LARNet; The Cyber Journal of Applied Leisure and Recreation Research 

Sexual orientation and team cohesion in women’s intercollegiate basketball
(July 2006)
Stacey R. Altman, J.D.
Cheryl Estes, PhD
Felicia Tittle, M.S.
<>Contact information:
Stacey R. Altman, J.D.
Eastern Carolina University
Department of Exercise and Sport Science

151 Minges Coliseum
Greenville, NC 27858-4353
Phone: 252-328-2973 (voice) / 252-328-4654 (fax)

The purpose of this study was to examine how sexual orientation affected athletes’ team cohesion in women’s collegiate basketball. Thirty-five athletes from three colleges completed a modified Group Environment Questionnaire (GEQ). The athletes’ scores were similar on task cohesion. However, their scores differed on social cohesion with athletes that self-identified as lesbians scoring lower than those athletes that did not identify as lesbians. The perceived presence of non-self-identified lesbians affected how athletes viewed the closeness of the team as a whole. While literature supports that task, rather than social, cohesion predicts effect on team performance, social cohesion is important to reaching student development goals for many collegiate athletic programs. The present study was an initial, descriptive step toward filling a gap between scholarly discourse and perceptions regarding sexual orientation in women’s intercollegiate sports. The results, while not generalizable to other female collegiate athletes, are empirical evidence and they create a foundation for future studies that will be needed to explore and validate findings in this important area.
KEYWORDS: team cohesion, Group Environment Questionnaire (GEQ), sexual orientation, coaching, sport administration

In the world of women’s basketball, lesbian participants who have shared their sexual orientation have historically suffered significant consequences, such as reduced playing time, the loss of scholarships and being isolated by teammates and/or coaches (McKay & Martin, 2001; Schwartz, 1997). The consequences described above are largely a result of stakeholders’ (i.e. athletes, parents, coaches, administrators, and media professionals) decisions, which are made with personal bias and/or the desire to conform to perceptions of what other stakeholders’ beliefs are in relation to women’s sport (McKay & Martin, 2001; Lenskyj, 2003). This tangled web of bias and perceptions of others’ perceptions creates a sporting atmosphere that maintains societal expectations for women that are based on arbitrary and narrow definitions of femininity and physicality. In this atmosphere it is often thought that societal expectations, as either broadly or narrowly construed, cannot be met if lesbians are present or thought to be present on a team or in a sport (Cahn, 1994; Griffin, 1998; Lenskyj, 2003).

In addition to concern for meeting social expectations, stakeholders often cite a belief that lesbian participants negatively affect a team’s performance as a reason for their actions. For example, Griffin (1998) states some coaches call attention to changes in team dynamics related to homosexual behaviors (including dating teammates) and prohibit participation by athletes who identify as lesbians for that reason. She goes on to explain that the same coaches do not find the changes in team dynamics created by “two heterosexual players dating the same male football player” as much of a concern (p. 195). Lenskyj (2003) notes that in the context of administration, “many sport administrators make hiring and firing decisions based on women’s known or perceived sexual orientations, and the resulting poisoned work environment forces some lesbians to quit their jobs or abandon their hopes of an athletic career” (p. 5).  It is not clear, however, that these beliefs, or perceptions of expectations, related to athletes or other stakeholders in athletics, have been formed through careful reflection or informed by research.

While discourse regarding the complexities of sexuality is a relatively recent addition to sport literature, the challenge involved in understanding group dynamics in sport is well established. There are many group dynamics that take place within a team, and one of the most important is cohesion. A team’s ability to “gel” or “have good chemistry” is related directly to its performance (Widmeyer, Brawley, & Carron, 1985). The more cohesive the team is, the more the team, as a collective unit, encourages peak performance in all of its members; if cohesion is lacking, it is likely the team will not reach its potential (Sugarman, 1998). There are many factors that are thought to contribute to a team’s cohesiveness, including coach’s personality, coach’s leadership style, and even how much time the team spends socializing (Turman, 2003). It is common for the skill level of participants to be relatively equal in the arena of highly competitive sport, therefore the level of team cohesiveness often makes the difference between winning and losing (Thiese, 1999).

It is a common assumption that homosexuality detracts from team or unit cohesion. Much attention has been given to the notion in the context of military units (Barkawi, Dandeker, Wells-Petry, & Kier, 1999). In sport, less scholarly attention and popular press has been given to the topic. To date, qualitative and quantitative research has addressed homophobia and managing the lesbian “stigma” in sport (Anderson, 2005; Blinde & Taub, 1997), but no researchers have empirically examined whether or not sexual orientation affects factors related to team performance, including team cohesion. In fact, “sport psychologists have largely avoided, through scholarly discourse, examining lesbianism in sport, thereby perpetuating ‘the silence so loud that it screams’” (Vealey, 1997, p. 165).
The purpose of the present study was to empirically examine whether sexual orientation affected perceptions of team cohesion in National Collegiate Athletic Association (NCAA) Division II Women’s Basketball Teams. While the present study is exploratory in nature, it provides a necessary foundation for future empirical studies.

