Author: Taylor, Matthew J
Date published: September 1, 2010
There is a longstanding tradition that suggests the notion that participation in sports enhances moral development. This concept is far from new, as Plato in his Republic asserts the importance of the integration of sport into academic pursuits [curricula] (Arnold, 1984). The first practical representation of this practice developed in the mid-nineteenth century in the English public schools, whereby sport was utilized as a means of social control, in that it replaced the less socially redeeming activities of bullying, poaching, drunkenness, and vandalism which often accompanied the leisure time of "schoolboys" (Donnelly, 1981; Segrave, 1 983). A similar phenomena took place in America, as sport began to play a prominent role in the American educational system (Schäfer, 1 97 1 ), again due in large part to its perceived ability to "keep young males on the straight and narrow [path]" (Camp, 1913).
The early discussion of sport and its multiple benefits did not include the potential of these outcomes for women and girls. In general, the history of sport suggests that it has been a male-dominated institution, "created [for] and shaped by men, without regard to the existence and experience of women" (Boutilier & SanGiovanni, 1983, p. 17; Brown, Frankel, & Fennell, 1989). Although in the late 19th century small numbers of women did begin to compete at the intercollegiate level, it was not until Title IX ( 1 972) that the numbers increased in earnest. For younger girls there has also been a change in participation levels. During the 2005-2006 high school year, almost 3 million girls participated in school-sponsored sports; representing roughly 41% of high school athletes and the highest rate of participation ever (National Federation of State High School Associations, 2006).
Impact of sport participation
Substance use. In terms of substance use, there have been data which suggest that participants in sports [physical fitness] programs are less likely to use alcohol and other drugs (Chung & Elias, 1996; Collingwood, Reynolds, Kohl, Smith, & Sloan, 1991 ; Koss & Gaines, 1993). One could surmise that protective mechanisms involved here are that athletes do not want to impede their performance or jeopardize their place on the team by using alcohol or other substances. Another protective hypothesis is that sport participation represents adolescent engagement in pro-social, peer oriented activities that occur under adult supervision, which lessen the likelihood of substance use (Wallace & Bachman, 1 99 1 ). However, a number of studies have found contradictory evidence, which suggests that delinquency substance use was [positively] related to sports participation (Blood, 1990; Bush & Iannotti, 1992; Eccles & Barber, 1999). DuRant, Middleman, Faulkner, Emans, and Woods (1997) found that girls in their sample that participated in sports reported higher levels of alcohol use, binge drinking, and marijuana use than non-participants. More recently and similarly, Miller et al. (2003) found that girls who self-identified as "jocks" had increased rates of binge drinking.
Self-esteem and school adjustment. Self-esteem, the evaluation of how one feels about one's self-concept, is believed to be positively impacted by sport participation. A number of studies have suggested that girls who participate in sports have higher levels of self-esteem and a more positive self-concept (Butcher, 1 989; Foon, 1 989; Tiggemann, 200 1 ; Weiler, 2003). The specific mechanism of this outcome of sport participation appears to have direct and indirect antecedents. It has been suggested that the sport participation-self-esteem relationship is mediated by enhanced feelings of physical competence (Richman, 2002; Richman & Shaffer, 2000). Other research has suggested that the relationship between sport participation and self-esteem may be mediated by school adjustment (Tracy & Erkut, 2002).
The domain of school adjustment can be operationalized as the combination of variables that tap into the overall school experience (e.g., attitudes toward school and teachers, connectedness to school, academic acumen, etc.). Positive school adjustment has been linked to decreased substance use (Chiarella, 2003; Marquait, 2003; Napoli, Marsiglia, & Kulis, 2003; Resnick et al., 1997). Related to sports participation, girls that played sports reported higher grades and more motivation to pursue post-secondary education, than their non-athlete counterparts (Perry-Bumey & Takyi, 2002). However, the exact mechanism through which sports participation impacts school adjustment is unknown. One could speculate that athletes feel external pressure to maintain certain grade point averages in order to remain academically eligible for participation. Another notion is less rooted in the academic realm and more focused on social elements and status. Namely, being an athlete in many school settings may increase popularity and social status, which in turn, would enhance positive feelings about school, and thus improve school adjustment (Buhrmann & Bratton, 1977; Melnick & Sabo, 1992).
Peers, school (adjustment), and substance use. Behaviors, both normative and deviant, are socially learned from the main sources of socialization which are family, school, and peers (Oetting & Donnermeyer, 1998). Rooted in the family constellation, the child is initially impacted almost exclusively by its socialization practices/processes. Upon entrance into an educational environment, whether it be classified as non-formal (e.g., daycare) or formal/ traditional (e.g., school), the child is exposed to additional socialization agents, the most influential of which include non-family member caregivers and teachers; and, to a lesser extent, the peer group. It is when the child reaches early adolescence and begins the process of psychologically (and physically) moving away from the family (and school agents), that the influence and role of the peer group, or "lifestyle" cluster, is paramount (Oetting & Beauvais, 1987a, 1987b).
Given a significant proportion of adolescent time is spent within a school setting, it follows that this represents a main forum within which the majority of peer interactions occur. As such, understanding the interplay between school-context variables and those of the various peer influences and the additive impact of all of these on individual behavior is of importance, especially as related to deviant behavior (i.e., substance use). Impaired peer relationships and negative interpersonal experiences coincide with school adjustment problems (Harris-Britt, 2004). There is some evidence to suggest mat socially isolated youth are more at risk for substance use (Pearson et al., 2006). Difficulties in adjusting to the school setting may result in expectations mat an adolescent will engage in norm-breaking behaviors (Maatta, Stattin, & Nurmi, 2006). Aside from external expectations, and the specter of selffulfilling prophesy, a number of studies strongly suggest that poor school adjustment (and a resultant dropout) increases involvement with drugs, violence, and other deviant behaviors (Beauvais, 1992; Chavez& Swaim, 1992; Lundman, 1986; Miller & Ohlin, 1985; Swaim, Beauvais, Chavez, & Oetting, 1997).
