The influence of gender-role socialization, media use and sports participation on perceptions of gender-appropriate sports.

Hardin, Marie ; Greer, Jennifer D.

Although the experiences of millions of girls and women in the United States indicate the contrary, research demonstrates that media consistently, and across all platforms, present sports as the purview of men (Duncan & Sayaovong, 1990; Hardin, Lynn, Walsdorf, & Hardin, 2002; Pedersen, 2002). Numerous studies have demonstrated that female athletes have been vastly underrepresented in media coverage (Bernstein, 2002; Pedersen, 2002). The reason for this could be that the most popular spectator sports in the United States are those considered masculine (Messner, 2002). Yet, since the passage of Title IX in 1972, the movement of girls and women into many sports that are not considered "feminine" has been phenomenal. Girls and women participate in virtually every type of sport, even those used to showcase the ultimate in hegemonic masculinity; several professional women's football leagues have operated in the United States since 2000 (About NWFA, n.d.; Associated Press, 2007).

Inclusion of more women than ever reflects changing values about their athletic aptitude. It seems logical to ask: Have perceptions of sports progressed in ways that mirror participation? In addition, participation in and media coverage of action sports has grown dramatically since earlier studies have examined attitudes toward gender-appropriate (1) sports. Men and women alike are competing in sports such as snowboarding, wakeboarding, and skateboarding--and sports broadcasters are there to capture the action.

Because of these trends, this study updates research on how U.S. sports are viewed in light of gender norms. Attitudes toward the masculinity of 14 sports were collected through a mass Internet survey of college-age men and women. Further, the study examined how these attitudes were related to sports participation, media consumption, and gender socialization.

Although gender-role differences as biological and "natural" exist in popular consciousness, research has long demonstrated that, instead, many are socially constructed (Bandura, 1986; Messner, 2002). Gender stereotyping is a ubiquitous, invisible regulator of relationships and opportunities. Hargreaves (1994) argues that individuals understand their gender because they are given names and treated in particular ways, such as dressed in pink for girls or blue for boys, that reflect social constructions of gender. Bandura's social cognitive theory is a key in understanding the factors in socialization. He argues that behavior, environmental events, and cognitive factors operate to shape attitudes and action. Individuals ponder action and its outcome, projecting consequences and adjusting accordingly. Thus, action is not a result of "imprinted histories" as much as it is a result of "cognized futures" (Bandura, 1986, p. 19). Bandura emphasizes the role of media in social learning, so much that, he argues, television influence has "dethroned" the primacy of interpersonal experience. As a consequence, life models the media (Bandura, 1986, p. 20).

Bandura leaves open the possibility for evolution in how activities are typed in terms of gender. Multiple models of men or women exhibiting consistent activity is the basis of the gender typing process; over time, "concordance gender-linked modeling can confer masculinity or femininity to previously neutral activities" (1986, p. 95). Thus, previously gender-typed activities, if modeled often enough by men and women, could eventually confer neutrality on them.

Perceptions of Sports as Gender-appropriate

As children are introduced to sports, their experiences are based on gender roles and expectations (Hargreaves, 1994; Nilges, 1998). The construction of sports as appropriate replicates gender-typed toys: rough-and-tumble symbols for boys, domestically oriented symbols for girls. Messner (2002) writes that day-to-day interactions of children with each other and with adults still privilege boys and men in the athletic status system and marginalize girls and women.

Early work on how sports are typed in regard to gender was done by Metheny (1965), who proposed a set of attributes used to categorize a sport as feminine or masculine; sports recognized as masculine involve contact and the use of force or heavy objects (Koivula, 2001). Later, Postow (1980) argued that sports-related attitudes such as devotion to a team, stamina, and competitive spirit also are perceived as masculine. Thus, team sports are considered more masculine than individual sports. Sports in which aggressiveness is considered an essential part of the game, including ice hockey and football, have been regarded as masculine (Koivula, 2001). Sports that have historically been perceived as feminine, such as figure skating or gymnastics, are those that allow women to exhibit gender-role attributes such as grace and beauty while participating (Koivula, 2001). These typologies reinforce ideas of difference; they showcase constructions of men as stronger and faster, thus deserving a higher rank in the overall social order, than women. Generally, men and women type sports similarly; exceptions sometimes occur with basketball, which may be categorized as a more masculine sport by boys than by girls (Riemer & Feltz, 1995).

Although Cashinore (2005) argues that the typologies developed by Metheny (1965) and others are "about as fresh as disco music and mullets" (p. 157), research indicates that even in recent years, sports have been gender-typed in traditional ways (Koivula, 2001; Matteo, 1986; Riemer & Feltz, 1995; Solmon, Lee, Belcher, Harrison, & Wells, 2003). More recent studies, however, have identified that some sports are perceived as more neutral--indicating a slight shift in perception that sports must be either masculine or feminine. A recent study (Koivula, 2001) involving 400 university students found that participants categorized sports as feminine, masculine, or gender-neutral based on their perceptions of the sports' aesthetics, speed, and risk. Sports such as tennis, volleyball, and swimming were ranked as neutral, gymnastics and aerobics were ranked as feminine, and baseball, soccer, and football were typed as masculine. Respondents incorporated the perceived purpose of a sport and its risk when assigning labels. Koivula (2001) points out that definitions of a gender-appropriate sport can change because gender is constructed based on historically and culturally specific conditions. Action sports, which have attracted more participants and more attention from media in recent years, have not been examined in past studies related to gender-typing.