Review of Literature
When women’s sport is labeled a lesbian activity, or when athletes who identify as lesbians are stereotyped and threatened within sport, all female athletes are harmed because a narrowing of gender and sexual expression for all athletes occurs (Adams, Schmitke, & Franklin, 2005; Dowling, 2001; Griffin, 1998) but lesbian athletes are especially affected (Krane, 1996). As noted above, lesbian athletes who have shared their sexual orientation have suffered consequences and all female athletes have been subjected to strategies intended to present a narrowly defined feminine image (e.g. uniforms with short skirts, gender-marking, and the use of monikers such as “Kittens” for female team mascots) (Adams, et al., 2005; Messner, 2002).  These strategies also seek to eliminate the actual or perceived presence of lesbian athletes on the team. Issues affecting lesbian athletes can be best understood by exploring general attitudes toward homosexual athletes.

Attitudes toward homosexual athletes
A brief explanation of the concepts of stereotyping, homophobia, and homonegativism helps in understanding how the beliefs and perceptions of expectations emerge and are maintained. Distinguishing these concepts illustrates that the behaviors described above may be of an intentional or unintentional nature. This is important because the authors argue that studies like the present one can help eliminate unintentional discrimination through education.

Harrison (2001) describes stereotyping as a “process of imposing characteristics on people based on their perceived group membership” (p. 98). A stereotype, he explains, is a cognitive schema, or a way of organizing information based on past experience or reactions. Stereotypes, it is argued, “are encoded subconsciously in memory and retrieved automatically” (Harrison, p. 99). In other words, because of the commonness of many stereotypes, some are bound to find a place in most people’s minds. It then follows that under stress or pressure, even those who normally express nondiscriminatory values will revert to stereotyping (Harrison).
Certainly, females entering sport, historically a male domain, coupled with the requirement to produce revenue at elite levels of sport, qualify as stressors for participants, coaches, and administrators.

The cognitive process of stereotyping is linked to dichotomous thinking, or seeking to categorize with little or no tolerance for ambiguity, which is in turn linked to what Harrison (2001) called intellectual laziness. It is possible to infer then that mental laziness is a factor in perpetuating negative attitudes toward lesbian athletes.

Homophobia has been defined as the irrational fear and intolerance of homosexuality, gay men, lesbians, or bisexual people (Griffin, 1998). Researchers interested in women’s sport have given this issue considerable scholarly attention. The attitudes and actions associated with homophobia have been reported to be a factor in media representations of female athletes (Cotton & Jackson, 1992; Knight & Giuliano, 2003; Kreigh & Kane, 1997), in job selection and satisfaction (Greendorfer & Rubinson, 1997; Lenskyj, 2003; Wellman & Blinde, 1997), and in the overall quality of the sport experience for women who self-identified as lesbians and women that did not self-identify as lesbians (Cahn, 1994; Griffin, 1998; Hargreaves, 2000; Nelson, 1994). However, attention to team performance issues, especially those related to psychology, and homophobia has been relatively absent.

The concept of homonegativism refers to the purposeful or systematic oppression of those who challenge traditional notions of sexuality and gender roles (Krane, 1997). This concept adds an important dimension to the discussion because it includes the view that actions are predicated on intent (suggesting rational and strategic behavior) to harm homosexuals rather than on irrational fear. It also indicts social institutions rather than just individuals for perpetuating discrimination.
When these concepts are thought of collectively, the complexity involved in changing attitudes and reforming actions in the conduct of women’s sport is apparent. To move toward this goal, specific myths about the impact of sexual orientation on image and performance factors, such as team cohesion, must be investigated. A few qualitative studies provide rich information regarding questions related to sexual orientation and team interaction. For example, Fasting (1998) found that teams with a lesbian presence were over all friendly and accepting of one another independent of sexual orientation. However, Fasting also found many athletes expressed negativity toward lesbian athletes who shared or flaunted their sexuality. These athletes, both those that identified as lesbian and those that did not, were particularly concerned about being unable to recruit young players whose parents objected to the team or sport image and/or environment. Other than this recruiting concern, there was no concern expressed by the players in connection with the impact on other team performance indicators (Fasting).

Theberge’s (2000) ethnographic study of women’s hockey revealed that athletes did not perceive the presence of lesbians on their team or in the league as having a negative affect on team performance or social interaction. In fact, the perception was that welcoming lesbian players enhanced performance, although team cohesion, as a performance measure or predictor, was not specifically considered.
Further study of these complex questions from an empirical perspective can build on the theoretical foundations established and will add valuable insight regarding performance factors like team cohesion.