Peer substance use is a well identified risk factor to adolescent use (Chabrol et al., 2006; Nation & Heflinger, 2006). A bi-directional, reciprocal relationship between individual substance use and that of peers/friends is well documented, in that individual use predicts a higher number of substance using peers, which, over time, impacts individual substance use progression (Dinges & Oetting, 1993; Simons-Morton & Chen, 2006). Simply stated, substance using adolescents have peer clusters that use substances and are thus more likely to support norms for drug use (Chavez, Deffenbacher, & Wayman, 1 996). Additionally, as a youth becomes more able to operate with some degree of independence from parental/caregiver monitoring, there are potential negative effects depending on the nature of the family dynamics. In the absence of strong familial bonds, youth may develop associations with peers experiencing similar situations and be more likely to participate in deviant activities (Shanahan, Elder, Burchinal, & Conger, 1996). Not only would these influences serve to impede the successful transmission of pro-social norms, but may lead to weaker family sanctions against adolescent substance use. Youth in this socialization scenario are prone to enter a peer cluster that is more tolerant of deviant behavior (Conger et al., 1991; Salts, Lindholm, Goddard, & Duncan, 1995).
One could surmise that in large urban areas, there would be some ability of a youth to more selectively choose her peer group, given larger numbers of potential peer clusters, and, if desired, have the potential for minimal overlap between groups. Adolescents located in less dense social networks (e.g., rural locales), by default have fewer peer "options" and fewer mutually exclusive peer groups. Therefore, they may be in closer proximity to (potentially) deviant peers, and thus more likely to engage in substance using behaviors (Ennett et al., 2006). An important tool in the prevention of adolescent substance use is to limit associations with deviant peers (Barnes, Hoffman, Weite, Farrell, & Dintcheff, 2006).
Rural youth and substance use. To date, the bulk of the sports participation research has focused primarily on four groups: males, urban dwellers, Whites, and college-age subjects. There remains some question about how salient the findings of the female sports participation research area are in relation to ethnic girls and/or adolescent girls, especially from rural areas. This is a significant oversight for there are some recent trends in rural areas that would seem to make this an important area of study. More than one-fifth of the United States population lives in a rural area, as characterized by a non-metropolitan, small urban cluster; with one third of this population being under the age of 1 8 (U. S. Census Bureau, 2000). Additionally, the most recent census data also suggest that 20-25% of African Americans reside in rural areas.
As the myth of the pure and pristine rural setting remains firmly entrenched in the collective psyche of the nation, a growing body of literature suggests that this notion is far from accurate. In recent decades, rural areas have seen a significant increase in problems and behaviors once viewed exclusively as "urban" occurrences (Cronk & Sarvela, 1997; Monsey, Owen, Zierman, Lambert, & Hyman, 1 995). Over the last 1 0 to 1 5 years, the study of substance use among adolescents living in rural areas has received a significant amount of attention (Edwards, 1992; Cronk & Sarvela, 1997;ResnicketaI., 1997). A number of these studies have concluded that the substance use rates of rural youth are only slightly lower than those found among their urban counterparts (Elliott, Huizinga, & Menard, 1989; Leukefeld, Clayton, & Myers, 1992). Moreover, for some substances (e.g., alcohol, cigarettes, and methamphetamine), use among rural adolescents is frequently higher than their urban counterparts, with shrinking differences in the use of other substances (Cronk & Sarvela, 1 997). The most recent Monitoring the Future data provide the most convincing evidence in that they suggests while adolescent substance use in urban areas has declined, there has not been a similar shift within non-urban locations (Johnston, O'Malley, Bachman, & Schulenberg, 2005).
A model of sport participation as a deterrent to adolescent substance use
Given the current state of the extant literature on sports participation and its impact on adolescent life, specifically substance use, an accurate conclusion is that its specific mechanism of action is far from apparent, although we believe that the link from sport participation to adolescent substance use is not direct, but rather operates through a series of moderating and mediating variables (e.g., school adjustment, self-esteem, and peer substance use). In an attempt to further understand and demonstrate the relationship between sports participation and adolescent substance use, we developed a model built upon the foundations of the health belief model, and the theories of social control and peer clusters. The health belief model (HBM) was developed to identify individual and behavior modifying factors which influence outcomes related to the initiation and maintenance of health promoting and preventative behaviors and the avoidance of unhealthy ones (Janz & Becker, 1984). The aforementioned findings which suggest that those who participate in sport are less likely to engage in negative health-related behaviors are partially explained by the HBM. Namely, individuals may believe that the positive psychological (e.g., self-esteem), physical (e.g., physical competence), and social (e.g., group belonging) benefits of sport are potentially compromised and threatened if they chose to use substances - and thus they refrain. Within the context of our model this is represented by the perceived relationship between sport participation and self-esteem, which mediates the relationship between participation and individual substance use.
Social control theory (Hirschi, 1 969) suggests that individuals participate in delinquent, counter-normative behaviors when their social bonds to society (and its groups) are weakened. The rationale as it relates to sport participation is that social identification with a prosocial entity (e.g., organized and supervised team) will result in adherence to pro-social values and the production of similar behavior. Consider that when an individual participates in a sport, they not only represent themselves, but also a collection of athletes and the school or community entity. Even in the context of more individualistic sports, such as tennis and track, individuals are still part of a group/team. Specific to school-based participation the social responsibility to preserve the opportunity to participate may come from the requirement to maintain grades to remain eligible or, at the very least, to be a good school citizen; both of which are represented in our model by the relationship with school adjustment.
As adolescents become more entrenched in the school experience and gravitate away from family, the peer group becomes the major force of socialization. Oetting and Donnermeyer (1998) suggest 4 types of peer influence; 1) peers in general; 2) peer groups (e.g., formal and informal groups such as a classroom); 3) peer lifestyle groups (e.g., the drug lifestyle group); and 4) peer clusters, which are "smaller (more intimate and influential) subsets of peer groups and lifestyle groups" (p. 1011). It is these peer clusters that have been found to be the most influential determinants and predictors of adolescent substance use; which is the foundation of peer cluster theory (Oetting & Beauvais, 1987a, 1987b). The impact of peers is represented in a variety of ways in our model. First, peer use represents the most direct and influential predictor of individual use and serves as a mediator in the relationship between sports participation and individual substance use. Additionally, our model suggests that sports participation is related to school adjustment and self-esteem outcomes that in turn influence adolescent "choice" of similarly oriented and behaving peers.