The Influence of Sports Participation

Since passage of Title IX, sports participation by girls and women has grown exponentially. In 1972, 1 in 27 girls played high school sports; in 1998, one in three did (Sports Illustrated for Women, 2002). Sports participation by boys also has increased, although not at the same rate (Carpenter & Acosta, 2005). Most growth in participation by girls and women has been in sports that have been typed neutral or masculine, such as soccer. The most frequent college varsity sports for women are basketball, volleyball, cross country, soccer, softball, tennis, track and field, golf, swimming, and lacrosse--none of which is aesthetically oriented (Acosta & Carpenter, 2004).

The expanding role of sports in the lives of girls (and boys) in the United States could lead to more progressive ideas about what constitutes a gender-appropriate sport, but research has not supported that possibility. Several studies have revealed that male athletes have more conservative, traditional attitudes toward gender roles than do male non-athletes (Andre & Holland, 1995; Boyle, 1997; Houseworth, Peplow, & Thirer, 1989). Studies in the 1980s and 1990s demonstrated that high school and college students judged participation in gender-appropriate sports as socially more desirable than participation in sports deemed gender-inappropriate; for instance, girls who participated in gymnastics were deemed more desirable as a date (for boys) and as a friend (for girls) than were girls who played golf or softball (Holland & Thomas, 1994).

Matteo (986) found that the more strongly a male college student adhered to traditional gender roles, the less likely he was to participate in sports not considered masculine. Young women, even if they strongly adhered to gender roles, were more likely to try masculine sports, perhaps because masculine sports are considered more valuable in U.S. culture (Matteo, 1986). Perceptions of a sport as masculine, feminine, or neutral also may impact perceptions of ability. Solmon et al. (2003) found that college-aged women who perceive a sport as gender-neutral are more confident about participating than are women who identify a sport as masculine.

Impact of Media Messages

Research indicates that the U.S. sports/media complex has positioned sports as male terrain; its "masculinist cultural center" has been a site for boys and men to learn hegemonic masculinity (Messner, 2002, p. 92). Messner has outlined lessons of the "televised sports manhood formula:" sports belong to men; aggression is integral to sports and to masculinity; and violence is natural and oftentimes necessary. Lessons from the televised sports manhood formula "are evident, in varying degrees, in the football, basketball, extreme sports, and SportsCenter programs and their accompanying commercials" (p. 124).

Media emphasize the "sports manhood formula" and overwhelmingly feature core men's sports (Bernstein, 2002; Bishop, 2003; Messner, 2002; Pedersen, 2002). Sports media generally dedicate only 5% to 8% of coverage to women's sports even though 40% of sports participation is by women (Adams & Tuggle, 2004; Kane, Griffin, & Messner, 2002). Further, network coverage emphasizes women's sports considered traditionally gender-appropriate. For instance, NBC's Olympic coverage showcases women's figure skating (winter) or gymnastics (summer) while Olympic sports such as women's shot put or discus are virtually invisible, and women's team sports receive less prime-time coverage than individual sports (Tuggle, Huffman, & Rosengard, 2002). Adams & Tuggle (2004) found that women's team sports such as basketball, soccer, and softball received less coverage in more recent years than in the early-to-mid 1990s.

U.S. sports media outlets enjoy great popularity. In the late 1990s, 94% of children surveyed said they consumed sports media, and many said they did so daily (Messner, 2002). "Sports media are thus likely to be one of the major influences on children's views of gender, race, commercialism, and other key issues" (2002, p. xix). Messner argues that children are socialized into traditional views of gender and sport even by the new genre of "action" sports (also called alternative or extreme sports) such as skateboarding and snowboarding. Such sports are so popular that teenage sports fans in 2002 voted skateboarder Tony Hawk "coolest big-time athlete" (Wheaton, 2004). Action sports have moved into the mainstream through heavily commercialized coverage of the "X Games" on television and the integration of snowboarding into the Winter Olympics in 2006. These sports are mostly individual activities that emphasize both risk (masculine) and aesthetics (feminine); they are also non-contact. Wheaton argues that action sports offer possibilities for more progressive ideas about gender.

Research Questions

A number of studies since the 1960s that have examined sex-typing of sports; however, none found in this exhaustive review sought to reassess the notion of gender-appropriateness of sports in light of increased female participation in recent years. Further, although the relationship between participation and sex-typing of sports has been explored, recent research has not accounted for media consumption, a key factor in social learning about gender roles. The goals of this study were two-fold: 1. To assess the gender-typing of sports in the wake of increased female participation in non-aesthetic, team sports such as soccer and basketball and in regard to newer action sports that are, because of their features (risk + aesthetic qualities), harder to classify; and 2. To assess the impact of lived experience (i.e., sports participation), gender-role socialization, and media consumption on gender-typing of sports. The following research questions were posed:

Where do certain sports, including newer action sports, fall on the masculine/feminine continuum, as rated by young adults reared in an environment with increased female presence in team sports? Are clear distinctions between masculine, feminine, and neutral sports still made in the current sports environment?

Do men and women differ in their gender-typing of these sports?

What role do media consumption (television viewing, sports viewing, and attention to coverage of specific types of sports), sports participation (participation in organized sports, personal fitness, and playing specific types of sports), gender role socialization (as measured by attitudes toward masculinity) and demographics (gender and ethnicity) play in the gender-typing of sports?

Does gender interact with sports participation, media consumption, gender role socialization, and ethnicity attitudes toward the masculinity of certain types of sports are examined?