Team Cohesion
Cohesion is intuitively seen as an important factor in sport team performance and most coaches devote considerable effort to enhancing cohesion in their teams. Team cohesion has been defined as a multidimensional construct, which includes group goals and objectives as well as development and maintenance of positive interpersonal relationships (Yukelson, Weinberg & Jackson, 1984). Team cohesion is a basic requisite for group maintenance, and without group maintenance a team may have difficulty attaining goals and objectives (Widmeyer et al., 1985).
Carron’s (1982) conceptual model illustrated an important distinction between the two dimensions of team cohesion: task and social. Task cohesion concerns achieving a group’s tangible goals and objectives, including productivity, work output, performance effectiveness, winning games and other team successes. Social cohesion concerns satisfying members’ social and emotional needs such as friendship and affiliation. While these two dimensions are related, researchers have discovered that team members do not necessarily have to socialize with each other off the court in order to play well on the court (Mullen & Copper, 1994).

Measuring Team Cohesion
Widmeyer et al. (1985) completed a comprehensive literature review and compiled their list of suspected antecedents to team cohesion into three categories: characteristics of the group members, characteristics of the group, and situations experienced by the group. These categories were used as the basis for designing the Group Environment Questionnaire (GEQ).
The GEQ was designed to measure individual and group members’ perceptions of team cohesion and has proven to be a reliable and valid measure for team cohesion (Widmeyer et al., 1985). The GEQ has eighteen questions that assess four measures of cohesiveness: Individual Attraction to Group-Task, Individual Attraction to Group-Social, Group Integration-Task, and Group Integration-Social (see Figure 1).

Figure 1

The theoretical basis for these four subscales must be described in order to further subsequent discussion. Two subscales concern Individual Attractions to the Group, including both task and social dimensions. These two scales focus on the athlete’s self, and she responds based on her perceptions and feelings. The Individual Attractions to the Group-Task (ATG-T) subscale reflects the individual team member’s feelings about the attractiveness of the group task, the group’s productivity, and the group’s goals and objectives (for specific examples of what each subscale addresses, see figure 1). The Individual Attractions to the Group-Social (ATG-S) subscale reflects the individual team member’s feelings about the attractiveness of the group as a social unit and the social interactions within the group.

The other two subscales concern group integration, where the athlete-individual assesses the group as a whole by answering questions that measure her perceptions of the team related to its coherence around task and social activities. The Group Integration-Task (GI-T) subscale is a measure of the task oriented similarity, closeness and bonding within the team as a whole. The Group Integration-Social (GI-S) subscale is a measure of socially oriented similarity, closeness and bonding within the team as a whole.

Team Cohesion Studies
In addition to Widmeyer et al.’s (1985) foundational basis for effectively measuring team cohesion, three studies examined relationships between team cohesion and performance effects. Other researchers have utilized the GEQ to examine how other variables effected, or were affected by, a team’s cohesion.

Relating to team cohesion and performance effects, Mullen and Copper (1994) conducted a meta-analysis of studies from various sub-disciplines in psychology (e.g. industrial, sport, military, social). Beauchesne, Turner, Brennan, and Hoopengardner (1997) examined the cohesiveness of one women’s field hockey team in relation to several variables, including performance. Carron, Bray and Eys (2002) studied how team cohesion is related to team success.

Mullen and Copper (1994) grouped 49 team cohesion studies into two different paradigms: (a) correlational studies where members of intact teams perceptions of cohesiveness were correlated with performance, and (b) experimental studies where high and low levels of team cohesion were experimentally induced. Each paradigm had strengths and weaknesses. While the experimental paradigm allowed greater control over determining direction of influence, the correlational paradigm allowed for greater naturalism. For the correlational studies, cohesiveness-performance effect decreased as a function of interpersonal attraction (ATG-S and GI-S, both social subscales) and increased as a function of commitment to task (ATG-T and GI-T, both task subscales). Interpersonal attraction (i.e. friendship) and group pride did not make independent contributions to cohesiveness-performance effect for either the correlational or experimental studies. Within each paradigm, commitment to task emerged as the critical predictor of performance. Further, the task components were the strongest predictors of cohesiveness-performance effect in the experimental paradigm and the only predictors of effect in the correlational paradigm. Therefore, Mullen and Copper concluded that commitment to task was the primary component of cohesiveness for predicting performance.

