Investigating leadership, gender, and coaching level using the Revised Leadership for Sport Scale.

Jambor, Elizabeth A. ; Zhang, James J.

Research often investigates the leadership styles of coaches (Chelladurai, 1978; Chelladurai & Carron, 1978; Fry, Kerr, & Lee, 1986; Offermann & Bell, 1992). These investigations look at personalities, behaviors, and other related variables in an attempt to better understand and define leadership. Chelladurai and Carron (1978) developed a multidimensional model to better define leadership. Within this model, these researchers provided three states of leadership behaviors (required, preferred, and actual) that were associated with group performance and member satisfaction. By using this leadership model, one can investigate the multidimensional aspects of leadership by including the environment (the setting, the competitive level), the group members and their characteristics, and the personal characteristics of the leader (gender, age, psychological characteristics). It is this multifaceted focus of leadership that allows research to better understand leadership issues.

Researchers provide contradicting evidence for the differences between male and female leaders. Freeman and Lanning (1989) demonstrated how males and females are similar in social power motivation (an element of leadership). Conversely, Chelladurai and Saleh (1980) found that male athletes preferred coaches to be more autocratic, yet more supportive than did female athletes. Following Chelladurai and Saleh's (1980) research on preferred leadership behavior, suggesting that behavior is dictated by member's preferences (athlete's attitudes), it would follow that coaches of males (typically male coaches) would be different than coaches of females (typically female coaches). Additionally, Lipman-Blumen (1992) reported that gender differences exist in leadership achievement styles. At present, the literature argues as to whether males and females differ or whether leadership models need re-evaluation to include a broader range of variables and behaviors (Lipman-Blumen, 1992).

Within the scope of required leader behavior and actual leader behavior, differences might exist between coaches at different levels. Chelladurai and Carron (1978) suggested that leader behavior occurs as the result of organizational constraints (required behavior) and personality, ability, and experience (actual behavior). It would follow that coaches at the primary and secondary school levels would differ from each other just as they would differ from coaches at the collegiate level. The organizational constraints and experience levels would dictate differences in leadership behavior.

Chelladurai and Saleh (1980) investigated five leadership behaviors: training and instruction, democratic, autocratic, social support, and positive feedback. These behaviors were used to measure coaches' perceptions of their own behavior. However, the situational leadership theory states that leaders should vary behaviors related to members (Hersey & Blanchard, 1977). Situational behavior is not present in the Leadership for Sport Scale (LSS) (Chelladurai & Saleh, 1980). Zhang, Jensen, and Mann (1996) have taken the investigation of leadership a step further in their development of the Revised Leadership Scale for Sport (RLSS), modeled after Chelladurai and Saleh's (1980) LSS. The RLSS defines six different behaviors of leadership: training and instruction; democratic; autocratic; social support; positive feedback; and situation consideration. The RLSS adds situation consideration to the variables used in the LSS to provide another behavior in which leaders may operate. Leaders can be described based on the situation in which they find themselves, offering more flexibility to the definition. However, beyond the research by Zhang, Jensen, and Mann (1996) in the development of the RLSS, little has been done to test its use with various populations.

In an attempt to extend the knowledge of leadership in sport, the purpose of this study was to determine possible differences in leadership behaviors, using the RLSS, between male and female coaches and among different coaching levels. The present study focused on the three states leadership behaviors, suggested by Chelladurai and Carron (1978), in male and female coaches among various levels of coaching. The scales by Chelladurai and Saleh (1980) and Zhang, Jensen, and Mann (1996) include methods for assessing athletes' perceptions and preferences and coaches' perceptions. For this study, only the coaches' perceptions, as measured by the RLSS, were used. First, it was hypothesized that male and female coaches would respond differently to the RLSS given the six dimensions in which a leader may exhibit behaviors. This hypothesis was based on Chelladurai's (1980) contention that certain factors (i.e., gender) set the boundaries for leadership behaviors. This relates to Chelladurai's required leader behavior within the multidimensional model of leadership. Second, it was hypothesized that differences on the RLSS would occur among the coaching levels: junior high school, high school, and college. This relates to the preferred leader behavior within Chelladurai's multidimensional model of leadership, in that situational characteristics influence preferred behavior.