Objectives & goals
The objective of this study was to explore the impact of sports participation on selfesteem, school adjustment, and substance use among rural and urban dwelling African American high school girls. To this end, there were two primary goals of this study. The first was to test the hypotheses that sports participation would predict: better school adjustment, higher self-esteem, association with peers who use fewer substance, and less individual substance use. The second goal of this study was to develop a model of adolescent substance use using the relationship of sports participation to school adjustment, self-esteem, peer use, and individual use. It was hypothesized that the effects of sports participation would be positively related to self-esteem and school adjustment and negatively associated with peer and individual (own) substance use. The fit of this model was compared for both rural and urban samples given: 1) the lack of rural-specific focus in the extant research; and 2) the potential differential effects of location (e.g., limited peer groups in rural areas).
Participants were 1 ,976 female African American high school students. Participants were from rural (n = 1 ,048; 53%) and urban (n = 928; 47%) areas distributed across the United States, the criteria for which established using the Metropolitan Proximity Index (Labao, 1 990). Rural was defined as non-metropolitan areas with a main population center containing 2,500-9,999 people and not adjacent to a county of the same size. Urban was defined as metropolitan areas and associated core counties containing more than 500,000 people. Only 657 (33 .2%) participated in sports, and a significantly larger percentage participated in sports in the rural sample (35.6%; ? = 373) than the in the urban sample (30.6%; ? = 284), ?2(1, N= 1976) = 5.52, ? = .019. The sample included 528 freshmen (26.7%), 5 1 8 sophomores (26.2%), 507 juniors (25.7%), and 423 seniors (2 1 .4%). The mean age of participants was 1 6.03 years (SD = 1 .24).
Data were obtained from a random selection of female, African American high school students from a previously collected dataset.1 Primary variables of study (school adjustment, self-esteem, and own and peer alcohol/drug use) were based upon items collected by the parent project using the Community Drug and Alcohol Survey (CDAS), which is based on adaptations of the American Drug and Alcohol Survey (ADAS) and the companion Prevention Planning Survey, (PPS; Oetting & Beauvais, 1990; Oetting, Beauvais, Edwards, & Waters, 1 984). Both instruments are anonymous self-report, paper-and-pencil surveys of alcohol (and drug) use, and items concerning risk and protective factors for alcohol/drug use (e.g., negative affect, family and peer relationships, school environment), respectively. Individual items were grouped into the scales and resultant variables of study using previously established instrument guidelines and procedures (Oetting & Beauvais, 1990; Oetting et al., 1984).
A dichotomous sports participation variable was created by participant responses to items about involvement in formalized sad structured 'sport activities at and away from school. To this end, sport activities, such as "pick up" games or other more informal scenarios (e.g., roller-blading with friends) were not included as participation. As such, affirmative responses to structured and organized sport participation, regardless of being school or communitybased, resulted in a participant being placed in the "play" group. For the school adjustment and self-esteem variables, individual item responses were comprised of 4-point Likert scales, ranging from "A lot" to "Not at all." School adjustment was operationalized by items associated with attitudes toward teachers (e.g., "I like my teachers"), toward the school experience (e.g., "I like school"), and school performance (e.g., "I do well in school") (6 items; a = .80). Self-esteem was comprised of items focused on social acceptance (e.g., "Other people my age like me"), social confidence (e.g., "I am good at games"), and social competence (e.g., "I like myself) (1 1 items; a = .79). In measuring peer substance use, participants were asked items that related to peer use, such as "How many of your friends use _____ (e.g., methamphetamines)?" and "How many of your friends get drunk?"; with responses ranging along a 4point Likert scale, from none to all of them (6 items; a = .74). Individual (own) substance use was comprised of items focused on alcohol, marijuana, and other illicit drugs (e.g., methamphetamines, cocaine, LSD) and included frequency of use in the last month and self-categorization of use as described by a 6-point Likert scale with the anchors of non-user to very heavy user (1 7 items; a = .73).
In accordance with the first hypothesis, we tested the multivariate prediction of four sets of dependent variables by two independent variables and their interaction. The four sets of dependent variables were school adjustment (attitudes toward teachers, attitudes toward school, and school performance), self-esteem (social acceptance, social confidence, and social competence), peer substance use (peer use of alcohol, marijuana, and other illicit drugs), and own substance use (own use of alcohol, marijuana, and other illicit drugs). The two independent variables were sports participation and location. Pairs of two-way between-subjects factorial multivariate analyses of variance (MANOVAs) were preformed for each set of dependent variables using the general linear model, first with the two independent variables to test the multivariate main effects and then again with the multiplicative relation between the two added to test the multivariate interaction. Significant multivariate main effects and interactions were followed up with post hoc univariate main effects tests.
In the pair of MANOVAs for school adjustment, the multivariate main effects for sports participation, F(3, 1971)= 11. 10,p=.000, and location, F(3, 1971)=5.21,p=.001, were significant, but the multivariate interaction, F(3, 1970) = 0.47, ? = .700, was nonsignificant. In the univariate ANOVAs corresponding with the significant multivariate main effects, those who participated in sports had significantly higher scores for attitudes toward teachers (M= 6.45, SD = 1 .33), attitudes toward school (M= 5.75, SD = 1 .7 1 ), and school performance (M= 6.77, SD = 1 .05) than did those who did not participate in sports (Ms = 6.27, 5.42, & 6.50, and SDs = 1 .37, 1.71,& 1.06, respectively), Fs(1, 1973) = 7.86, 15.88,&26.23,ps = .005,.000,&.000, A=.13,.19, & .24, respectively. Those in urban locations had significantly lower scores for school performance (M= 6.49, SD = 1 .07) than did those in rural locations (M= 6.69, SD = 1 .05), F(1 , 1 973) = 14.91 , ? = .000, d=A7, but the differences in scores for attitudes toward teachers and attitudes toward school of those in urban and rural locations were nonsignificant, Fs(1, 1973) = 0.39 & 0. 14,ps = .534& .71 1, ufe = .03 & .02, respectively.