Beauchesne et al. (1997) examined the cohesiveness of a 15-member women’s field hockey team using the GEQ, a leadership questionnaire, and six sociometric questions. Details of the GEQ analysis revealed that this team was much more cohesive in social aspects than in task-related areas. The social dimension subscales, ATG-S and GI-S were in the 75th and 80th percentiles respectively, while the task subscales ATG-T and GI-T were in the 55th and 30th percentiles (on the norms for females from Widmeyer et al., 1985). While these team members were moderately cohesive overall and showed a high concern for relationships, people and social issues, they were lower in task cohesiveness. This team did not do well in terms of wins and losses during the season of this research.

The primary purpose of Carron et al.’s (2002) study was to examine the relationship between task cohesiveness (GEQ subscales ATG-T and GI-T) and win/loss percentages in elite basketball and soccer teams. Carron et al. focused on the relationship between task cohesion and team success only, and not the relationship between social cohesion and team success, due to participant burden and conceptual rationale (see Mullen & Copper, 1994). Participants included 294 Canadian intercollegiate and club athletes from 18 basketball and 9 soccer teams. Researchers hypothesized that GI-T and ATG-T would both be positively related to team success in terms of wins and losses and the GI-T would show a stronger relationship to team success than ATG-T. The intraclass correlation coefficients and eta-squared statistics for both subscales were consistent with group effects (wins/losses), lending support for the hypothesis that task cohesion is related to a team’s win/loss record. Contrary to the hypothesis, the summation of effect sizes showed that, within the total sample, the relationship between GI-T and success was very strong (r=.57), but the relationship between ATG-T and success was even stronger (r=.67).  Therefore, the individual’s attraction to the group’s success on tasks was an even stronger predictor of team success than the individual’s perception of how well the team worked on task functions.Relating to how other variables effected, or were affected by, a team’s cohesion, Turman (2003) conducted a study on coaching techniques and strategies. Mathes (1997) examined team performance and Eys et al. (2003) explored pre-competition anxiety.

Turman (2003) examined how specific coaching techniques and strategies were utilized by coaches to foster and promote cohesion among their players through the use of interviews. Turman’s findings identified a number of techniques that diminished cohesion (i.e. inequity, embarrassment and ridicule) and a number of techniques that increased cohesion (i.e. bragging, sarcasm and teasing, motivational speeches, quality of opponent, athlete directed techniques, team prayer, and dedication).
Mathes (1997) studied how team cohesion was affected by winning and losing on interacting teams (groups playing together, e.g. lacrosse and basketball) and co-acting teams (individuals competing for a team score, e.g. swimming and gymnastics) using all four subscales of the GEQ. In contrast to earlier studies, Mathes’ study found that co-acting teams scored significantly higher ( p < .05) on the ATG-T and GI-T subscales after a loss, rather than a win.

Eys et al. (2003) examined the association between athlete perceptions of task cohesiveness (ATG-T and GI-T subscales) and the degree to which perceptions of the intensity of pre-competition anxiety symptoms were viewed as facilitating or debilitating. Findings showed that athletes who perceived their anxiety as facilitative had higher scores on one or both of the task subscales. While many other researchers have studied team cohesion with various populations, a review of these is not directly related to this paper.

Studies correlating GEQ scores with team performance provide general support for the theory that higher task-cohesion is correlated with winning. While researchers have studied the relationships of a number of variables (coaching techniques, winning and losing, and pre-competition anxiety) on team cohesion, to date, sexual orientation has not been examined. Recent studies related to homophobia and lesbian athletes were concerned primarily with an individual’s sexuality and not the effect it has on team cohesion or team success. This gap in the literature begs an important question: does the presence of lesbian athletes affect team cohesion in women’s sport? While previous studies have provided valuable knowledge and helped frame the discussion about the effects of lesbian athletes in women’s sports, the present study seeks to expand that knowledge through exploratory research into the difficult, and as yet unaddressed question, of how the sexual orientation of women’s collegiate basketball players affects team cohesion.

The research questions addressed by the present study were: (a) do athletes that self-identify as lesbians and athletes that do not self-identify as lesbians have different scores on team cohesion? and (b) does the perceived presence of non-self-identified lesbians on their team affect the athlete’s team cohesion scores?

The design of this study was a nonexperimental evaluation utilizing research questions intended to lead to insights, that in turn, can be examined in future studies. The variable of interest, sexual preference, is nonmanipulable necessitating this design (Shadish, Cook & Campbell, 2002). While nonexperimental design cannot support strong causal claims and is open to potential threats (e.g. history and maturation—team member’s previous experiences and developmental states may vary) (see Campbell & Stanley, 1963), the authors believed this exploratory research was warranted because the results of these studies may prove useful for educating stakeholders about women’s sport and sexuality based on more than myth and innuendo.

