LSS Dimensions Training Democratic Autocratic Social support Positive feedback

.48** .33** –.30** .53** .46**

.51** .20** –.20** .35** .33**

.55** .28** –.23** .47** .41**

CBAS-PBS Categories Reinforcement Nonreinforcement Reinforcement + instruction Mistake-cont. encouragement Mistake-cont. tech. instruction Punishment Punitive technical instruction Ignoring mistakes Keeping control General technical instruction General encouragement Organization General communication

.44** –.42** .44** .39** .42** –.22** –.23** –.22** –.01 .42** .36** .36** .39**

.26** –.27** .32** .17** .34** –.04 –.08* –.27** .10** .41** .27** .36** .26**

.39** –.39** .46** .29** .44** –.08* –.16** –.30** .10** .48** .41** .42** .31**

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

for general communication. Nonreinforcement was the strongest negative predictor of liking, exceeding negative correlations involving the two punitive categories. The instructional categories and organization were most highly correlated with perceived knowledge of the sport. Positive evaluations of the coach’s teaching skills were most strongly related to the CBAS-PBS categories of reinforcement + instruction, the two instructional categories, organization, and general communication. Nonreinforcement and ignoring mistakes were negatively correlated with rated teaching ability. Comparative Relations With Coach Evaluations. To compare the overall ability of the LSS and CBAS-PBS measures to predict evaluative attitudes toward the coach, we conducted multiple regression analyses, regressing the attitude scores onto the five LSS scales and the 13 CBAS-PBS categories in separate analyses. These analyses yielded large multiple correlations across evaluative measures for both scales, ranging from .51 to .63. On “like playing for the coach,” the CBAS-PBS behaviors accounted for 39% of the variance, the LSS subscales accounted for 37%. On “coach’s knowledge of your sport,” the LSS subscales accounted for 30% of the variance, the CBAS-PBS categories accounted for 26%. For “coach’s teaching

212 / Cumming, Smith, and Smoll

ability, the CBAS-PBS behavior scores accounted for 40% of the variance, the LSS accounted for 37%. Thus the two measures accounted for substantial and similar amounts of variance in athletes’ evaluative reactions toward the coaches.

Conclusion The present study was the first to examine relations between Chelladurai’s theory-based sport leadership measure and all of the empirically derived behavioral categories of the Coaching Behavior Assessment System. Our results indicate considerable convergence between measures of perceived coaching behaviors that originate from different theoretical and methodological traditions, and their areas of convergence are also reflected in their relations with athletes’ evaluative responses to their coaches. Less impressive, however, was the level of discriminant validity shown in the relations between the LSS and the CBAS-PBS. Both the LSS and CBAS-PBS categories account for significant variance in athletes’ evaluative reactions, and they appear to be tapping conceptually meaningful behavioral differences among coaches. We should point out, however, that because our results are correlational, causal interpretations are tentative. It is tempting to interpret results of studies like ours as indicating that differences in coaching behaviors are the cause of athletes’ evaluative reactions to the coach. However, we must remember that the athletes are the source of both the perceived behavior data and the evaluative responses. It is therefore possible that positive or negative evaluations of the coach due to other, unmeasured factors create a tendency to attribute positively or negatively appraised behaviors to the coach. The behaviors of coaches play a major role in the athletic environment and influence performance, skill acquisition, enjoyment, group processes, and the psychosocial growth of athletes. They clearly deserve the empirical attention they have received, but their study depends on the availability of reliable and valid measures. In this study, we found that the LSS and the CBAS-PBS measures tap behavioral phenomena that are both theoretically meaningful and related to athletes’ attitudes toward the coach. Nonetheless, we agree with Chelladurai (1993) that a complete understanding of coaching behaviors will require the exploration of the antecedents of coaching behaviors, relations among observed and athlete-perceived behaviors, and the consequences of those behaviors and perceptions.