In the pair of MANOVAs for self-esteem, the multivariate main effect for sports participation, F(3, 1 97 1) = 20.25, ? = .000, was significant, but the multivariate main effect for location, F(3, 1971) = 0.39,p = .760, and the multivariate interaction, F(3, 1970) = 0.84,p = .472, were nonsignificant. In the univariate ANOVAs corresponding with the significant multivariate main effect, those who participated in sports had significantly higher scores for social acceptance (M= 1 3 .82, SD = 2.54) and social competence (M= 1 3 .95, SD = 2. 1 5) than did those who did not participate in sports (Ms = 1 3 .33 & 1 3.23, and SDs = 2.62 & 2. 1 8, respectively), Fs(I , 1973) = 15.53 & 46.75, ps = .000 & .000, ds = .18 & .31, respectively, but there were no differences in scores for social confidence, F(1 , 1 973) = 0. 14,p = .7 1 0, d= .02.
In the pair of MANOVAs for peer substance use, the multivariate main effects for sports participation, F(3, 1 971) = 2.57, p = . 053, and location, F(3, 1971)= 1.22,p = . 302, were nonsignificant, but the multivariate interaction, F(3, 1 970) = 3 .96, ? = .008, was significant. Given the nonsignificant main effects, we examined the significant interaction by simply comparing urban and rural locations in terms of the overall pattern of mean differences in scores on the three dependent variables of those who did and did not participate in sports. Accordingly, those who participated in sports in rural locations had higher scores on all three dependent variables, peer use of alcohol (M =2.16, SD = 0.90), marijuana (M= 1 .9 1 , SD = 0.9 1 ), and other illicit drugs (M= 4.45, SD = 1 .69), than those who did not participate in sports (Ms =2.10, 1.88, & 4.23, SDs = 0.90, 0.86, & 0.99, and ds = .07, .03, & . 1 8, respectively). Comparatively, those who participated in sports in urban locations only had higher scores for peer use of alcohol (M = 2.12, SD = 0.96) and instead had lower scores for peer use of marijuana (M= 1 .80, SD = 0.86) and other illicit drugs (M= 4.25, SD = 1 . 1 1) in contrast to tiiose who did not participate in sports (Ms = 2.0 1, 1. 87, 4.36, SDs = 0.9 1, 0.88, & 1.36, and as = . 12,. 08, &. 08, respectively).
In the pair of MANOVAs for own substance use, the multivariate main effects for sports participation,/^, 1971)= 1 .94,p=. 121, location, F\3, 1971) =2.09, ? = .099, and me multivariate interaction, F(3, 1970) = 0.29,p = .833, were all nonsignificant.
Consistent witìi the hypothesized prediction of school adjustment, self-esteem, and peer and own substance use by sports participation, multivariate main effects of sports participation were found for school adjustment and self-esteem. As expected, exploration of the corresponding univariate main effects suggested mat sports participation generally predicted higher school adjustment (higher for all three dependent variables) and self-esteem (higher for social acceptance and social competence but no difference for social confidence). Although a multivariate main effect of sports participation on peer substance use was expected, only a multivariate interaction witìi location was found, and an exploratory examination of die overall pattern of means did not provide an unambiguous interpretation. It appears as though sports participation predicts higher peer use of alcohol, marijuana, and other illicit drugs in rural locations but only higher peer use of alcohol and instead lower peer use of marijuana and other illicit drugs in urban locations. Thus, as expected, sports participation predicted lower peer use of marijuana and otiier illicit drugs, but only in urban locations, whereas sports participation unexpectedly predicted higher peer use of alcohol regardless of location and higher use of marijuana and other illicit drugs in rural locations. Contrary to expectation, no multivariate main effect of sports participation was found for own substance use. Regarding the possible roles of location, in addition to the multivariate interaction of location with sports participation on peer substance use, a multivariate main effect of location on school adjustment was found. Exploration of the corresponding univariate main effects suggested that compared to rural locations urban locations were predictive of higher school adjustment for school performance but not for attitudes toward teachers and attitudes toward school.
Structural Equation Modeling
In accordance with die second hypothesis, the hypothesized structural equation model (SEM) was tested on the urban and rural samples, initially as separate SEM analyses and then as a multisample SEM analysis. LISREL 8.72 was used to carry out the SEM analyses. The factor loadings from die latent variables to dieir respective manifest variables in the measurement model were all significant in both the separate and multisample SEM analyses, so the remainder of the results focuses primarily on the structural model.
Overall, the hypothesized model fit well in the urban sample, as indicated by the following fit indices: ?2(70) = 228.58 (? < .05), GFI = .97, TLI = .93, CFI = .94, and RMSEA = .049 (? = .540). Although the chi-square was significant, which means that the model did not meet the unrealistic expectation that it fits perfectly, the GFI value exceeded .95 suggesting that the absolute fit of the model is good. The TLI and CFI values exceeded .90, which suggest that the improvement in fit of the model over the null model is adequate. The RMSEA value was under .05 and low enough for the test of lack of close fit to be nonsignificant, which suggests that the approximate fit of the model is good. As shown in Figure 1 , all paths were significant except for the ones from sports participation to self-esteem and from own substance use to school adjustment.
Overall, the hypothesized model did not fit well in the rural sample, as indicated by the following fit indices: ?2 (70) = 362.80 (? < .05), GFI = .95, TLI = .87, CFI = .90, and RMSEA = .063 (? = .000). The GFI value did not exceed .95 suggesting that absolute fit of the model is only adequate, and the TLI and CFI values did not exceed .90, which suggest that the improvement in fit of the model over the null model is not adequate. The RMSEA value was not under .05 and high enough for the test of lack of close fit to be significant, which suggests that the approximate fit of the model was not good, though it was below .08 suggesting that the approximate fit of the model was reasonable. As shown in Figure 1 , all paths were significant except for the one from self-esteem to peer substance use and from own substance use to school adjustment.