Method

Overview

To examine the relationship between demographics, attitudes, and behavior and the sex-typing of sports, researchers developed a Web survey and received approval from the human subjects committee at a large, Research I university in the Northeast. (2) The university's football program is extremely popular among students; its men's and women's basketball programs also draw many spectators although neither team was nationally ranked last year. Because of cool weather, baseball, softball, and other outdoor spring sports are not large draws. Although the university offers volleyball as an intercollegiate sport for only women, it also offers volleyball as a club sport for men, and the men's matches are also popular among spectators.

A convenience sample of 370 students enrolled in a Web-based introductory grammar class at that university was selected. Students were given the option of completing this Web survey or an online grammar assignment for extra credit. While this sample is problematic for any descriptive data, only the fast research question includes a descriptive aspect. The rest of the study seeks to explain the relationship among variables rather than to describe attitudes, mitigating some of problems associated with convenience samples.

In 12 days of data collection in April 2006, 365 responses were collected online. After duplicate submissions and partially completed surveys were excluded, the number of usable responses dropped to 340, a response rate of 91.9%. Because the non-response rate was so low (8.1%), researchers do not believe there was any systematic bias created by the non-participation. However, it must be noted that this sample cannot be considered representative of the university population from which it was drawn or of the broader population of U.S. college students.

Sample Characteristics

Of those 340 who returned usable surveys, 108 (31.8%) were men, 230 (67.6%) were women, and 2 did not answer the question. The average age was 19.07 (SD = 1.32). Ages ranged from 18 to 34, with 97.9% of respondents 21 or younger. The vast majority (83.8%, 285) identified themselves as Caucasian, 18 (5.3%) as African-American, 14 (4.1%) Asian, 9 (2.6%) Hispanic/Latino, 9 (2.6%) "other," 2 (0.6%) Pacific Islander, 1 (0.3%) Native American, and 2 (0.6%) did not answer the question. (3)

The Questionnaire

To assess the gender-typing of 14 sports of interest in this study, respondents were asked to rate each sport from 1 (feminine) to 5 (masculine) with 3 marked neutral. The following prompt was given: "Please rate the following sports on a gender-appropriateness scale. Which sports do you believe are more feminine or masculine?" The 14 sports chosen (see Table 1) did not include some popular sports such as baseball, auto racing or figure skating. The list was compiled to get a range of sports that included several that were theoretically very masculine (football and rugby) and very feminine (gymnastics) based on research by Metheny (1965), Koivula (2001) and others. Other sports chosen were added for one of two reasons: a) they were new-genre action sports that had not been sex-typed in previous studies; or b) they were sports with potential to be coded neutral because of healthy participation rates among men and women in recent years whether for personal fitness or for competition, especially among those of college age. As Koivula (2001) and Bandura (1986) both suggest, repetitive participation (or images of) by men or women in a particular sport could impact perceptions of its appropriateness.

To assess their experience with sports, subjects answered questions about their time spent on organized sports and personal fitness activities each week (both measured on a 1 to 4 scale with 1 equaling 0 hours and 4 equaling 10 or more hours). Then they indicated whether they had participated recreationally or competitively or had not participated in each the 14 sports.

Gender role socialization was assessed on Brannon's Masculinity Scale (BMS, Brannon & Juni, 1984). Brannon's original scale contained 58 items; 12 items were selected for this study. These items were scored on a 1 = strongly disagree to 4 = strongly agree response format with no neutral midpoint.

Media use was measured by asking subjects how many hours they watched television each week and how many hours of sports media (including Internet, magazines, newspapers, etc. as well as television) they consumed each week. Each of these measures was collected with a 1 to 5 response format, with I equaling none and 5 equaling 20 or more hours a week. In addition, respondents were asked to rate from 1 to 5 the amount of attention they paid to media coverage of the 14 sports of interest in the study.

Results

The first research question asked where the 14 sports of interest in this study fell on the masculine/feminine continuum with the goal of examining whether, at least for this convenience sample, respondents still made firm distinctions between masculine, feminine, and neutral sports in the wake of changes in team sports, women's participation, and action sports in recent years. First, the sports attitudes were investigated through a principal component factor analysis, using a Promax oblique rotation with Kaiser normalization. This was done to examine whether groupings emerged and whether these were consistent with previous studies. Four distinct factors emerged, as noted in Table 1: Hyper masculine sports (Football, Weightlifting, Rugby, and Basketball); Action sports (Motocross, Skateboarding, Snowboarding, Wakeboarding, and Surfing); Neutral sports (Soccer, Swimming, and Tennis); and Feminine sports (Volleyball, Gymnastics). (4) This analysis indicates that as a whole, respondents do group sports into clear categories. Some of the categories that emerged were consistent with previous studies; action sports, however, emerged as a separate category.

To investigate how these groupings relate to the perceived masculinity of a sport, the factors were examined in terms of overall means on the attitude scores. The first data column of Table 2 lists the average scores by sport from most masculine (5) to most feminine (1). Football to surfing fell on the masculine side of the scale (above 3); Soccer, swimming, and tennis hover near the neutral midpoint (around 3); and volleyball and gymnastics are viewed as feminine sports (below 3). Looking more closely at the overall means in Table 2, clear breaks occur between the groups suggested by the factor analyses. Together with the Principal Components Factor Analysis, these means suggest gender-typing of sports into masculine, neutral, and feminine areas. The means of action sports indicated that these were seen as in between masculine and neutral sports identified in this study and in previous studies.

