Gender bias: a study of high school track & field athletes' perceptions of hypothetical male and female head coaches.
Gender bias toward female coaches may be derived from many different levels within the coaching environment, ranging from hiring practices of athletic directors to athletes' perceptions. The prevalence of gender bias in the hiring of coaches has been well documented in previous studies (Hasbrook, Hart, Mathes, & True, 1990; Stangl & Kane, 1991).
Another form of gender bias, specifically that of athletes' perceptions of their coaches, was the focus of several past investigations. Williams and Parkhouse (1988) studied the magnitude of gender bias exhibited by female high school varsity basketball players. Subjects were broken up into four groups based upon the gender of their coach and team success. Each group was asked to indicate their preference between a hypothetical male or female coach who was classified as either successful or unsuccessful, based on their team's win/loss record. In every situation, all groups showed a preference for the male coach except when given the choice between an unsuccessful male coach and a successful female coach. Even in this scenario, 40% of the female athletes still preferred the unsuccessful male coach.
In a study designed to examine attitudes of male and female athletes toward male and female coaches, Weinberg, Reveles, and Jackson (1984) used a sample of male and female junior high school basketball players, high school basketball players, and college basketball players. All the male athletes played for male coaches, and all the female athletes played for female coaches. The sample was divided into male and female athletes and each group was asked to indicate their attitudes and feelings towards an equally qualified hypothetical male or female coach. The male athletes generally exhibited more negative attitudes towards female coaches than did female athletes, while both groups did not differ in their view of male coaches.
In a more recent study, Medwechuk and Crossman (1994) investigated the effects of gender bias on the evaluation of male and female swim coaches who vary in status. They reported that both male and female swimmers did not differ in their evaluation of male coaches' ability to motivate, and male coaches' potential for future success. Male coaches were rated as equal regardless of their low or high status. Unlike the basketball players in the Weinberg et al., (1984) and the Parkhouse and Williams (1986) studies, the swimmers in the Medwechuk and Crossman (1994) study exhibited a clear preference for a same sex coach.
The Weinberg et al., (1984) and the Parkhouse and Williams (1986) studies have focused on high school basketball, which is an interactive sport activity, and thus, may not be applicable to individual sports, such as swimming, track and field, gymnastics, etc. which constitute coactive sport activities. The coaching arrangement of track and field, for example, differs from basketball and other team sports in that the same coach may often be responsible for both the boys' and girls' teams, whereas team sports usually have separate coaches for each program. In individual sports, such as track and field, it is not uncommon that one coach would be directing both the male and female programs. Combined programs in which a female head coach is directing both the male and female track and field teams exist throughout Southern California at the high school level.
Past studies explored situations in which female athletes were coached by either male or female coaches, and situations in which male athletes were coached exclusively by males. The combined program, which track and field affords, allows for studies in which attitudes about coaching sports where male athletes that are coached by female coaches may also be studied. This type of data is practically non-existent. In a combined program, with one head coach, both male and female athletes, who are members of the same team, and coached by the same coach, can be surveyed to measure any gender bias when they assess male and female coaching competence. The ability to compare the responses of male and female athletes who are coached by males, with the responses of those athletes who are coached by females, may provide additional insight into the dynamics of gender bias.
The purpose of the present study was to test the effects of athletes' gender and their coach's gender on gender bias. In addition, the generality of the findings reported by Medwechuk and Crossman (1994), and Weinberg et al. (1984) to high school track & field programs were explored. based on the studies discussed earlier the first hypothesis was that male track and field athletes, whether coached by a male or a female, will express greater negative gender bias toward a hypothetical female coach (described as equally qualified to her male counterpart) when compared to female high school track and field athletes. Second, it was predicted that male high school track and field athletes coached by male coaches will voice greater negative gender bias toward a hypothetical female coach (described as equally qualified to her male counterpart) when compared to male high school track and field athletes coached by a female. Finally, it was predicted that female high school track and field athletes coached by male coaches will support greater negative gender bias toward a hypothetical female coach (described as equally qualified to her male counterpart) when compared to female high school track and field athletes who are coached by a female.
Two hundred and sixteen (112 male & 104 female) high school track and field (33% throwers and 67% runners) athletes (freshmen to senior standing, ages 15-18) in the Los Angeles and Riverside counties of the California Interscholastic Federation Southern Section took part in this study. The ethnic makeup of the participants was evenly distributed among the studied high school teams and included Afro-Americans (21%), Asian/Pacific Islanders (19.7%), Caucasians (18%), and Latino/Latina (41%). The freshmen respondents who had little experience with their high school track coach at the time of data collection were evenly distributed among the groups studied.
