School Sports Participation and Academic Achievement in Middle and High School

School Sports Participation and Academic Achievement

Abstract

• Objective: Examine the relationship between school sports participation and academic outcomes, mediated by positive body-weight image and global self-esteem.
• Method: Cross-sectional analysis of 3,186 students (6th-9th grade) from 14 schools in one North Carolina county using the School Success Profile survey. Structural equation modeling was employed.
• Results: School sports participation was significantly associated with academic achievement (standardized effect = 0.225), positive body-image perceptions, and self-esteem.
• Conclusions: School sports may promote outcomes of interest to social workers. Further research is needed to investigate the mechanisms underlying these relationships.
• Keywords: school sports, academic achievement, self-esteem, body image, structural equation modeling

Introduction

• Increasing youth physical activity is a priority in schools to promote positive child development.
• Schools offer accessibility to students, long periods of exposure, and existing resources for physical-activity interventions.
• Data indicates that children's physical activity rates remain low and may be declining.
• Schools may be overly focused on core academic areas due to legislation like No Child Left Behind.
• Promoting children's physical activity may be perceived as detrimental to core academic subjects.
• Research shows that physical activity promotes classroom performance and should be emphasized in schools.
• Evidence is needed to demonstrate the potential for physical activity to enhance grade-based academic achievement.

The Role of School Sports

• Student participation in physical activity within the school setting includes:
  • Physical education classes
  • School sports
  • Free-time activity
• Physical education is often the dominant form of physical activity in schools.
• School sports vary by school level and type, grade level, district resources, and individual student preferences.
• School sports can occur in team (e.g., soccer, basketball) or individual (e.g., tennis, cross country) contexts.
• Participation in school sports is rarely mandated by schools or districts.
• Most evidence on the ability of school-based physical activity to enhance academic achievement comes from studies examining physical education or specialized physical-activity interventions.
• There has been little research on school sports participation.
• School sports may promote physical activity and student academic outcomes due to social contexts, increased involvement with school resources, and identification with school values.
• Nationally representative data indicate that more than half of high school students annually participate in school sports.
• There exists a great need for additional research that extends the robust findings linking general physical activity and academic achievement to school sports.

Sports and Academics: Biological Links

• Academic achievement has long been of interest to social workers concerned with child development.
• Research has consistently found that physical activity can promote children’s academic outcomes.
• One key causal mechanism underpinning the relationship between physical activity and academics is a direct, physiological link between activity and cognition.
• Physical activity activates the allocation of cognitive resources and promotes faster cognitive processing through stimulus encoding.
• Physical activity can promote cell stability and ameliorate the negative effects of stress on the body.
• Sibley and Etnier (2003) conducted a meta-analysis and found a mean effect size of 0.32 (p < .05) for physical activity on cognitive development.
• This enhancement of cognition is directly related to core academic ability and its manifestation in outcomes such as grades and test scores.
• Multiple regression analysis found that increasing students' physical-activity levels from less than 2.5 hours per week to 7.0 hours per week was associated with 5.7%-9.1% increases in students’ grades.
• Rasberry and colleagues (2011) systematically reviewed 127 positive associations between school-based physical activity and academic achievement across 50 unique studies.
• Other observational analyses have consistently found that increases in physical activity are directly associated with improvements in academic-achievement outcomes.