The second research question examined whether men and women differed in typing these 14 sports. Based on the factor analysis from RQ1, averages were created for each category of sport and a Multiple Analysis of Variance (MANOVA) was run on these four dependent measures (Hyper masculine, Action, Neutral, and Feminine) by sex of respondent. Overall, attitudes were related to whether the respondent was male or female (F = 11.88, p < .001). The effect of gender was significant for the Hyper masculine sports, as Table 2 shows. To investigate the specific sports that male and female respondents evaluated differently, means by gender were compared on the 14 individual sports investigated in the first research question through a second MANOVA. As Table 2 shows, men and women did not vary in their assessment the masculinity of most sports. However, significant gender differences did emerge for three sports. Women saw weightlifting and basketball as significantly less masculine than did the men. In contrast, women rated swimming as significantly less feminine than did men. It is interesting to note that each time men and women differed significantly, women were more likely to place the sport closer to the neutral mid-point on the scale.

The third research question asked when all factors were considered (sports participation, media consumption, gender socialization, and demographics), which factors were most related to the typing of sports. To answer this research question, all the variables of interest described above were examined through a global MANOVA (Multiple Analysis of Variance) that examined main effects on attitudes toward the four categories of sports that emerged in the analysis for research question one (Hyper masculine, Action, Neutral, and Feminine). This test was run because it can be used to consider the main effects of multiple categorical and continuous independent variables on several continuous dependent variables, while controlling for Type I error inflation. For the 15 independent variables entered into the analysis multi-colinearity diagnostics were performed. None were correlated above .48, so the diagnostics didn't suggest a problem; therefore, all were retained. Significant models were found for attitudes toward Hyper masculine and Neutral sports, as Table 3 shows. A model atp < .1) was found for Action sports, which also is included in Table 3. As the model for Feminine sports was not significant, it is excluded from the table.

Distinct patterns emerged on the sources of variation related to different types of sports. For Hyper masculine sports, main effects emerged for participation in personal fitness, attention paid to Hyper masculine sports media, and ethnicity. Those who spent the greatest time on personal fitness weekly had the lowest masculinity ratings for the Hyper masculine sports. In contrast, those who paid greater attention to Hyper masculine sports media rated the sports as the most masculine. So doing, in the form of personal fitness, was related to attitudes viewing these sports as less masculine, while media consumption of these types of sports was linked to heightened views of the masculinity. Respondents who identified themselves as belonging to an ethnic minority group also saw the sports as significantly less masculine than did whites. Examining ethnicity by group revealed a significant difference between Caucasians, who gave Hyper masculine sports the highest masculinity rating by ethnic group (4.11), and all other ethnic groups, which had mean scores for Hyper masculine sports ranging from 3.75 to 3.94.

A weak model emerged for attitudes toward Action sports. Because study of attitudes toward these sports is largely exploratory, that model is detailed here. A main effect again was found for participation in personal fitness, and a weak effect emerged for the BMS score (attitudes toward masculinity). As with attitudes toward Hyper masculine sports, those who spent the most time in personal fitness activities rated these Action sports as the least masculine. The same trend was found for BMS average--those who held the most traditional views toward masculinity saw Action sports as less masculine than those with more egalitarian views toward gender. However, caution should be used in drawing firm conclusions, as the effects of BMS average were weak.

The most robust model emerged for attitudes of the masculinity of Neutral sports ([R.sup.2] = .223). Interestingly, the sources of variation in this model were quite different than found in the models for Hyper masculine and Action sports. Here, participating in and following coverage of feminine and neutral sports were key. For each, the rating of these sports' masculinity climbed with more action (participation in neutral and feminine sports) and attention to sports media (coverage of neutral and feminine sports). In addition, those with more traditional views toward masculinity saw Neutral sports as being the most masculine.

The final research question examined whether these variables interacted with gender of respondent--in short, whether men and women varied in their assessment of the masculinity of sports depending on their sports participation, their sports media consumption, and other characteristics. The global MANOVA run for the previous question also examined the interaction of gender of respondent and all other independent variables to see how men's and women's attitudes differed in light of these other variables. As Table 4 shows, five significant interactions with gender of respondent emerged: two participation variables, two sports media variables, and attitudes toward masculinity (BMS). Interestingly, no significant interactions emerged in the model for attitudes toward Hyper masculine sports, indicating that men and women hold similar attitudes toward these sports, even when examined through participation, media consumption, and other variables.

Where gender of respondent did make a difference was for attitudes toward Action and Neutral sports, as Table 4 shows. For participation in personal fitness, women held fairly similar attitudes across the board, with those women who spent the most time in personal fitness seeing Action and Neutral sports as slightly more masculine than women who worked out less. In contrast, men who spent the most time on personal fitness saw both of these sports groups as significantly less masculine than men who spent fewer hours on personal fitness and than women. The same pattern was found for the other variables in the model for attitudes toward neutral sports. While women held roughly similar attitudes toward the masculinity of Neutral sports regardless of their participation in feminine sports, attention to Neutral sports, or attention to Feminine sports, men viewed the sports as significantly more masculine as their participation in Feminine sports and attention to Feminine and Neutral sports coverage increased. In short, men saw Neutral sports as more masculine when they participated in Feminine sports or consumed sports media about Feminine and Neutral sports. Finally, for attitudes toward masculinity (BMS), women again were fairly consistent in their ratings regardless of masculinity scores; for men, the ratings of the masculinity of Action and Neutral sports rose with their BMS scores. However, in both cases these were relatively small but significant differences.