A modified version of the Group Environment Questionnaire (GEQ) (Widmeyer et al., 1985) was used to investigate the effects of sexual orientation on team cohesion among Division II women’s collegiate basketball players. A description of the instrumentation, sampling, data collection and analysis follows.

Widmeyer et al. (1985) developed the GEQ during three studies, which resulted in the 18-item questionnaire and four subscales (see Figure 1). Each subscale total score is considered a relatively unique value and is moderately related to the other three. Therefore, results of the GEQ are calculated as four individual subscales and not as one total GEQ score. Each of the 18 questions on the GEQ has a 9-point Likert-type scale, ranging from 1 to 9, with 1 being “strongly disagree” and 9 being “strongly agree”. Twelve of the questions are worded in the negative and five are worded in the positive. The eighteen questions are grouped into the four subscales, Widmeyer et al. (1985) reported reliability coefficients for ATG-T, ATG-S, GI-T and GI-S as being .75, .64, .70 and .76 respectively (Widemeyer et al.). They concluded the GEQ demonstrated evidence of reliability, validity, and good internal consistency. Questions worded in the negative were reverse coded to determine subscale scores. The total possible high scores are: ATG-T = 36; ATG-S = 45; GI-S = 36; and GI-T = 45. A normative table was developed for the GEQ with regard to female athletes for each of these four subscales (see Widmeyer et al.). Subscale comparisons can be made using these normative percentages thus revealing what weight various team members, or an entire team, give to each of the four aspects of cohesion.

Six demographic and gender-related questions were developed and added to the GEQ in order to address the research questions posed by the present study: (a) What year are you in school? (b) How many years have you been on this team? (c) Did you play high-school basketball? (d) Do you consider yourself to be a lesbian? (e) Are there any self-identified lesbians on your team, and if yes, how many? and (f) Are there any individuals on your team who you believe are lesbians, but do not identify themselves, and if yes, how many?

Recruiting teams to participate was anticipated to be difficult due to busy team schedules and the controversial topic of the study. A convenience sample limited to one region was necessary due to the decision to personally administer surveys rather than relying on mail or Internet survey techniques and the constraints of time, funding and assurance of anonymity. Twelve randomly selected NCAA Division I women’s basketball teams were initially contacted by letter, phone, and E-mail. None of these coaches agreed to participate, so letters were sent to head coaches at three randomly selected NCAA Division II women’s basketball programs. One coach responded positively and assisted in recruiting two other Division II teams.

Data Collection
The surveys were administered to three Division II women’s basketball teams during February 2003, mid-way through the team’s playing seasons. Timing of the survey allowed for ample time for first-year players to obtain enough experience with their team to respond to the questionnaire. Athletes were verbally assured that confidentiality would be maintained and neither individual nor team names would be included in the study. Completion of the questionnaire was optional, and all athletes on each of the three teams willingly completed the survey. Coaches were not present during the administration of the surveys.

Data were entered into SPSS®, and GEQ questions written in the negative were reverse coded in order to determine subscale scores. Descriptive statistics and the Mann Whitney statistic were calculated in order to address the research questions. Mann Whitney was chosen because of the small sample size. The data were carefully examined to verify that distributions of subscale scores for athletes that self-identified as lesbians and athletes that did not self-identify as lesbians met the necessary criteria of being similar in shape, and the reliability of team cohesion subscales was checked. Alpha was reported at p < .10 due to the exploratory nature of the study. For research question one, the four subscales that measure team cohesion were compared for athletes who were self-identified lesbians to those who were not self-identified lesbians. For research question two, GEQ subscale scores were compared between athletes who stated there were lesbians on the team who had not self-identified and athletes who stated there were not any non-self-identified lesbians on the team.

The sample consisted of 35 female athletes from three Division II women’s basketball teams in one region of the United States. Regarding year in school, 9 (25.7%) were freshman, 9 (25.7% ) were sophomores, 12 (34.3%) were juniors, and 5 (14.3%) were seniors. Twenty-eight (80.0%) athletes had been on the team between 0 and 2 years and 7 (20.0%) athletes had been on the team 3 to 4 years. Thirteen (37.1%) athletes were self-identified lesbians and 22 (62.9%) were not.  Thirty-four (97.1%) of the athletes indicated that there were self-identified lesbians on their team. When asked how many self-identified lesbian athletes there were on the team, replies ranged from one athlete stating 2, four athletes stating 3, fifteen athletes stating 4, fourteen athletes stating 5, with one athlete not responding. Fourteen (40.0%) of the athletes indicated there were individuals on their team who were lesbians but did not identify themselves as such, and 21 (60%) of the athletes indicated there were not any non-self-identified lesbians on their team.