Methods

Participants

The participants were 162 coaches (male = 118; female = 44) at the junior high school (n = 25), high school (n = 99), and college level (n = 38). All participants participated on a volunteer basis only.

Instrument

The Revised Leadership Scale for Sport (Zhang, Jensen, & Mann, 1996) was used to measure six leadership behaviors. The behaviors were:

1. Training and Instruction: planning training practices and evaluating the performance of the athletes; having knowledge and being responsible.

2. Democratic: encouraging involvement of the athletes; admitting mistakes and confronting problems.

3. Autocratic: making independent decisions; using commands and punishment.

4. Social Support: helping athletes with personal problems; making sport part of enjoyment of an athlete's life.

5. Positive Feedback: encouraging an athlete after making a mistake; correcting the behavior rather than blaming the athlete.

6. Situation Consideration: setting up individual goals and clarifying ways to reach goals; differentiating coaching methods at different maturity stages and skill levels.

The 60 statements were preceded by "In coaching, I:" and followed by a Likert scale of 1-5:1 = never; 2 = seldom; 3 = occasionally; 4 = often; and 5 = always. Each of these relates to the percentage scale (0-100%) used by on the Leadership Scale for Sport (Chelladurai & Saleh, 1978).

Procedures

Scales were given in a variety of settings: classrooms, gymnasiums, practice fields, and offices. Participants were asked to respond honestly to each statement, describing their own coaching behavior. Confidentiality was stressed so that the coaches might feel more at ease in responding. Participants were informed, after completing the scale, as to the nature of the study. Any participant who requested feedback from the results of their responses was provided with that feedback.

Results

Reliability Measures for the RLSS

Reliability was tested for each behavior section of the RLSS to lend support to the following analyses. The internal consistency for each section was: .84 for training and instruction; .66 for democratic; .70 for autocratic; .52 for social support; .78 for feedback; and .69 for situational consideration. These measures of reliability were believed to be significant enough to use the RLSS in investigating possible gender differences in leadership behavior.

RLSS and Gender

A MANOVA was used to determine the existence of significant differences between male and female coaches when looking at the leadership behaviors in total. Results indicated no significant differences between male and female coaches ([Lambda] = .77; p [less than or equal to] .23). Male and female coaches gave similar responses for the combined six leadership behaviors.

When examining the six behaviors individually, males and females differed on only one leadership behavior: social support (F (1,160) = 4.47; p [less than or equal to] .05). Females scored significantly higher on this behavior than did the males, indicating a greater tendency for helping athletes with personal problems and making sport part of the enjoyment of an athlete's life.

RLSS and Coaching Level

A MANOVA was used to determine the existence of significant differences between three different levels of coaching (junior high school, high school, and college) when looking at the leadership behaviors in total. Results indicated significant differences between the three coaching levels ([Lambda] = .61; p [less than or equal to] .02). The three different levels of coaching did not respond similarly to the six leadership behaviors of the RLSS.

When examining the six behaviors individually, the three coaching levels differed on three behaviors. Democratic leadership behaviors were significantly different (F(2, 158) = 4.14; p [less than or equal to] .009). As indicated by a Scheffe post hoc test, college coaches and high school coaches were significantly different from one another with high school coaches indicating a higher degree of democratic behaviors. Training and instruction behaviors were significantly different (F(2, 158) = 3.46, p [less than or equal to] .02). As indicated by a Scheffe post hoc test, junior high school coaches were significantly lower in training and instruction behaviors than were the high school and college coaches. Finally, social support behaviors were significantly different (F(2, 158) = 3.66, p [less than or equal to] .02). As indicated by a Scheffe post hoc test, junior high school coaches reported significantly less social support behaviors than did high school and college coaches.