Discussion

The results of this study must be viewed in light of its limitations. The study involved college students, who are not representative of the larger U.S. population and who may have different media consumption habits. Further, the study involved college students in a particular region of the country, where some sports are both less available for participation and less popular for spectatorship than others. Another limitation of this study was in the sports it used; sports such as weightlifting, which can be interpreted in different ways (for instance, as a "power" sport or as a fitness activity designed to tone the body) must be taken into account in interpreting the results. It also could be argued that other sports, such as lacrosse, could have been added because of their growing participation rates in the United States. Further, it must be acknowledged that this research may have forced respondents to think of sports in ways that were not natural to them; in other words, would respondents have chosen, without prompting, to describe any of the sports in this research as masculine or feminine? Thus, the rating scale itself is limiting.

Even so, the results of this survey clearly indicate that, when asked, youths who have grown up with Title IX and who have had real-life experience with a wider array of sports than any previous American generation are almost as likely to use traditional gender roles to type sports as previous generations. Metheny's typology of gendered sports, formulated in 1965, seems as valid today as it was 40 years ago, although attitudes seem to have slightly shifted into acceptance of some sports as neither entirely masculine or feminine.

Overall, men and women see sports similarly; they make clear distinctions among sports they rate as masculine and feminine. Clear differences emerged based on means, and certain sports clearly grouped together in factor analysis. For the most part, these findings were consistent with past studies that found respondents typing certain sports as masculine, feminine, and neutral. Action sports emerged as a separate category in factor analysis (discussed below), although the means for these sports were on the masculine end of the response format used to measure gender typing. Based on the means, respondents categorized far more sports as masculine on the 1-to-5 scale than neutral, demonstrating an orientation toward sports in general as a masculine endeavor. Predictably (according to traditional gender-typing), sports that emphasized overt displays of aggression or strength were typed as masculine, and non-contact sports that are either traditionally dominated by women (volleyball) or emphasize aesthetics (gymnastics) were typed as feminine. It seems that little has changed, even in the light of more liberal attitudes about women's sports participation in general. Further, Postow's (1980) assertion about the masculinity of team sports also still has relevance, despite the fact that team sports are a part of the sporting lives of millions of girls and women. The only two team sports that did not skew masculine were soccer and volleyball. Competitive volleyball is played by far more girls than boys in the United States; further, it is a non-contact sport. Soccer, although it may be considered a contact sport (in the same vein as basketball), may be typed as feminine for a couple of reasons: It has been positioned in U.S. society as an acceptable sport for girls, especially since the U.S. women's World Cup championship in 1999, which made Mia Harem and Brandi Chastain household names; and, unlike in other nations, U.S. soccer is not a popular spectator sport. Scholars speculate on the reasons for this, including the fact that it is not seen as "American" as are other sports such as football (Markovits & Hellerman, 2001).

As in previous studies, men and women did view the masculinity of the sports studied here similarly, although there were some differences. For instance, women saw basketball as less masculine than did men. This could be because basketball is one of the fastest-growing team sports for women and is one of the only team sports for women in the United States that gets even minimal media attention. Research suggests that if women perceive themselves as being able to participate in such a sport, they would tend to see it as neutral (as opposed to masculine). Other factors may have been at work in the assessment by men of swimming as more feminine than by women, such as the fact that swimming does not involve contact and is not a team sport and that it does not receive the same levels of media coverage as other sports rated as more masculine, for instance.

Action Sports

Sports commonly seen on ESPN's annual "X Games" and perceived as action sports were included on this survey because of their recent emergence as popular spectator sports and because of their combination of masculine (risk) and feminine (aesthetic) elements. All action sports in the survey loaded together in factor analysis, and all were rated as masculine on the measure of that construct, to varying degrees, by both men and women. It is important to note, however, that they clearly were not correlated with the masculine, neutral, or feminine sports; they emerged as their own group. This suggests that four decades after Metheny (1965) found evidence of gender-typing of sports as masculine and feminine, attitudes toward sports and gender roles may now be more nuanced. In addition to a clear neutral category of sports in people's minds, young adults see action sports as a different ballgame altogether.

These results provide reason to speculate that perhaps, as some scholars have argued, action sports are expanding definitions of masculinity because these sports are not contact or team sports and are judged on the look of the participant, much as figure skating or gymnastics. However, it is difficult to know how much less these sports would have been rated masculine if it were not for the risk factor, which is heavily touted in ESPN coverage through features such as on-screen graphics that report the number of broken bones and accidents that each competitor has had. Also, ESPN coverage generally excludes female competitors, which is also a factor (as discussed below). Unfortunately, results indicating that viewers see these sports as masculine provide encouragement for ESPN to continue excluding women, creating a circular relationship that reinforces these sports as masculine.

Intervening Factors in Gender-Typing

Participation. Previous studies have addressed the interaction of participation and attitudes about gender and sports; this research found interesting relationships not covered in previous research. For instance, participation in personal fitness activities seems to neutralize, to a certain degree, perceptions of Hyper masculine and action sports. One reason for this could be that these sports are highly performance/spectator oriented; those who pursue fitness and participate in sports-related activity may feel more connected to these sports, more likely seeing them as participation opportunities than as purely all-male, power/risk spectacles.

On the other hand, sports that were generally typed as neutral and feminine were typed as more masculine by individuals who also reported fitness and sports participation. We speculate that a reason for this could be that sports participation informs individuals about the necessity for the strength, power and stamina required for a broad range of sports, even those that may not overtly display these attributes; for instance, consider the power behind a hit across the net in a volleyball game or the force behind moving a ball downfield during a soccer match. Because sporting attributes such as strength are considered masculine, those who have participated are more likely to type even sports that don't "look" masculine as, indeed, more masculine.