Although a comparison of size of the unstandardized path coefficients and the AIC values, lower values for which indicate better fit after adjusting for parsimony, clearly suggest that the hypothesized model fits better in the urban sample (AIC = 298.58) than it does in the rural sample (AIC = 432.80), a multisample SEM analysis was performed to compare its relative fit in the two samples. In comparison with the model in which none of the parameters were constrained to be equal across samples, ?2(140) = 591 .38, the model in which the factor loadings in the measurement model were constrained to be equal across samples, ?2( 1 50) = 598.84, did not fit significantly worse, ??2(10) = 7.46,/? > .05. Likewise, the model in which the paths in the structural model were also constrained to be equal across samples, ?2(1 57) = 608.03, did not fit significantly worse, ??2 (7) = 9 . 19, ? > .05 . Thus, the relative fit of the hypothesized model in the urban and rural samples was not significantly different.
When the findings from the separate and multisample SEM analyses are considered together, it appears as though the hypothesized model is supported in urban, but possibly not, rural locations. It fit well in the urban sample according to the fit indices, but although the multisample SEM analysis found the relative difference in fit between the urban and rural samples to be nonsignificant, the AIC was substantially higher for the rural sample than for the urban sample and it did not fit well in the rural sample according to the fit indices.
Figure 1. Structural equation models for the urban and rural samples. For visual clarity, only the latent variables of the structural model are shown. Solid lines indicate significant paths, and dashed lines indicate nonsignificant paths. The unstandardized path coefficients are reported, such that the variance of sports participation is fixed to 1 and the other latent variables are scaled according to the first of each set of manifest indicator variables. (The standardized path coefficients are reported in parentheses.)
The main objective of this study was supported by two goals. The first was to provide a preliminary exploration of the potential role of sports participation in the lives of African American adolescent girls; given this is a group, especially rural dwelling, which has received little focus in the extant literature. In correspondence with our hypotheses, the participants who played sports had higher levels of school adjustment than the students who did not. More specifically, this finding was rooted in the indicators of more positive attitudes towards teachers and the overall school experience, yet not necessarily always related to school performance (e.g., grades). The fact that the sports participants in rural areas also had a higher level of school adjustment may suggest that in these locales social status gains were driving these outcomes (Buhrmann & Bratton, 1977; Melnick & Sabo, 1992). Students in smaller geographical settings may have fewer options in terms of extracurricular activities aside from athletics, thus through such endeavors, especially in a setting where anonymity is scarce and athletics are highly valued, students may have increased celebrity status and recognition; and this ultimately may relate to enhanced feelings about the school experience. An interesting feature of note here is the finding that within the sample, the girls in rural areas were more likely to participate in sports compared to those living in urban ones, which may relate to the availability of extracurricular activities. In an urban setting, similar experiences and social outcomes of elevated status due to sport participation may be less likely to occur as status may be attained and showcased through a wider variety of activities and with more obscurity because of increased population density.
The finding that self-esteem was higher in girls that participated in sports is consistent with previous research on different populations of study (Foon, 1989; Tiggemann, 200 1). The elements underlying this relationship were enhanced social acceptance (from peers) and increased social competence. The implication of this finding is that social status, and the subsequent positive interpersonal interactions, may have increased as a result of sports participation and this then leads to enhanced self-esteem. The indicator of selfesteem that was not significant as an individual indicator was social confidence. As such, sports participation appeared to be related to enhanced self-esteem stemming more from external mechanisms, namely interpersonal acceptance and competence, and less the augmentation of internal self-perceptions interpersonal abilities (e.g., social confidence). This finding addresses the potential social outcomes of sport impacting self-esteem, whereby previous work examined the sport-self-esteem relationship more as a function of enhanced physical abilities (e.g., physical competence) (Richman, 2002; Richman & Shaffer, 2000). As future work expands to develop a better understanding the meaning of participation, and why girls and boys participate in sports, this notion of sport as a social activity, witìi accompanying social rewards is critical. This potential relationship would appear to be of some consequence in rural communities where the potential for fewer social activities exists; thus if a youth is not playing sports, they may be missing out on prime interpersonal exchanges. Furthermore, while it has long been thought that the participation of girls in sports was due in large part to the social aspect of it and that for boys these activities were considered "more serious" endeavors, with college scholarship implications, these overly simplistic, sex-stereotype-based assumptions have begun to be challenged (Perry-Burney & Takyi,2002).
The notion that sport was a protective tool in relation to peer and individual substance use received mixed support, as there were differential effects of sports participation found between rural and urban settings. Participants who played sports in rural areas reported higher levels of overall peer drug use for all types of substances (i.e., alcohol, marijuana, and other illicit drugs). In urban settings, overall peer use rates were lower, but this was only specific to marijuana and other drugs, not alcohol. Some caution should be used in a dogmatic approach to these conclusions due to small effects sizes. Of note, we found no differences in own self-reported substance use between sport-non-sport and rural-urban groups. However, the (continued) association of athletes with substance using peers, especially in rural areas, would seem to be a potential risk factor to own use, but such an outcome should not necessarily be assumed. There are a multiple themes that would interact to possibly heighten the likelihood of use, regardless of sports participation. Previous work has found that "athlete" status itself linked to substance use, especially alcohol and binge drinking (DuRant et al., 1 997; Miller et al., 2003). Moreover, associations with a using peer group place an adolescent in an interpersonal environment (e.g., peer cluster) which not only supports norms for drug use, but also models the very behavior targeted by prevention efforts (Chavez et al., 1996; Dinges & Oetting, 1993; Simons-Morton & Chen, 2006). As such, simply affording youth an opportunity to participate in sports may oversimplify the interpersonal processes at work. It would make sense to consider the implementation of a multifaceted prevention approach that, in addition to providing a "sporting outlet," is accompanied by corresponding efforts to effectively lessen associations with deviant peers, or at least their impact, specifically in rural areas where complete avoidance of this peer group would not be practical.