Participants were selected from eight high school programs that were matched paired and broken into two groups based on the gender of the head coach, program location, size and ethnic makeup of the school, win/loss record, and event coaching specialties of the head coach for the 1994 season. The eight studied programs were evenly split into four teams having male head coaches, and four teams having female head coaches. Schools chosen for this study had (a) a coach that led both male and female athletes in the same event areas, and (b) had been the head track coach at the school for at least two seasons. All athletes that were coached in their event areas by the head coach were surveyed.
A modified version of the Weinberg et al., (1984, p.450) "Attitudinal Questionnaire" was utilized for the identification of gender bias in the sample. Since the original text in Weinberg et al.'s questionnaire was addressing team sport athletes (basketball) changes were required to make the language more appropriate for individual sport athletes (track & field). For example, the statement in Weinberg et al., "He could make me want to play better," was modified in this study to, "He (or she) could motivate me to perform better." The statement, "He might be a head coach in 20 years" was modified to "He (or she) might have a future as head coach" to better fit the perspective of a high school athlete. To insure internal validity, questions three, five, and nine were phrased in such a way as to ask essentially the same question. Respondents who contradicted themselves when answering any combination of these questions, had their questionnaire deemed unusable. Participants' answers were rated on a 1 (not at all) to 11 (very much) Likert scale.
Participants were told that their cooperation was appreciated for the completion of a study about their reactions and attitudes toward a hypothetical coach. The athletes were informed that neither the investigation nor their responses had anything to do with the evaluation of their current coach and were instructed to react to the ten questions based on how they felt about a hypothetical coach ("Michael" or "Mary" both described as very successful recent college track competitors with Masters degrees specializing in the scientific basis of coaching) in the imaginary event that he/she were to become the respondent's coach for next season. In addition, participants were asked to rely on their first impressions when rating the statements, to be as honest as possible in their reactions, and were assured that their responses would be strictly confidential.
Test reliability was secured by the utilization of the same investigator for the collection of all data and the strict following of the procedures described hereafter. The administration of the questionnaire was performed at the track and field practice site of the participating high schools. The questionnaires, and a supply of pencils, were brought to the practice site on a predetermined date that was arranged with the head coach. Materials were distributed from separate folders that contained a presorted set of questionnaires that asked for an evaluation of either a hypothetical male or female coach. One folder contained questionnaires destined for female subjects, while the other folder contained questionnaires for male subjects. This method was designed to insure that both male and female subject groups received an equal number of questionnaires asking them to evaluate either a hypothetical male or female coach, while giving the appearance that the researcher was distributing the questionnaires randomly.
The participants were asked to individually turn in their completed questionnaires and were instructed that questionnaires returned in groups would not be accepted. The survey dates ranged from the middle to the end of the 1994 track and field season. A questionnaire was deemed unusable if filled out incorrectly (e.g., subject circled more than one number on the 11 point spread Liken scale for one item) or if any evidence of patterned response or contradictory answering patterns were detected. Patterned response was defined as a series of identical responses to each question, such as giving a rating of 11 for all ten questions, causing the subjects to contradict themselves when giving their evaluation (following the above guidelines fifteen questionnaires or approximately 5% were deemed unusable).
Respondents were first split into two groups based on the gender of the participant. Within each original group two more groups were formed based on the gender of the hypothetical coach that was evaluated by each athlete. This pairing yielded four groups based on the gender of the subject and the gender of the hypothetical coach being evaluated. Means and standard deviations were calculated for the ten questions for each group, and one-way and two-way ANOVAs were performed to study the main effects and interaction effects of the gender of subject and gender of the hypothetical coach for each question. Differences between groups due to interaction effects were described by comparing the means (see Table 1) as recommended by Thomas and Nelson (1990, p. 152). This analysis was a replication of the study by Weinberg et al., (1984).
An extension of the Weinberg et al., (1984) study was performed by splitting the previously mentioned groups into groups based on the gender of the athlete's coach. This arrangement provided eight groups determined by the gender of the athlete, gender of the athlete's coach, and the gender of the hypothetical coach that was evaluated. A table of means and standard deviations were calculated for each of the four groups based on participant's gender and the gender of her/his coach (see Table 2). Means and standard deviations were calculated for each group for the ten questions and a three-way ANOVA was performed to study the main effects of the participant's gender, gender of her/his coach, and gender of the hypothetical coach ("Michael" or "Mary") for each question.
Summary of Respondents
Of the 231 completed questionnaires 15 were deemed unusable due to respondents' failure to closely follow the given instructions. The analyzed sample included a total of 216 usable questionnaires, or approximately 95% of the original data (112 males and 104 females).