Media consumption. Following sports through media coverage also seems to impact the sex typing of sports, and in intriguing ways. It seems here that Messner's (2002) "manhood formula" is relevant; he postulates that sports media coverage reinforces traditional masculinity through an emphasis on toughness and dominance in depictions. Thus, masculinity and sports are conflated. It is not surprising, then, that those who consume mediated sports would rate them as more masculine.

The findings here, however, suggest that adherence to traditional masculinity (as measured by the abbreviated BMS scale) makes little significant difference in the gender-typing of sports. We wonder, then, if it is not so much the emphasis on traditional masculine values in sports depictions, but simply the ubiquitous images of men, that ultimately conflate sports with men/masculinity. An example of this possibility is action sports; the sports themselves, with the exception of motocross, are centered on aesthetics. In other words, it is the way the athlete looks (grace, coordination: feminine) that determines excellence. Yet, these sports were still typed as masculine, likely because the majority of images are of men. Certainly, ESPN's framing of these sports as high-risk endeavors cannot be overlooked; even so, they are about performance, they are individual sports, and they do not involve demonstrations of force, which means they should, by Metheny's (1965) and Postow's (1980) arguments, be typed as neutral-to-feminine.

Social Learning and the Stagnation of Gender-Typing

Thus, it seems that traditional views of masculinity become less important than the overwhelming proportion of mediated sports images that depict men jumping, running, hitting, throwing, kicking, lifting, riding, and skating. Although factors such as participation and media use interact with gender to problematize this conclusion under some circumstances for some sports, sports continue to be conceptualized as a generally masculine endeavor with the exception of a few activities. Even individuals who do not follow sports coverage cannot escape the presentation of sports on television, in magazines and in newspapers as an activity for and by men. Is it too far a reach to consider that if enough depictions of men performing gymnastics were presented, that gymnastics would eventually become gender masculine by virtue of the images themselves? Conversely, consistent, repetitive images of women playing football could increase acceptance of that sport as gender neutral.

This is what Bandura (1986) suggests. With enough repetition of an activity by a men or women, the "gendering" of that activity can be changed. However, he adds that media influence may be more powerful than even lived experience in the gendering of activities. The results of this study seem to support that assertion. The problem, then, becomes changing the images, as they seem to be more powerful than even participation rates in many sports. Until this happens, we predict that participation rates will do relatively little more than provide degrees of difference in the way men and women conceptualize sports as gender-appropriate.

The consequences of the stagnation of gender-typing in sports are not insignificant. First, as Witt (1997) points out, ideas about what is appropriate and not appropriate limits the potential of both men and women. The social pressure to conform is great, as demonstrated in research about sport participation and perceived desirability as a friend or date (Holland & Andre, 1994). The price is higher for women than for men; as they ponder their futures and make decisions about appropriate gendered activity, teenage girls drop out of sports at a rate that is six times higher than that of boys (Benefits, 2000). In doing so, they miss out on the benefits of sports participation.

The questions raised by the results of this research provide opportunities for exploration. For instance, survey responses indicate some interaction between ethnicity and gender-typing of sports. Research could explore how race interacts with perceptions of sports as gendered. Further, qualitative research involving younger participants and a wider array of sports would also help illuminate the interaction of television consumption with other types of socialization in attitudes about sports and gender.

References

About NWFA (n.d.). National Women's Football Association Web site. Available at: http://www.womensfootballcentral.com/about.html

Acosta, R. V. & Carpenter, L. J. (2004). Women in intercollegiate sport: A longitudinal, national study. National Association for Girls and Women in Sport. Retrieved January 1, 2005 from http://www.aahperd.org/nagws/template.cfm?template=acostacarpenter.html

Adams, T., & Tuggle, C. A. (2004). ESPN's SportsCenter and women's athletics: 'It's a boys' club.' Mass Communication & Society, 7, 237-248.

Andre, T., & Holland, A. (1995). Relationship of sport participation to sex role orientation and attitudes toward women among high school males and females. Journal of Sport Behavior, 18, 241-254.

Associated Press (2007, 16 Sept.). Love of game leads women to pro league. The New York Times Web site. Available at: http://www.nytimes.com/2007/09/16/sports/baseball/16wpfl.html

Bandura, A. (1986). Social foundations of thought and action: A social cognitive theory. Englewood Cliffs, N.J.: Prentice Hall.

Benefits--Why sports participation for girls and women: The foundation position. (2000, Aug. 14). Women's Sports Foundation Web site. Available at: http://www.womensportsfoundation.org/cgi-bin/iowa/issues/body/ article.html?record=577

Bernstein, A. (2002). Is it time for a victory lap? International Review for the Sociology of Sport 37(fall-winter), 415-428.

Bishop, R. (2003). Missing in action: Feature coverage of women's sports in Sports Illustrated. Journal of Sport and Social Issues, 27, 184-194.

Boyle, J. E. (1997). Organized sports participation, masculinity, and attitudes toward women. Unpublished thesis. Virginia Polytechnic Institute and State University, Blacksburg.

Brannon, R., & Juni, S. (1984). A scale for measuring attitudes about masculinity. Psychological Documents, 14, Doc. #2612.

Carpenter, L. J., & Acosta, R. V. (2005). Title LX. Champaign, I11.: Human Kinetics.

Cashmore, E. (2005). Making sense of sports (4th ed.). London: New York.

Duncan, M. C., & Sayaovong, A. (1990). Photographic images and gender in Sports Illustrated for Kids. Play and Culture, 3, 91-116.

Hardin, M., Lynn, S., Walsdorf, K., & Hardin, B. (2002). The framing of sexual difference in Sports Illustrated for Kids editorial photos. Mass Communication and Society, 5, 341-360.