GEQ Scores for Lesbian and Non-Lesbian Athletes
Reliability coefficients for the study sample on each of the four subscales ATG-T, ATG-S, GI-T and GI-S were .11, .74, .75 and .78 respectively. The authors speculated that the low reliability for ATG-T might have occurred because of the small sample size and the high discrimination power of the 9-point spread on the Likert scale. The mean GEQ score for the three reliable subscales were compared for self-identified lesbian athletes and non-lesbian athletes (see Table 1).
Table 1

While there were no significant differences for the GI-T subscale, the GEQ subscales for the social dimensions of team cohesion (ATG-S and GI-S) were both significantly higher for non-lesbian athletes than lesbian athletes (p < .05 and p < .10 respectively) (see Table 1). The ATG-S mean of 32.67 for non-lesbian athletes was higher than the mean of 25.00 for lesbian athletes (p=.021). The GI-S mean of 23.33 for non-lesbian athletes was higher than the mean of 17.69 for lesbian athletes (p=.068) (see Table 1).

Does the perceived presence of non-self-identified lesbians affect team cohesion?
Research question two was examined by comparing GEQ subscale scores for athletes who stated there were lesbians on their team who had not self-identified and athletes who stated there were not any non-self-identified lesbians on their team (see Table 2). One finding emerged with regard to the Group Integration-Social (GI-S)

Table 2
subscale, which measured the athletes’ perceptions of the socially oriented similarity, closeness and bonding of the team as a whole. Those athletes perceiving there were non-self-identified lesbian athletes present on their team scored lower with a GI-S subscale mean of 17.66, while athletes perceiving there were not any non-self-identified lesbian athletes present on their team scored higher with a mean of 23.33 (p = .068) (see Table 2).

Interpreting the results according to the meaning of each reliable subscale leads to the following observations: (a) both self-identified lesbian athletes and non-lesbian athletes had similar feelings about the team as a whole with regards to task oriented similarity, closeness and bonding within the team; (b) non-lesbian athletes felt more positively than self-identified lesbian athletes about their individual attraction to the appeal of the team as a social unit, including social interactions within the group; (c) non-lesbian athletes felt more positively than lesbian athletes about the team as a whole with regard to socially oriented similarity, closeness and bonding, (d) athletes who perceived there were non-self-identified lesbians on the team scored somewhat lower on their views of the team’s socially oriented similarity and closeness and bonding of the team as a whole. These results are interesting in view of findings from the literature.

Both self-identified lesbian and non-lesbian athletes in the present study provided their most positive responses to the group integration-task questions (see Figure 1 and Table 1). The meta-analysis by Mullen and Copper (1994) supported the concept that the task subscales, and not the social subscales, were the strongest predictors of cohesiveness-performance effect (i.e. if a team has strong task cohesion, they will win more games). Mullen and Copper also found that the social subscales did not predict, and were actually negatively correlated with, performance-effect. Carron et al. (2002) further reinforced the finding that higher task scores were the strongest predictors of team success. Therefore, if a team’s primary concern is winning games, task subscale scores are the most important measures of cohesion to consider. Notably, in the present study both non-lesbian and self-identified lesbian athletes were very similar in their ratings of how well the team pulled together on group-oriented tasks. No observations can be made about the individual attractions to the group-task subscale due to poor reliability. Therefore, these athletes perceived they were contributing equally to the team’s task performance without regard to sexual orientation.

Non-lesbian athletes were more positive than self-identified lesbian athletes regarding the appeal of the team for social interactions on both individual and group levels (see Figure 1 and Table 1). While high social cohesion has not been correlated with winning games (Carron et al., 2002), and may actually have some negative consequences in terms of winning (Hardy, Eyes & Carron, 2005), it should be noted that many collegiate sports teams have other goals. For programs where quality social experiences and personal development of athletes is considered important, athletes’ scores on social subscales may be cause for concern. Stakeholders may consider adding more direct, positive opportunities for social interaction between team members, such as team building activities like completing challenge courses, creating skits, or hosting celebrations. Additionally, facilitation of social interaction during competitions themselves may be less cumbersome for organizational personnel and may be better received by athletes. Theberge (2000) notes, that in women’s hockey, players report increased social interaction when competition is formatted as a tournament rather than home and home contests. Since self-identified lesbian and non-lesbian athletes differed on aspects of social cohesion, a more comprehensive study to further explore the nature of these differences could inform stakeholders about ways to improve social cohesion for all athletes.

Athletes perceiving the presence of non-self-identified lesbians on their team were lower in their ratings of the team as a socially similar, close, socially oriented unit (see Figure 1 and Table 2). These results may not be surprising because athletes might be expected to provide lower ratings for group members whom they perceived were less than honest (e.g. by not revealing their sexuality). Though honesty as a factor affecting cohesion has not been explored to any great extent in the sport setting, honesty is studied more thoroughly in the general context of unit or group cohesion. For example, Brown and Ayres (2004) suggest that policy and/or custom requiring a lack of honesty regarding sexual orientation in the military has impeded “the very honesty and intimacy that helps to forge strong bonds” (p. 187). The responses of the athletes surveyed in the present study suggest that honesty may be important to cohesion in the sport setting as well.