The other goals of the study were to the fit of a model of sports participation, school adjusdnent, self-esteem, peer use and own use on rural and urban samples and to assess its efficacy with both samples. While both applications of the model yielded many significant paths for both populations, there was some variability in the absolute fit. Fit indices suggested a better fit of the model when applied to the urban sample. We believe that these results are best explained by the community characteristics of rural areas and that a possible explanation can be offered by discussing some findings of note with attention paid to their respective ecological contexts. To begin with, sports participation was significantly and positively related to school adjustment for both samples; where as sport to self-esteem path was only significant for the rural sample. With regard to school adjustment, this finding is similar to previous work (Perry-Burney & Takyi, 2002). It has been suggested with other populations of study that the impact of sports participation upon self-esteem may be mediated by school adjustment, and our results also suggest this relationship with our sample of African American girls (Tracy & Erkut, 2002). Although, in general, the strength of these relationships was not significantly large, the direct path from sports participation to self-esteem for the rural sample was twice the magnitude than that of the urban one. As previously noted, this may again point to evidence that sports participation facilitates social status, especially in rural areas where the school is often a focal point of community attention, which would then heighten self-esteem of athletes, as a "big fish in a small pond."
Peer use was more strongly associated with individual use in urban areas. This finding is indeed interesting and may suggest that in spite of the risk factor of "proximity" to using peers for rural youth, this is no assurance that individual use will result (Ennett et al., 2006). Peer use was also more negatively associated with school adjustment is urban areas, which suggests that urban youth may be more prone to engage in behaviors similar to that of this particular peer cluster. As previously noted, rural youth have peers that use more, yet may be better able to counter these potentially negative influences, perhaps by being more tolerant of the behavior, but choosing not to mimic it. The finding that individual use was more negatively related to school adjustment in urban areas may offer another piece of evidence, in that youth substance use in rural areas may be seen as more normative and, as such, less taboo and disruptive. Shedler and Block (1 990) found in their sample of adolescents that non-users were less well-adjusted and had more anxiety than "experimenters." This normative use hypothesis would seem to correspond to historical data trends which suggest that the substance use rates for rural youth are higher than urban for certain substances (i.e., alcohol) and have not shown a recent decrease like urban rates (Cronk & Sarvela, 1997; Johnston et al., 2005).
While this project represents an addition to the current literature on sports participation, especially as it relates to the understudied populations of rural, African American girls, care must be taken in interpreting these findings due to the following limitations. As a project that utilized secondary data, there were a number of data issues beyond our control. There was only minimal information concerning the details of subjects' sports participation. Although the structured nature of our participants' involvement could be easily inferred, more specifics about the sport setting (i.e., time spent being involved, degree of participation by friends, level of enjoyment in participation) was not available. Moreover, of particular interest would have been an assessment of the meaning of sport participation, as the motivation and focus of a student doing so for the attainment of a college scholarship may be very different than one participating because close friends are too. The cross-sectional nature of the data, while useful and contributory to our understanding of the variables of study, was a limitation it that it did not offer an opportunity to gauge longitudinal relationships, especially given that as girls advance through high school, they are more likely than their male counterparts to drop out of sports participation (Tucker Center for Research on Girls & Women in Sport, 2007). Another limitation involves the fact that in general, alcohol and drug use rates among our participants were very low. Our noted rates of use of this sample are far from unique, as previous work has documented a similar finding that African American adolescents, especially girls, typically have lower substance use compared to other adolescents (Centers for Disease Control and Prevention, 2006; Johnston, O'Malley, Bachman, & Schulenberg, 2008). This being said, we suggest caution be used in the broad generalization of our finding to other populations, as limited research has specifically focused on these phenomena with (rural) African American girls, and our work represents but an early exploration. And finally, although our explicit intent was to focus our attention specifically on African American youth, a comparison group of non-African American adolescents could have added a layer of understanding to the analyses.
The purpose of this study was to offer some evidence about the impact of sports participation on the lives of African American girls. The results represent an addition to the extant literature and begin to shed some light on how sport impact aspects of adolescent life. The role and mechanism of sports participation and its potential outcomes remains elusive. Although there were some interesting findings of note our results generated a number of areas for future study. Why was self-esteem positively related to peer substance use, for example, and why only significantly so in urban areas? Future research should include a more in depth exploration of the meaning behind sports participation. We would speculate that if an adolescent is participating in sport solely for the social motivations, and is not overly invested in the outcomes, then the sport as prevention deterrent may be a non-factor for such an individual. Another question relates to the role of team sports compared to individual sports, namely, does one type have more or less of a substance use deterrent property? Additionally, what are the community characteristics (e.g., changes in industrial structure, poverty) that impact the role of sport and corresponding outcomes, especially in rural areas, given the lack of focus upon them? And finally, how does culture (e.g., racial/edinic culture) play a role in how sport and its participation are viewed.
NIDA grant ROl DA09349, Adolescent Drug Use in Rural America. Principal investigator: Ruth W Edwards, Ph.D., Tri-Ethnic Center for Prevention Research, Colorado State University.
Arnold, P. J. (1984). Sport, moral education and the development of character. Journal of Philosophy of Education, 18, 275-28 1 .
Barnes, G M., Hoffman, J. H., Weite, J. W., Farrell, M. P., & Dintcheff, B. A. (2006). Effects of parental monitoring and peer deviance on substance use and delinquency. Journal of Marriage and Family, 68, 1084- 1 104.
Beauvais, F. (1992). Indian adolescent drug and alcohol use: Recent patterns and consequences. [Special issue]. American Indian and Alaska Native Mental Health Research, 5, v-78.
Boutilier, M. A., & SanGiovanni, L. (1983). The sporting woman. Champaign, IL: Human Kinetics.
Blood, K. J. (1990). Non-medical substance use among athletes at a small liberal arts college. Athletic Training, 25, 335-338.
Brown, B. A., Frankel, B. G, & Fennell, M. P. (1 989). Hugs or shrugs: Parental and peer influence on continuity of involvement in sport by female adolescents. Sex Roles, 20, 397412.
Buhrmann, H. G, & Bratton, R. D. (1977). Athletic participation and status of Alberta high school girls. International Review of Sport Psychology, 12, 57-67.