Hargreaves, J. (1994). Sporting females: Critical issues in the history and sociology of women's sports. London: Routledge.

Holland, A., & Thomas, A. (1994). Athletic participation and the social status of adolescent males and females. Youth & Society 25, 388-407.

Houseworth, S., Peplow, K., & Thirer, J. (1989). Influence of sport participation upon sex role orientation of Caucasian males and their attitudes toward women. Sex Roles, 20, 317-325.

Kane, M. J., Griffin, P., & Messner, M. (2002). Playing unfair: The media image of the female athlete. [Motion picture]. (Available from Media Education Foundation, 60 Masonic Street, Northampton, MA 01060)

Koivula, N. (2001). Perceived characteristics of sports categorized as gender-neutral, feminine and masculine. Journal of Sport Behavior, 24, 377-393.

Markovits, A. S., & Hellerman, S. L. (2001). Offside: Soccer and American Exceptionalism. Princeton, N.J.: Princeton University Press.

Matteo, S. (1986). The effect of sex and gender-schematic processing on sport participation. Sex Roles, 15, 417-431.

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

Metheny, E. (1965). Symbolic forms of movement: The feminine image in sports. In E. Metheny (Ed.), Connotations of movement in sport and dance (pp. 43-56). Dubuque, IA: Brown.

Nilges, L. M. (1998). I thought only fairy tales had supernatural power: A radical feminist analysis of Title IX in physical education. Journal of Teaching in Physical Education, 17, 172-194.

Pedersen, P. M. (2002). Investigating interscholastic equity on the sports page: A content analysis of high school athletics newspaper articles. Sociology of Sport Journal, 19, 419-432.

Postow, B. C. (1980). Women and masculine sports. Journal of the Philosophy of Sport, VII, 51-58.

Riemer, B. A., & Feltz, D. L. (1995). The influence of sport appropriateness and image on the status of female athletes. Women in Sport & Physical Activity Journal, 4, 1.

Solmon, M. A., Lee, A. M., Belcher, D., Harrison, L., & Wells, L. (2003). Beliefs about gender appropriateness, ability, and competence in physical activity. Journal of Teaching in Physical Education, 22, 261-279. Sports Illustrated for Women (2002). Media Kit. New York: Time, Inc.

Tuggle, C., Huffman, S., & Rosengard, D. (2002). A descriptive analysis of NBC's coverage of the 2000 Summer Olympics. Mass Communication & Society, 5,361-375.

Marie Hardin

Penn State University

Jennifer D. Greer

University of Alabama

(1) The term "sex-appropriate" is also sometimes used in literature about masculinity, femininity, and sports. We choose the term "gender" to recognize the distinction between sex as referring to that which is within the biological realm, and gender, referring to identification, behaviors and characteristics outside the biological realm ascribed to men and women (Caudwell, 1999).

(2) Sports offered at the university for intercollegiate competition and for fans to attend are as follows: Men's intercollegiate sports are baseball, basketball, cross country, football, gymnastics, indoor track, soccer, swimming, tennis, outdoor track, and wrestling. Women's sports are basketball, cross country, field hockey, golf, gymnastics, indoor track, rowing, soccer, softball, swimming, tennis, outdoor track, and volleyball.

(3) For analyses, ethnicity was coded as Caucasians vs. other groups. This was done because of the small number of respondents in each group other than Caucasian. However, post hoc analyses were performed on any significant finding for ethnicity to help better understand the idea.

(4) The names of the factors were chosen based on previous research that identified the sports in these factors as masculine, feminine, or neutral. In addition, the researchers looked at the means of these sports on the masculine-feminine continuum to ensure that the names were appropriate. Action sports, because they emerged as distinct factor, were named as such without reference to where they fell on the masculine-feminine continuum.

Address Correspondence To: Marie Hardin, Penn State University, 222 Carnegie, University Park, PA, 16801, Phone: 814-865-1395, E-mail: Mch208@psu.edu

Table 1. Oblique Promax rotation of 4 factors of 14 masculinity
attitude variables.

Sport Factor 1 Factor 2 Factor 3 Factor 4
 Action Hyper Masculine Neutral Hyper Feminine

Wakeboarding .818 .303 .056 -.047
Surfing .790 .330 -.017 -.095
Snowboarding .761 .297 .046 -.271
Skateboarding .668 .555 .118 -.411
Motocross .653 .560 .058 -.334

Football .229 .763 .088 -.262
Basketball .304 .665 -.135 .122
Weightlifting .366 .659 -.147 -.258
Rugby .398 .637 .078 -.114

Swimming -.018 -.160 .798 .050
Tennis -.057 -.015 .683 .457
Soccer .134 .050 .432 -.091

Volleyball -.186 -.231 .050 .844
Gymnastics -.193 -.638 .329 .430

Table 2. Attitudes toward masculinity of span.