Several limitations in the present study must be acknowledged: (a) the sample size is small and limited to 35 athletes from three Division II women’s collegiate basketball teams, (b) the small sample size combined with the high discriminatory power of the 9-point Likert scale resulted in poor reliability for the ATG-T subscale so these results could not be reported, (c) the three teams were a convenience sample all from the same state, and (d) one of the three teams openly discussed players’ sexual orientation while completing the questionnaire. This creates a potential threat to internal validity, but the comparison of this team’s GEQ scores to the other two teams revealed no obvious differences. Therefore, these results, while not generalizable to other female collegiate athletes, should be considered as the first empirical evidence of their kind, which creates a foundation for future studies and discussions.

Conclusions and Recommendations
Three observations can be made based on the results of the present study. First, non-lesbian and self-identified lesbian athletes provided similar ratings with regard to task-related questions from the Group Environment Questionnaire. It should be noted that other researchers found that task-oriented teams are more likely to win, thus the similar commitment to task present in both self-identified lesbian and non-lesbian athletes suggests that they might contribute equally to the success of the team (Carron et al., 2002; Mullen & Copper, 1994).

Second, the differences that did exist between non-lesbian and self-identified lesbian athletes were all related to social aspects of team cohesion. Non-lesbian athletes scored significantly higher on questions related to the teams’ ability to satisfy members’ social and emotional needs. To date, researchers have not found a strong connection between social cohesion and winning (Carron et al., 2002; Mullen & Copper, 1994). Therefore, lower social cohesion scores for lesbian athletes may not be cause for concern except when stakeholders perceive there is a desired social benefit to be derived from sport participation. If this finding is validated through additional research, it could be recommended that teams with goals related to socialization and character development of athletes may want to conduct more education, teambuilding and trust activities. If it can be established through additional research that non-lesbian athletes have no problem playing with self-identified lesbian athletes, more openness can be encouraged.

Finally, while a number of athletes perceived the presence of non-self-identified lesbians on their teams, such presence did not significantly affect the athletes’ cohesion scores for the group integration-task subscale. However, some differences were evident in the Group Integration-Social (GI-S) subscale indicating that those athletes who felt there were non-self-identified lesbians on their team were less likely to feel closeness and bonding with the team as a whole with regard to social activities.

In closing, we would like to re-iterate Krane’s (1997) observation that prejudice and discrimination affects homosexuals and heterosexuals alike in that all women in sport are potentially harmed by myths associated with lesbian athletes. While the present study is an initial, descriptive step toward filling a gap between scholarly discourse and perceptions regarding sexual orientation in women’s collegiate sports, much work remains to be done. Future studies are needed replicate and validate these findings with greater sample sizes that lend themselves to generalizability. The authors propose that in addition to team cohesion, other variables of interest could include win/loss records, athletes’ personality types, effect of coaching behaviors, and impact of extraordinary events such as the firing of a coach or loss of a team member. Future studies should continue to explore athletes’ perceptions and attitudes with focused discussion on issues relevant to the goals of collegiate athletics and stakeholder education.


Adams, N., Schmitke, A., & Franklin, A. (2005). Tomboys, dykes, and girly girls: Interrogating the subjectivities of adolescent female athletes. Women Studies Quarterly, 33(1-2), 17-34.

Anderson, E. (2005). In the game: Gay athletes and the cult of masculinity. Albany, NY: State University of New York Press.

Barkawi, T., Dandeker, C., Wells-Petry, M., & Kier, E. (1999). Rights and fights: Sexual orientation and military effectiveness. International Security, 24(1), 181-210.

Beauchesne, S., Turner, E., Brennan, R., & Hoopengardner, S. (1997, December). An analysis of the correlation between leadership and cohesion in a women’s field hockey team. Retrieved November 13, 2002 from

Blinde, E. M., & Taub, D. E. (1996). Women athletes as falsely accused deviants: Managing the lesbian stigma.  Sociological Quarterly, 33, 521-535.

Brown, J., & Ayres, I. (2004). The inclusive command: Voluntary integration of sexual minorities into the military. Michigan Law Review, 103, 150-188.

Cahn, S. (1994). Coming on strong: Gender and sexuality in twentieth century women’s sport. Cambridge, MA: Harvard University Press.

Campbell, D. T., & Stanley, J. C. (1963). Experimental and quasi-experimental designs for research. Chicago, IL: Rand McNally College Publishing Company.