Bush, P. J., & Iannotti, R. J. (1992). Elementary school children's use of alcohol, cigarettes, and marijuana and classmates' attribution of socialization. Drug and Alcohol Dependence, 30,21 '5-287.
Butcher, J. E. (1989). Adolescent girls' sex role development: Relationship with sports participation, self-esteem, and age at menarche. Sex Roles, 20, 575-593.
Camp, W. (1913). What are athletics good for? Outing, 63,259-272.
Centers for Disease Control and Prevention. (2006, June 9). Youth risk behavior surveillance, United States, 2005: Surveillance Summaries. Retrieved September 10,2008 from http://www.cdc.gov/mmwr/PDF/SS/SS5505.pdf
Chabrol, H., Chauchard, E., Mabila, J. D., Mantoulan, R., Adele, ?., & Rousseau, A. (2006). Contributions of social influences and expectations of use to cannabis use in highschool students. Addictive Behaviors, 31, 21 16-21 19.
Chavez, E. L., Deffenbacher, J. L., & Wayman, J. C. (1996). A longitudinal study of drug involvement in Mexican American and White Non-Hispanic high school dropouts, academically at risk students, and control students. Free Inquiry for Creative Sociology, 24,185-193.
Chavez, E. L., & Swaim, R. C. (1992). An epidemiological comparison of Mexican-American and white non-Hispanic 8th- and 12th-grade students' substance use. American Journal of Public Health, 82, 445-447.
Chiarella, M. C. (2003). Family and peer influences on Mexican American adolescent alcohol use: Moderating effects of school adjustment. Dissertation Abstracts International, 63 (9-B), 4362.
Chung, H., & Elias, M. (1996). Patterns of adolescent involvement in problem behaviors: Relationship to self-efficacy, social competence, and life events. American Journal of Community Psychology, 24, 771-784.
Clark, C. M. (1992). Deviant adolescent subcultures: Assessment strategies and clinical interventions. Adolescence, 27, 283-293.
Collingwood, T. R., Reynolds, R., Kohl, H. W., Smith, W, & Sloan, S. (1991). Physical fitness effects on substance abuse risk factors and use patterns. Journal of Drug Education, 21, 73-74.
Conger, R. D., Lorenz, F. O., Elder, G H., Jr., Melby, J. N., Simons, R. L., & Conger, K. J. (1991). A process model of family economic pressure and early adolescent alcohol use. Journal of Early Adolescence, 11, 430-449.
Cronk, C. E., & Sarvela, P. D. (1997). Alcohol, tobacco, and other drug use among rural/small town and urban youth: a secondary analysis of the Monitoring the Future data set. American Journal of Public Health, 87, 760-764.
Dinges, M. M., & Oetting, E. R. (1993). Similarity in drug use patterns between adolescents and their friends. Adolescence, 28, 253-266.
Donnelly, P. (1981). Athletes and juvenile delinquents: A comparative analysis based on a review of literature. Adolescence, 76,415-432.
DuRant, R. H., Middleman, A. B., Faulkner, A. H., Emans, S. J., & Woods, E. R. (1 997). Adolescent anabolic-androgenic steroid use, multiple drug use, and high school sports participation. Pediatric Exercise Science, 9, 150-158.
Eccles, J. S., & Barber, B. L. (1 999). Student council, volunteering, basketball, or marching band: What kind of extracurricular involvement matters? Journal of Adolescent Research, 14,10-43.
Edwards, R. W. (Ed.). (1992). Drug use in rural American communities. New York: Harrington Park Press.
Elliott, D. S., Huizinga, D., & Menard, S. (1989). Multiple problem youth: Delinquency, substance use, and mental health problems. New York: Springer-Verlag.
Ennett, S. T, Bauman, K. E., Hussong, A., Faris, R., Foshee, V. A., Cai, L., et al. (2006). The peer context of adolescent substance use: Findings from social network analysis. Journal of Research on Adolescence, 16, 159-186.
Foon, A. E. (1989). Sport participation among adolescents: Sex differences and effects on academic achievement, self-esteem, affiliation patterns and locus of control. Journal of Applied Research in Coaching and Athletics, 4, 157-175.
Harris-Britt, A. (2004). Students' negative experiences with peers: An assessment of prevalence, negative affect, and coping responses in relation to school adjustment. Dissertation Abstracts International, 64 (11-B), 5784.
Hirschi, T. (1969). Causes ofDelinquency. Berkeley, CA: University of California Press.
Janz,N.K.&Becker,M.H. (1984). The Health Belief Model: A decade later. Health Education Quarterly, 11, 1-47.
Johnston, L. D., O'Malley, P. M., Bachman, J. G, & Schulenberg, J. E. (2005). Monitoring the Future national results on adolescent drug use: Overview of key findings, 2004 (NIH Publication No. 05-5726). Bethesda, MD: National Institute on Drug Abuse.
Johnston, L. D., O'Malley, P. M., Bachman, J. G, & Schulenberg, J. E. (2008). Monitoring the Future national results on adolescent drug use: Overview of key findings, 2007 (NIH Publication No. 08-6418). Bethesda, MD: National Institute on Drug Abuse.
Koss, M. P., & Gaines, J. A. (1993). The prediction of sexual aggression by alcohol use, athletic participation, and fraternity affiliation. Journal of Interpersonal Violence, 8, 94-108.
Labao, L. (1990). Locality and inequality: Farm and industry structure and socioeconomic conditions. Albany, NY: State University Press of New York.
Leukefeld, C. G, Clayton, R. R, & Meyers, J. A. (1992). Rural drug and alcohol treatment. In R. W. Edwards (Ed.), Drug use in rural American communities (pp. 95- 1 1 6). New York: Harrington Park Press.
Lundman, R. J. (1986). Control of delinquency: Can we learn from the past? Contemporary Psychology, 31, 279-281.
Marquait, B. S. (2003). The effect of school transition on rural seventh and ninth grade students' substance use. Dissertation Abstracts International, 63 (9-A), 3152.