Sport Overall Men Women
 mean

Football 4.81 4.87 4.78
Weightlifting 4.54 4.64 4.50
Rugby 4.21 4.37 4.13
Basketball 3.65 3.95 3.51
 (+) Hyper masculine avg. 4.08 4.22 4.02

Motocross 4.19 4.10 4.23
Skateboarding 4.09 4.00 4.13
Snowboarding 3.48 3.50 3.46
Wakeboarding 3.45 3.36 3.49
Surfing 3.32 3.41 3.28
 (+) Action average 3.70 3.68 3.72

Soccer 3.08 3.06 3.09
Swimming 2.89 2.76 2.94
Tennis 2.78 2.80 2.78
 (+) Neutral average 2.92 2.87 2.94

Volleyball 2.38 2.44 2.35
Gymnastics 1.93 1.83 1.98
 (+) Feminine average 2.16 2.14 2.16

 Avg. attitude, all 3.49 3.51 3.48
 sports

Sport Difference by Effect size, Gender
 Gender (F) [R.sup.2],
 Adjusted [R.sup.2]

Football 3.012 .009 (.006)
Weightlifting 3.788 .0l1 (.008)
Rugby 5.963 * .018 (.015)
Basketball 29.991 *** 084 (.081)
 (+) Hyper masculine avg. 19.106 *** .054 (.051)

Motocross 1.730 .005 (.002)
Skateboarding 2.796 008 (.005)
Snowboarding .316 .001 (-.002)
Wakeboarding 2.729 008 (.005)
Surfing 3.574 .011 (.008)
 (+) Action average .495 .001 (-.002)

Soccer .816 .002 (-.001)
Swimming 11.138 *** .033 (.030)
Tennis .002 .000 (-.003)
 (+) Neutral average 2.991 009 (.006)

Volleyball 1.224 004 (.001)
Gymnastics 3.107 .009 (.006)
 (+) Feminine average .117 000 (-.003)

 Avg. attitude, all
 sports

Note. (+) = Separate MANOVA for these four averages. * p <.05;
*** p < .001 based on a comparisons of all sports by gender in a
Multiple Analysis of Variance (MANOVA). Effect sizes for Gender on
DVs.

Table 3. Global MANOVA, Variables related to attitudes toward
categories of sports

Sources of variation Hyper Action
 masculine Type III SS
 Type III SS ([F.sup.sig])
 ([F.sup.sig])

Part. in organized sports (1-4) .06 (39) .16 (.66)
Part. in personal fitness (1-4) 1.09 (7.52 **) 1.79
 (7.19 **)
Part. in Hyper masculine .32 (2.2) .14 (.54)
 sports (0-4)
Part. in X sports (0-4) .31 (2.15) .003 (.01)
Part. in neutral sports (0-4) .12 (.80) .05 (.20)
Part. in feminine sports (0-4) .09 (.63) .11 (.43)
Time watching TV (1-5) .06 (.42) .01 (.05)
Time using sport media (1-5) .18 (1.25) .19 (.78)
Watch Hyper masculine sports (1-5) .63 (4.33 *) .10 (.42)
Watch Action spats (1-5) .39 (2.66) .15 (.62)
Watch Neutral sports (1-5) .000 (.00) .19 (.75)
Watch Feminine sports (1-5) .000 (.00) .39 (1.57)
BMS average (1-4) .003 (.02) .69 (2.75 (+))
Sex (1=male; 2=female) .14 (.99) .66 (2.69)
Ethnicity (1=white, 2=minority) 1.11 (7.64 **) .52 (2.08)
Corrected model: F & sig. 2.73 *** 1.41 (+)
Model summary: [R.sup.2] .209 (.133) .121 (.035)
 (Adjusted [R.sup.2])

Sources of variation Neutral Pillars
 Type III SS Trace
 ([F.sup.sig]) ([F.sup.sig])

Part. in organized sports (1-4) .04 (.43) 1.227
Part. in personal fitness (1-4) .17 (1.9) 2.690 *
Part. in Hyper masculine .05 (.52) .798
 sports (0-4)
Part. in X sports (0-4) .01 (.12) 1.119
Part. in neutral sports (0-4) .29 (3.31 (+)) 1.020
Part. in feminine sports (0-4) .39 (4.38 *) 2.134 (+)
Time watching TV (1-5) .15 (1.68) .778
Time using sport media (1-5) .69 (7.85 **) 2.541 *
Watch Hyper masculine sports (1-5) .08 (.93) 1.741
Watch Action spats (1-5) .01 (.15) .948
Watch Neutral sports (1-5) 3.59 10.804 ***
 (40.75 ***)
Watch Feminine sports (1-5) 1.32 5.229 ***
 (15.04 ***)
BMS average (1-4) .44 (4.98 *) 3.000 *
Sex (1=male; 2=female) .08 (.87) .923
Ethnicity (1=white, 2=minority) .22 (2.49) 2.666 *
Corrected model: F & sig. 4.24 ***
Model summary: [R.sup.2] .291 (.223)
 (Adjusted [R.sup.2])

Note. (+) p < .1 * p < .05; ** p < .01; *** p < .001

Table 4. Significant interactions with sex of respondent from Global
MANOVA.

Interaction Sex and Hyper masculine Action
 Type III SS Type III SS
 ([F.sup.sig]) ([F.sup.sig])

Part. in personal fitness (1-4) .12 (.82) .81 (3.25 (+))
Part. in feminine sports (0-4) .39 (2.69) .56 (2.26)
Watch neutral sports (1-5) .22 (1.49) .05 (.18)
Watch feminine sports (1-5) .03 (.24) .35 (1.40)
BMS average (1-4) .07 (.50) 1.32 (5.28 *)

Interaction Sex and Neutral
 Type III SS Pillai's Trace
 ([F.sup.sig]) ([F.sup.sig])

Part. in personal fitness (1-4) .59 (6.65 **) 2.586 *
Part. in feminine sports (0-4) .98 (11.14 ***) 3.680 **
Watch neutral sports (1-5) 2.58 (29.34 ***) 7.635 ***
Watch feminine sports (1-5) 1.16(13.20 ***) 4.479 **
BMS average (1-4) .44 (4.99 *) 3.000 *

Note. (+) p < .1 * p < .05; ** p < .01; ** p < .001