[TABULAR DATA FOR TABLE 1 OMITTED]
[TABULAR DATA FOR TABLE 2 OMITTED]
A Comparison of Subjects' Responses by Gender of Subject and Gender of Coach
The responses from the questionnaires were tallied and broken down into four groups of means and standard deviations based upon the gender of the subjects answering the questionnaires and the gender of the coaches being evaluated within the questionnaire (see Table 1). A one-way ANOVA revealed a significant main effect, F(1,205) = 4.16, p [less than] .04, for question number six, "I might feel angry (mad) if he/she yelled at me," that was dependent on the hypothetical coach's gender. A significant interaction effect was also found between the gender of the subject and the gender of the hypothetical coach, F(1,205) = 4.65, p [less than] .03.
Females (see Table 1) displayed the most negative feelings about "being yelled at" by a female coach (M = 7.0) and the least negative feelings about "being yelled at" by a male coach (M = 5.2). Of the eight significant differences between male and female perceptions in the Weinberg et al., (1984) study, only the "...yelled at" statement was replicated in the present study.
Responses were further analyzed by a 2 x 2 x 2 (gender of subject by gender of hypothetical coach by gender of subject's coach) ANOVA. Results indicated that there were significant main effects for questions one ("like as coach"), two ("perform better"), three ("something wrong"), and four ("confidence to improve"). The main effect revealed a significant difference in perceptions of a new hypothetical coach that was dependent on the gender of the athlete's actual coach. A two-way ANOVA (gender of athlete by gender of her/his coach) revealed a significant interaction (p [less than] .04) for question #3 "...done something wrong" (see Table 3).
The purpose of the present study was to test the effects of the athlete's gender, and her/his actual coach's gender on her/his evaluation of a new, equally qualified, hypothetical male or female coach. In addition, the generality of the findings reported by Medwechuk and Crossman (1994), and Weinberg et al. (1984) to high school track & field programs was explored.
An analysis of the effect of the gender of the subject's coach upon the subject's attitudes toward a hypothetical new coach revealed that the athletes of male coaches responded more positively to a new coach than athletes who were coached by females. This fact was true for four of the ten questionnaire items. Results also indicated a significant interaction effect between gender of athlete and gender of actual coach thus confirming the findings by Medwechuk and Crossman (1994). The first interaction effect revealed that male athletes coached by males were more receptive of being told that they had done something wrong than females coached by either a male or a female. A second interaction effect revealed that male [TABULAR DATA FOR TABLE 3 OMITTED] athletes coached by females were less receptive of being told they had done something wrong when compared to the other groups.
Males and females coached by males had more positive attitudes toward a hypothetical new coach than did subjects coached by females. This difference in attitudes was demonstrated in athletes' degree of liking of a new hypothetical coach, their estimate of whether that coach could get them to improve or motivate them to perform better, and whether they could take it when that coach told them that they had done something wrong.
One possible explanation of the discrepancy in attitudes would be that an actual difference exists between the way subjects of male coaches feel toward a new coach when compared to subjects of a female coach. This interpretation posits that the gender of the subject's coach plays a role in facilitating these differences. This conclusion does not seem plausible because it implies that all female coaches must do something different, when compared to all male coaches, to facilitate the discrepant feelings displayed by their athletes. By comparing groups of subjects from two different clusters of schools, it is more likely that differences in the responses of the subjects may be due to the differences between the high schools and their respective track and field programs, than the differences facilitated by the gender of the coach.
Of the eight Weinberg et al., (1984) questions that demonstrated significant differences in the way male and female athletes evaluated a hypothetical coach, only one question (Q #6 "Angry if yelled at...") turned out to be significant in this study. The replication of only one finding of Weinberg et al., raises the question as to why these results couldn't be reproduced, and what conditions existed that did not allow for replication. The present study utilized a similar questionnaire and methodology to that used by Weinberg et al. (1984). However, several differences between the two studies still remain. The first major difference is the fact that the present study was conducted ten years later than the study by Weinberg and colleagues. Within this ten year span, some of the accepted norms about the perceptions and ability of female coaches may have changed. It is quite likely that the increased female participation may have facilitated a change in attitudes toward the acceptability of females taking a leadership position in sports. Results of the present study with track athletes and the Medwechuk and Crossman (1994) study of swimmers seem to reflect such a trend. Another important distinction between the two studies was the difference in sampling procedures. Great care was taken in the present study in match pairing the high schools from which the male and female subjects came from by controlling for size, ethnic makeup, location of school, and the level of success of the track and field program. Since Weinberg et al., did not mention any special provisions for sampling procedures some of their findings may have been determined by factors other than bias over the perceived ability of the hypothetical coach.