Sports and Academics: Psychosocial Links

• Recent inquiries have advanced understanding of the relationship between physical activity and academic achievement beyond the biological mechanism toward more complex conceptual models that involve potential psychosocial variables in causal pathways.
• Developing such conceptual models identifies factors that are particularly interesting to social workers, moving physical-activity promotion from a health-focused intervention into the realm of a comprehensive, multifaceted means of enhancing children’s development and well-being.
• It also identifies factors that can be specifically targeted in physical-activity interventions, potentially increasing overall efficacy and population- and context-specific efficacy.
• Tomporowski, Lambourne, and Okumura (2011) have outlined a comprehensive theoretical model for physical-activity promotion among children whereby psychosocial factors related to self-based worth, efficacy, and esteem partially mediate the relationship between physical activity and mental functioning outcomes, such as cognition and academic achievement.
• The model also includes physical health factors (i.e., obesity, sleep, fatigue) as further mediators of the relationship between physical activity and academics.
• Two psychosocial factors in particular—global self-esteem and positive body-weight image—deserve further attention for their evidence-based role in mediating physical activity’s effect on academics.
• Global self-esteem describes a core component of adjustment that captures how a person perceives and evaluates her or his self within the context of their environment.
• Recent empirical research now supports this hypothesis, with numerous studies finding that physical activity enhances children’s self-esteem, which in turn may enhance other attitudes and behaviors related to academic achievement.
• A longitudinal observational analysis of 11,957 students in grades 7 to 12 found that increases in self-reported daily physical activity were significantly positively associated with increases in global self-esteem.
• A meta-analysis of randomized controlled trials found small to moderate effects of physical activity on self-esteem for school-aged children, with an effect size of 0.49 for exercise-only studies (n = 8) and 0.51 for comprehensive studies (n = 12) that comprised exercise and other components.
• Results from the well-fitting model showed that students’ self-reported home- and school-based weekly physical activity was associated with self-reported grades both directly ($b = .08$, p < .01) and indirectly via global self-esteem ($b = .02$, p < .01).
• Body image (BI) is an individual’s valuation of their physical body.
• BI varies conceptually but broadly relates to an individual’s perceptions and attitudes related to his or her physical body, typically in terms of weight, fatness, appearance, and overall body size.
• BI can be described as an individual’s internal perceptions of his or her external body.
• BI is increasingly examined as a key predictor of self-esteem, a variety of daily life actions, and long-term developmental outcomes.
• In recent decades, physical-activity researchers have extended theories of self- esteem by positing that physical-activity participation leads directly to body improvements such as decreases in fat, increases in stamina, and others that bolster valuation of an individual’s BI.
• This boost in a person’s opinion of his or her physical image, in turn, fosters general confidence that can pay dividends in other domains, including those related to self-esteem and academic achievement.
• Thus, it is believed that BI may mediate the relationship between physical activity and self-esteem.
• Most research linking physical activity with BI has been conducted on adults, with research on children being relatively scant.
• An observational study of 988 German high school students found that participation in regular physical activity and/or exercise was statistically significantly associated with more favorable self-reported BI.
• A meta-analysis of observational studies (n = 11) at the adolescent and high school level found that when comparing athletes versus nonathletes, the former had an over- all effect size of 20.38 (n = 42) on measures of negative body image.
• There is a continued need for more and newer research on how physical activity might increase positive BI among children.

Current Study

• Goal: Expand research regarding the connections between physical activity and school achievement among middle school and high school students.
• Builds upon the general model linking physical activity with academic achievement outlined by Tomporowski and colleagues (2011), as well as the structural equation modeling analysis by Kristjánsson and colleagues (2010), to add to research in three ways.
  • Combines the self-esteem theories long promulgated in children’s physical- activity promotion with more recent concerns regarding the sociocultural importance of BI, representing a more complete picture of how physical activity may increase academic achievement outcomes via direct increases in global self-esteem and indirect increases in positive BI.
  • Adds to the growing body of research regarding the school sports subtype of school physical activity.
  • Tests the conceptual model using a large sample and advanced statistical methods that control for clustering, missing data, nonnormality, and other important characteristics.
• These study features have potential to inform the development, implementation, and evaluation of school sports interventions to promote school success.
• The study features four explicit research questions (RQs) assessed using structural equation modeling (SEM).
  • RQ1: Do middle and high school students who participate in school-based sports demonstrate direct increases in grades-based achievement?
  • RQ2: Is the relationship between school-based sports and grades-based achievement indirectly influenced by increases in global self-esteem?
  • RQ3: Is the relationship between school-based sports and grades-based achievement indirectly influenced by increases in positive BI perceptions?
  • RQ4: Is the school-based sports model distinguishable from an alternative that models nonsports extracurricular activity as directly and indirectly influencing grades-based achievement via positive BI and global self- esteem?