RLSS, Gender, and Coaching Level

A MANOVA was used to determine the possible interaction of gender and coaching level when looking at the leadership behaviors in total. Results indicated no significant interactions ([Lambda] = .66; p [less than or equal to] .06). For the present sample, the combination of gender and level of coaching was not reflected in self-reported leadership behavior.

Discussion

The results of this study did not support the first hypothesis that male and female coaches would respond differently to the behaviors of the Revised Leadership for Sport Scale. Differences in preferred coaching style is often related to an athlete's gender (Chelladurai & Saleh, 1980; Zhang & Smith, 1994), yet does not distinguish between the genders of coaches. The only behavior in which differences did exist, when viewing the six leadership components separately, was the social support behavior. Social support behavior was denoted as helping athletes with personal problems and making sport port of enjoyment of an athlete's life. In the present study, female coaches reported providing more support in an athlete's personal life and making sport more enjoyable. This finding relates to earlier work by Salminem, Liukkonen, and Telama (1990). These researchers reported female coaches to be more supportive and rewarding than male coaches.

The present research supports previous studies that have concluded that males and females are not inherently different in their leadership styles, but rather that a new model of leadership is needed to define leadership behaviors. Lipman-Blumen (1992) suggests that the traditional masculine leadership model is not enough for the integrated world. This idea of integration is especially true within the athletic world. To say that a male is a different leader than a female is an out-dated method of leadership evaluation. Additionally, as many investigators of leadership have suggested, the members and the situation play important roles in defining leadership (Chelladurai, 1990; Chelladurai & Carron, 1978). This lack of gender differences additionally supports Chelladurai's (1990) multidimensional model of leadership which includes leader characteristics yet not gender. Therefore, one should not expect gender differences, as supported by the present results.

The second hypothesis, differences on the RLSS would occur among the three different coaching levels, was supported by the results. The coaching levels (junior high, high school, and college) were significantly different when viewing leadership in total. When investigating leadership in its six components, the coaching levels differed on three behaviors: democratic, training and instruction, and social support. Democratic leadership behavior was defined as encouraging involvement of the athletes, admitting mistakes, and confronting problems. Training and instruction leadership behavior was described as planning training practices and evaluating the performance of the athletes, as well as having knowledge and being responsible. Social support leadership behavior was defined as helping athletes with personal problems and making sport part of the enjoyment of an athlete's life.

Within the area of democratic leadership behavior, high school (HS) coaches reported this behavior to a significantly higher degree than did college coaches. Perhaps the differences within each environment, where college athletes are at a different level of involvement than are the HS athlete, requires differences in leadership.

When looking at training and instruction, junior high school (JH) coaches indicated this behavior at a significantly lesser amount than did HS and college coaches. One explanation might be the "feeder system" used within the school environment in which the study took place. In the feeder system, coaches at the JH level are often directed to train athletes in a manner similar to that of the HS at which the JH athlete will attend. Because much of the training and instruction at the JH level is directed by the HS level, JH coaches may not perceive themselves as having the leadership behavior defined as training and instruction. College coaches are vastly different than JH coaches in that their success is dependent upon their own training and instruction. Further, college coaches within this study had greater autonomy in developing and evaluating performances than the JH coaches within the feeder system.

As with the training and instruction leadership behavior, JH coaches reported significantly less social support leadership behavior than did HS and college coaches. With the differing commitments at the three levels, JH coaches may not have the additional time to interact with athletes on a personal level. JH coaches in this study had greater teaching commitments and athlete to coach ratios than did the HS or college coaches. This lack of time may impact the amount of involvement coaches have with their athletes.