The fact that gender bias by male track and field athletes toward female track and field coaches was not demonstrated in this study, may also be explained by the nature of the sport under investigation. When compared with the sport of basketball, which has been the sport of choice in which to conduct gender bias research in the past, track and field and swimming appear to be a more acceptable activity for females to participate in by virtue of their noncontact task structure. In addition, track and field and swimming are individual unitary or self-paced (closed-skills) sport activities, in which athletes are encouraged to compete with themselves rather than against each other. Track athletes as well as swimmers are also expected to have a more vital role in their own planning of personal objectives and training programs. Basketball players, on the other hand, are engaged in an interactive, open-skills activity which takes place in an environment that is closely supervised. Thus, the basketball coach may be expected to exhibit a more autocratic leadership style, which is thought to be a masculine quality (Staurowsky, 1990).
Another difference in coaching styles between basketball and track and field, may be the way in which performance differentials between athletes are measured. Field event marks or performance times are a relatively objective measurement of achievement. On the other hand, overall basketball performance is a more complex combination of defensive, offensive, and leadership skills. Thus, differences between basketball players' performance on the court may be subjectively judged and decided by the coach. The possibility that basketball players may nol readily agree with the coach's evaluation of their ability may make it harder for male basketball players to accept an apparently subjective judgement from a female figure hypothetical coach. This interpretation could also explain why more recent related studies (Parkhouse and Williams, 1986; Williams and Parkhouse, 1988) that used basketball players as subjects, reported gender bias against female coaches of successful professional status.
The present study did, however, replicate one finding reported by Weinberg et al. (1984). Female subjects displayed more negative feelings toward female coaches than did male subjects with regard to whether they would be angry if yelled at by a hypothetical coach. This finding suggests that females may not be as accepting of a female authority figure in this context. In addition, results indicate that females were more accepting of a male coach who yelled at them which suggests that the female athletes were more comfortable with a male authority figure in that situation. Male responses did not vary in their perceptions of a male or female hypothetical coach with regards to how angry they would feel when yelled at.
Practical applications for this finding may be used by female coaches who are training female athletes. Since this finding was supported by studies of both track and field, and basketball athletes, coaches from related sports may also benefit from these results. Female coaches need to be aware of the differences in the way male and female high school athletes may react if yelled at. The significance of this finding as it pertains to relationships between coaches and athletes and athletes' performance is subject for further study.
Both the present and the Medwechuk and Crossman (1994) studies provide preliminary evidence of a trend toward a more balanced view of the female coach. Subsequent investigations utilizing rigorous sampling procedures still need to be implemented in order to establish whether this trend is also present in team sports. The actual coach's leadership style and its effects on gender bias in her/his athletes would be an additional important venue for further investigation of gender bias among athletes.
Hasbrook, C. A., Hart, B. A., Mathes, S. A., & True, S. (1990). Sex bias and the validity of believed differences between male and female interscholastic athletic coaches. Research Quarterly for Exercise and Sport, 61, 259-267.
Medwechuk, N., & Crossman, J. (1994). Effects of gender bias on the evaluation of male and female swim coaches. Perceptual and Motor Skills, 78, 163-169.
Parkhouse, B. L., & Williams, J. M. (1986). Differential effects of sex and status on evaluation of coaching ability. Research Quarterly for Exercise and Sport, 57, 53-59.
Sisley, B. L., & Capel, S. A. (1986). High school coaching filled with gender differences. Journal of Physical Education, Recreation, & Dance, 57, 39-43.
Stangl, J. M., & Kane, M. J. (1991). Structural variables that offer explanatory power for the under-representation of women coaches since Title IX: The case of homologous reproduction. Sociology of Sport Journal, 8, 47-60.
Staurowsky, E. J. (1990). Women coaching male athletes. In M. A. Messner, & D. F. Sabo (Eds.), Sport, men, and the gender order: Critical feminist perspectives (pp. 163-169). Champaign, Illinois: Human Kinetics.
Thomas, J. R., & Nelson, J. K. (1990). Research methods in physical education. Champaign, Illinois: Human Kinetics.
Weinberg, R., Reveles, M, & Jackson, A. (1984). Attitudes of male and female athletes towards male and female coaches. Journal of Sport Psychology; 6, 448-453.
Williams, J. M, & Parkhouse, B. L. (1988). Social learning theory as a foundation for examining sex bias in evaluation of coaches. Journal of Sport and Exercise Psychology, 10, 322-333.
|Printer friendly Cite/link Email Feedback|
|Author:||Frankl, Daniel; Babbitt, Donald G., III|
|Publication:||Journal of Sport Behavior|
|Date:||Dec 1, 1998|
|Previous Article:||Leader Member Exchange Theory and sport: possible applications.|
|Next Article:||Personality characteristics of competitive and recreational cyclists.|