References Amorose, A.J., & Horn, T.S. (2000). Intrinsic motivation: Relationships with collegiate athletes’ gender, scholarship status, and perceptions of their coaches’ behavior. Journal of Sport & Exercise Psychology, 22, 63-84. Bandura, A. (1969). Principles of behavior modification. New York: Holt, Rinehart & Winston. Barnett, N.P., Smoll, F.L., & Smith, R.E. (1992). Effects of enhancing coach-athlete relationships on youth sport attrition. The Sport Psychologist, 6, 111-127. Campbell, D.T., & Fiske, D.W. (1959). Convergent and discriminant validation by the multitrait-multimethod matrix. Psychological Bulletin, 56, 81-105. Chelladurai, P. (1993). Leadership. In R.N. Singer, M. Murphy, & K.L. Tenant (Eds.), Handbook of research in sport psychology (pp. 647-671). New York: Macmillan. Chelladurai, P., & Reimer, H.A. (1998). Measurement of leadership in sport. In J.L. Duda

Athlete-Perceived Coaching Behaviors / 213 (Ed.), Advances in sport and exercise psychology measurement (pp. 227-253). Morgantown, WV: Fitness Information Technology. Chelladurai, P., & Saleh, S.D. (1980). Dimensions of leadership behavior in sport: Development of a leadership scale. Journal of Sport Psychology, 2, 34-45. Dwyer, 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. Horn, T.S. (1985). Coaches’ feedback and changes in children’s perceptions of their physical competence. Journal of Educational Psychology, 77, 174-186. Mischel, W. (1973). Toward a cognitive social learning reconceptualization of personality. Psychological Review, 80, 252-283. Reimer, H.A., & Chelladurai, P. (1995). Leadership and satisfaction in athletics. Journal of Sport & Exercise Psychology, 17, 276-293. Smith, R.E., & Smoll, F.L. (1990). Self-esteem and children’s reactions to youth sport coaching behaviors: A field study of self-enhancement processes. Developmental Psychology, 26, 987-993. Smith, R.E., Smoll, F.L., & Christensen, D.S. (1996). Behavioral assessment and interventions in youth sports. Behavior Modification, 20, 3-44. Smith, R.E., Smoll, F.L., & Curtis, B. (1978). Coaching behaviors in Little League Baseball. In F.L. Smoll & R.E. Smith (Eds.), Psychological perspectives in youth sports (pp. 173-201). Washington, DC: Hemisphere. Smith, R.E., Smoll, F.L., & Curtis, B. (1979). Coach effectiveness training: A cognitivebehavioral approach to enhancing relationship skills in youth sport coaches. Journal of Sport Psychology, 1, 59-75. Smith, R.E., Smoll, F.L., & Hunt, E. (1977a). A system for the behavioral assessment of athletic coaches. Research Quarterly, 48, 401-407. Smith, R.E., Smoll, F.L., & Hunt, E. (1977b). Training manual for the Coaching Behavior Assessment System. JSAS Catalogue of Selected Documents in Psychology, 7, 2 (Ms. no. 1406). Smoll, F.L., & Smith, R.E. (1989). Leadership behaviors in sport: A conceptual model and research paradigm. Journal of Applied Social Psychology, 19, 1522-1551. Smoll, F.L., & Smith, R.E. (2002). Coaching behavior research and intervention in youth sports. In F.L. Smoll & R.E. Smith (Eds.), Children and youth in sport: A biopsychosocial perspective (2nd ed., pp. 211-234). Dubuque, IA: Kendall/Hunt. Smoll, F.L., Smith, R.E., Barnett, N.P., & Everett, J.J. (1993). Enhancement of children’s self-esteem through social support training for youth sport coaches. Journal of Applied Psychology, 78, 602-610. Smoll, F.L., Smith, R.E., & Hunt, E. (1978). Toward a mediational model of coach-player relationships. Research Quarterly, 49, 528-541.

Acknowledgment This research was supported in part by Grants 1066 and 2297 to Ronald E. Smith and Frank L. Smoll from the William T. Grant Foundation. Manuscript submitted: March 16, 2005 Revision accepted: February 17, 2006