The differences found at the various coaching levels supports earlier research by Chelladurai (1990) who suggested that leadership is only significant within the context of the group. Further, leaders are required to behave in relation to the demands and parameters of the environment. The members of the different athletic environments and the environments themselves vary greatly. What is an important, or vital, behavior in one athletic environment may not be in another. Additionally, there may be some outside influences which directly, or indirectly, impact leadership that are not measured by the present study. The RLSS, however, was able to distinguish between the coaching levels, as did the LSS (Chelladurai & Saleh, 1978), but not discriminate between the genders. The present study provides support for use of the RLSS in future leadership studies. Future research using the RLSS might focus on possible differences among sports at the same competitive level. As suggested by Cribben (1981), the group's perception of the leader is more important than the effectiveness of the leader's behaviors. Chelladurai (1990) put forth that situational characteristics and the environment impact a leader's behavior. The environment of one sport may be much different than another and require different behaviors from its leaders. Differences in athletes and their respective sports may better facilitate and be facilitated by specific leadership behaviors. Perhaps differences in leadership behavior are influenced by the specific sport environment in which the leadership takes place.

Other lines of research might focus on the personality traits associated with specific leadership behaviors. Chelladurai (1990) suggests that along with ability and experience, personality has a strong influence on leader behavior. Certain personality characteristics might enable a leader to succeed in one type of environment and fail in another. Research is needed to provide a better understanding of the leadership-personality link. Finally, research using the RLSS is needed to investigate athletes' perceptions and preferences in relation to the coaches' perceptions of leader behavior. This has been successfully done with the LSS (Chelladurai, 1978; Chelladurai & Carron, 1978). To provide more support for the RLSS, a better understanding of the perceptions and interactions of coaches and athletes is required.

The present study also supports the use of the RLSS as a viable measurement of leadership behavior. It does not discriminate between genders, yet is able to distinguish between environments. It can be a useful tool in the defining and understanding of leadership behavior in the sporting world.

References

Chelladurai, P. (1990). Leadership in sports: A review. International Journal of Sport Psychology, 21, 328-354.

Chelladurai, P. (1978). A contingency model of leadership in athletics. Unpublished doctoral dissertation, Department of Management Sciences, University of Waterloo, Canada.

Chelladurai, P., & Carron, A. V. (1978). Leadership. Ottawa: Sociology of Sport Monograph Series, Canadian Association for Health, Physical Education, and Recreation.

Chelladurai, P., & Saleh S. D. (1980). Preferred leadership in sports. Journal of Sport Psychology, 2, 34-45.

Cribben, J. J. (1981). Leadership: Strategies for organizational effectiveness. New York: AMACOM.

Freeman, B., & Lanning, W. (1989). A multivariate analysis of the relationship between social power motivation and personality characteristics in college students. Journal of College Student Development, 30(6), 522-527.

Fry, L. W., Kerr, S., & Lee, C. (1986). Effects of different leader behaviors under different levels of task interdependence. Human Relations, 39(12), 1067-1082.

Hersey, P., & Blanchard, H. K. (1977). Management of organizational behavior. (3rd ed.). Englewood Cliffs, NJ: Prentice-Hall.

Lipman-Blumen, J. (1992). Connective leadership: Female leadership styles in the 21st-century workplace. Sociological Perspective, 35(1), 183-203.

Offermann, L. R., & Beil, C. (1992). Achievement styles of women leaders and their peers: Toward an understanding of women and leadership. Psychology of Women Quarterly, 16(1), 37-56.

Salminem, S., Liukkonen, J., & Telama, R. (1990, July). The differences in coaches' and athletes' perception of leader behavior of Finnish coaches. Presented at the AIESEP Congress, Loughborough, England.

Zhang, J. J., Jensen, B. E., & Mann, B. L. (1996). Modification and revision of the Leadership for Sport ScaPe. Journal of Sport Behavior, 19(4).

Zhang, J. J. & Smith, D. W. (1994, November). An investigation on leadership qualities necessary for effective intercollegiate athletic administration. Presented at the annual meeting for the Texas Association for Health, Physical Education, Recreation, And Dance, Austin, TX.

For further information, please contact:

Elizabeth A. Jambor Dept. of HPER Texas Tech University P.O. Box 41121 Lubbock, TX 7409-3011