Melnick, M. J., & Sabo, D. F. (1992). Education effects of interscholastic athletic participation on African American and Hispanic youth. Adolescence, 27, 295-309.
Miller, K. E., Hoffman, J. H., Barnes, G. M., Farteli, M. P., Sabo, D., & Melnick, M. J. (2003). Jocks, gender, race, and adolescent problems drinking. Journal of Drug Education. 55,445-462.
Miller, A. D., & Ohlin, L. E. (1 985). Delinquency and community: Creating opportunities and controls. Beverly Hills, CA: Sage.
Monsey, B., Owen, G., Zierman, C, Lambert, L., & Hyman, V. (1995). WhatworL· inpreventing rural violence: Strategies, risk factors, and assessment tools. St. Paul, MN: Amherst Wilder Foundation/Publishing Center.
Napoli, M., Marsiglia, F. F., & Kulis, S. (2003). Sense of belonging in school as a protective factor against drug abuse among Native American urban adolescents. Journal of Social Work Practice in the Addictions, 3, 25-41.
Nation, M., & Heflinger, C. A. (2006). Risk factors for serious alcohol and drug use: The role of psychosocial variables in predicting the frequency of substance use among adolescents. American Journal of Drug and Alcohol Abuse, 32, 415-433.
National Federation of State High School Association. (2006). High school athletic participation survey (2005-2006). Retrieved February 1, 2009, from http:// www.nfhs.org/core/contentmanager/uploads/2005_06NFHSparticipationsurvey.pdf
Oetting, E. R., & Beauvais, F. (1987a). Common elements in youth drug abuse: Peer clusters and other psychosocial factors. Journal of Drug Issues, 17, 133-151.
Oetting, E. R., & Beauvais, F. (1987b). Peer cluster theory, socialization characteristics, and adolescent drug use: A path analysis. Journal of Counseling Psychology, 34, 205213.
Oetting, E. R., & Beauvais, F. (1990). Adolescent drug use: Findings of national and local surveys. Journal of Consulting and Clinical Psychology, J<S, 385-394.
Oetting, E. R., Beauvais, F., Edwards, R. W, & Waters, M. (1984). The drug and alcohol assessment system: Book II: Instrument development, reliability, and validity. Ft. Collins, CO: Rocky Mountain Behavioral Sciences Institute.
Oetting, E. R., & Donnermeyer, J. F. (1998). Primary socialization theory: The etiology of drug use and delinquency. Substance Use and Misuse, 33, 995-1026.
Pearson, M., Sweeting, H., West, P., Young, R., Gordon, J., & Turner, K. (2006). Adolescent substance use in different social and peer contexts: A social network analysis. Drugs: Education, Prevention, & Policy, 13, 519-536.
Perry-Burney, G. D., & Takyi, B. K. (2002). Self-esteem, academic achievement and moral development among adolescent girls. Journal of Human Behavior in the Social Environment, 5, 15-28.
Resnick, M. D., Bearman, P. S., Blum, R. W., Bauman, K. E., Harris, K. M., Jones, J., et al., (1 997). Protecting adolescents from harm: Findings for the National Longitudinal Study on Adolescent Health. Journal of the American Medical Association, 278, 823832.
Richman, E. L. (2002). Girls' self construction: Influence of early sport participation on selfesteem in late adolescence. Dissertation Abstracts International, 62, 9-B. (UMI No. 3025382).
Richman, E. L., & Shaffer, D. L. (2000). "If you let me play sports": How might sport participation influence the self-esteem of adolescent females? Psychology of Women Quarterly, 24, 189-199.
Salts, C. J., Lindholm, B. W., Goddard, H. W, & Duncan, S. (1 995). Predictive variables of violent behavior in adolescent males. Youth & Society, 26, 377-399.
Schafer, W (1971). Some sources and consequences of interscholastic athletics: The case for participation and delinquency. International Review of Sport Psychology, /,117137.
Segrave, J. O. (1983). Sport and juvenile delinquency. Exercise and Sports Sciences Reviews, 11, 181-209.
Shanahan, M. J., Elder, G H., Jr., Burchinal, M., & Conger, R. D. (1996). Adolescent paid labor and relationships with parents: Early work-family linkages. Child Development, 67,2183-2200.
Shedler, J., & Block, J. (1990). Adolescent drug use and psychological health: A longitudinal inquiry. American Psychologist, 45, 612-630.
Simons-Morton, B., & Chen, R. S. (2006). Over time relationships between early adolescent and peer substance use. Addictive Behaviors, 31, 1211-1223.
Swaim, R. C, Beauvais, E, Chavez, E.L., & Oetting, E. R. (1997). The effect of school dropout rates on estimates of adolescent substance use among three racial/ethnic groups. American Journal of Public Health, 87, 51-55.
Tiggemann, M. (2001). The impact of girls' life concerns and leisure activities on body dissatisfaction, disordered eating, and self-esteem. Journal of Genetic Psychology, 162, 133-142.
Tracy, A. J., & Erkut, S. (2002). Gender and race patterns in the pathways from sports participation to self-esteem. Sociological Perspectives, 45, 445-467.
Tucker Center for Research on Girls & Women in Sport. (2007). The 2007 Tucker Center research report: Developing physically active girls: An evidence-based multidisciplinary approach. University of Minnesota, Minneapolis, MN: Author.
U.S. Bureau of the Census. (2000). 2000 census of population: General population characteristics. Washington, DC: U.S. Government Printing Office.
Wallace, J. M, & Bachman, J. G (1991). Explaining racial/ethnic differences in adolescent drug use: The impact of background and lifestyle. Social Problems, 38, 333-357.
Weiler, J. (2003). The athletic experiences of ethnically diverse girls. Retrieved February 14, 2009, from http://www.eric-web.tc.columbia.edu/digest.digl31.asp
Matthew J. Taylor and Greg M. Turek
University of Missouri, St. Louis
Address Correspondence to: Matthew J. Taylor, Ph.D., Department of Psychology, University of Missouri-St. Louis, 325 Stadler Hall, One University Blvd., St. Louis, MO63121. Email: firstname.lastname@example.org.