Carron, A., Bray, S., & Eys, M. (2002). Team cohesion and team success in sport. Journal of Sports Sciences, 20, 119-126.

Carron, A. V. (1982). Cohesiveness in sport groups: Interpretations and considerations. Journal of Sport Psychology, 4, 123-138.

Cotton, S., & Jackson, S. (1992). Addressing the issue of homophobia in sport. The New Zealand Coach, 1(4), 2-3.

Dowling, C. (2001). The frailty myth: Women approaching physical equality. New York: Random House.

Eys, M. A., Hardy, J., Carron, A. V., & Beauchamp, M. R. (2003). The relationship between task cohesion and competitive state anxiety. Journal of Sport & Exercise Psychology, 25(1), p. 66-77.

Fasting, K. (1998). Sexual stereotypes in sports: Experiences of female soccer players.Society’s Watchdog - or Showbiz’ Pet? Retrieved August 30, 2001,

Greendorfer, S., & Rubinson, L. (1997). Homophobia and heterosexism in women’s sport and physical education. Women in Sport and Physical Activity Journal, 6(2), 189-210.

Griffin, P. S. (1998). Strong women, deep closets: Lesbians and homophobia in sports. Human Kinetics Press.

Hardy, J., Eyes, M. A., Carron, A. V. (2005). Exploring the potential disadvantages of high cohesion in sports teams. Small Group Research, 36(2), 166-187.

Hargreaves, J. (2000). Heroines of sport: the politics of difference and identity. NY: Routledge.

Harrison, L. (2001). Understanding the influence of stereotypes: Implications for the African American in sport in physical activity. Quest, 53, 97-114

Knight, J., & Giuliano, T. (2003, September) Blood, sweat, and jeers: The impact of the media's heterosexist portrayals on perceptions of male and female athletes.
          Journal of Sport Behavior, 26(3), 272-285.

Krane, V. (1996). Lesbians in sport: Toward acknowledgment, understanding, and theory. Journal of Sport and Exercise Physiology, 18(3), 237-246.

Krane, V. (1997). Homonegativism experienced by lesbian collegiate athletes. Women in Sport and Physical Activity Journal,6(2), 141-163.

Kreigh, L., & Kane, M. (1997). A novel idea: Portrayals of lesbians in young adult sports fiction. Women in Sport and Physical Activity Journal, 6(2), 23-62.

Lenskyj, H. (2003). Out on the field: Gender, sport and sexualities. Toronto: Women’s Press.

Mathes, H. (1997). The effect of winning and losing on female interactive and coactive team cohesion. Journal of Sport Behavior, 20(3), 284-299.

McKay, J., & Martin, R. (2001). Courting Lesbianism. In Miller, T. (Ed.), Sportsex (pp. 103-125). Philadelphia, PA: Temple University Press.

Messner, M. (2002). Taking the field: Women, men, and sports. Minneapolis, MN: University of Minnesota Press.

Mullen, B., & Copper, C. (1994). The relation between group cohesiveness and performance: An integration. Psychological Bulletin, 115, 210-227.

Nelson, M. (1994). The stronger women get, the more men love football: Sexism and the American culture of sports. New York, NY: Harcourt Brace.

Schwartz, H. L. (1997, March 18). Out of bounds. Advocate, 56-59.

Shadish, W. R., Cook, T. D., & Campbell, D. T. (2002). Experimental and quasi-experimental designs for generalized causal inference. New York: Houghton
          Mifflin Company.

Sugarman, K. (1998, November 9). Cohesion. Retrieved October 20, 2002 from
 team cohesion.html

Theberge, N. (2000). Higher goals: Women’s ice hockey and the politics of gender. New York: SUNY Press.

Thiese, K. (1999). Together we stand, divided we fall: Using goal setting to increase
          Retrieved August 1, 2002 from

Turman, P. (2003). Coaches and cohesion: The impact of coaching techniques on team cohesion in the small group sport setting. Journal of Sport Behavior, 26(1),

Vealey, R. S. (1997). Transforming the silence on lesbians in sport: Suggested directions for theory and research in sport psychology. Women in sport and physical
          activity journal, 6(2), 165-188.

Wellman, S., & Blinde, E. (1997). Homophobia in women’s intercollegiate basketball. Women in Sport and Physical Activity Journal, 6(2), 63-82.

Widmeyer, W., Brawley, L., & Carron, A. (1985). The measurement of cohesion in sport teams: The group environment questionnaire. Ontario: Sports Dynamics.

Yukelson, D., Weinberg, R., & Jackson, A. (1984). A multidimensional group cohesion instrument for intercollegiate basketball teams. Journal of Sport
          Psychology,  6, 103-111.