Method

• Sample
  • Researchers collected a convenience sample in 2008 from 3,186 students in the 6th to 9th grades in 14 schools.
  • The mean student age was 12.8 years, with a range of 9 to 16. However, nearly all (99.0%) of the students were ages 11 to 15.
  • Slightly over half (50.6%) of the sample was male.
  • White students comprised the majority (83.7%) of the sample, with mixed race (5.5%) and African American (3.9%) students comprising the next two largest subpopulations.
  • Taken together, the schools comprised a locally representative convenience sample from a semirural, mountainous county within North Carolina.
  • The majority (87.7%) of students in the sample described their health as “good” or better, with 31.0% reporting “excellent” health.
  • English was the primary language spoken in 95.5% of the students’ homes, and the majority (74.8%) resided in a two-parent family.
  • The sample was relatively high achieving academically, with recent grades of “mostly As and Bs” reported by 58.3% of the students.
  • Accordingly, most students (66.8%) reported that they “definitely” aspire to attend college.
• Measures
  • Students completed the School Success Profile (SSP), a well-validated social- environmental assessment for middle and high school students that has been widely used to examine academic performance and other school-success outcomes .
  • The SSP includes 263 closed-ended items across six modules and these items are typically grouped and analyzed using 30 validated subscales.
  • Broadly, these subscales address stu- dents’ perceptions of their social environment (i.e., neighborhoods, schools, friends, and families), and about their own physical and psychological health and school performance (i.e., individual adaptation).
  • The SSP takes approximately 60 minutes to complete and students are allowed to decline or skip questions without penalty.
  • Academic achievement was measured with a single performance-based item asking students to report their grades on “[their] most recent report card.”
  • The five responses included the standard SSP options of mostly Ds and Fs (1), mostly Cs and Ds (2), mostly Cs (3), mostly Bs and Cs (4), and mostly As and Bs (5).
  • School sports participation (1 = Participation, 0 = No participation) was measured with a single, dichotomous student-reported item.
  • Students were first asked whether they had “ … participated in any school activities during the current school year that are not part of class work, such as sports or school clubs.”
  • If students answered positively, they were then prompted to check as many of the nine options for extra- curricular activities listed on the SSP as were applicable to them.
  • Any affirmative response to the option of “sports or athletics” (inclusive or exclusive of other extra- curricular activities) was coded as school sports participation.
  • The SSP self-esteem subscale (Bowen & Richman, 2008) assessed students’ global self-esteem.
  • All items were adapted from Rosenberg’s widely used scale for measuring self-esteem among adolescents.
  • Each item featured three ordinal responses of Not like me (1), A little like me (2), and A lot like me (3).
  • Higher scores indicate higher levels of global self-esteem.
  • The observed scores for the self-esteem composite variable ranged from 5 to 15 and Cronbach’s alpha for the five self-esteem items in the sample was 0.88.
  • The positive body-weight image SSP subscale (Bowen & Richman, 2008) was measured with three items asking about weight-based body image perceptions: (a) “I weigh too much,” (b) “I wish that I could lose weight,” and (c) “I worry a lot about my weight.”
  • Response options were identical to those for self-esteem and the original, negatively worded response options were recoded so that higher scores indicated more positive perceptions of one’s body weight.
  • The observed scores for the composite body-weight image variable ranged from 3 to 9 and Cronbach’s alpha for the three body-weight image items in the sample was 0.85.
  • Items from the SSP permitted covariates for race (White = 1, Non-White = 0), Hispanic/Latino ethnicity (Yes = 1, No = 0), biological sex (Male = 1, Female = 0), age (9–17), number of parents with whom the student lives (Two parents = 1, Other = 0), and primary home language (English = 1, Other = 0) on each of the latent variables in the model (Bowen & Richman, 2008).
• Data Analysis
  • Three sets of preliminary checks were conducted for normality, clustering, and missingness on the items used in the analysis.
  • Items were largely negatively skewed (90.0%< 0.0; range: -1.50–0.185) and platykurtic (60.0% < 3.0; range: 1.03–4.29) and these findings indicated high levels of nonnormality.
  • Results showed intraclass correlation coefficients signaling statistically significant clustering effects (r ≤ .038) in the data, indicating that a small percentage of the variance of item scores was due to school-level characteristics.
  • Missingness due to item nonresponse was found for 0.3% of cases on academic achievement, 8.7% of school sports, upward of 1.6% on self-esteem, up- ward of 1.4% on body-weight image, and upward of 0.3% on the demographic co- variates.
  • Based on the nature of the data and the specified research questions, latent variable SEM was chosen as the most optimal analytic technique.
  • Model specification: The null hypothesis in an SEM analysis is that the reproduced covariance matrix is not statistically significantly different from the ob- served matrix (Bollen, 1989; Bowen & Guo, 2012).
  • Model estimation: All SEM analyses were conducted in Mplus (Version 7.3 for Macintosh; Muthén & Muthén, 2014).
  • Model evaluation: The $x^2$ statistic assesses the absolute fit of the model but is problematic with large samples, thus, $x^2$ was not used to assess fit here and has been reported only for reference.
  • Instead, model fit was assessed using three other metrics: the comparative fit index (CFI), Tucker-Lewis index (TLI), and the root mean square error of approximation (RMSEA).
  • Fit cutoffs were based on recommended values: (a) CFI ≥ .95, (b) TLI ≥ .95, and (c), RMSEA point estimate and upper bound 90% con- fidence interval acceptable at ≤ .08 and preferred at ≤ .06.

Results

• Descriptive Characteristics
  • A significant proportion (45.4%) of all students reported participating in some form of school sports.
  • The sample demonstrated high academic-achievement scores (M = 4.25, SD = 1.10), with 58.3% of students reporting recent grades of “mostly As and Bs” and 23.8% reporting “mostly Bs and Cs.”
  • Raw academic achievement scores were 6.1% higher for students participating in school sports compared with those who were not.
  • Self-esteem scores were also high (M = 12.69; SD = 2.63), with 41.4% of students reporting perfect scores of 15.
  • Self-esteem scores were 4.8% higher for students participating in school sports compared with nonparticipants.
  • Positive body-weight image scores were somewhat low overall (M = 6.98; SD = 2.03), with slightly over one third (36.0%) of students reporting perfect scores but one quarter (25.3%) scoring in the bottom half of the range (≤ “5”).
• Measurement Model
  • The measurement model based on the polychoric correlation matrix exhibited acceptable fit according to the prespecified criteria: $x^2$ (32, N = 3184) = 325.584, p < .001; CFI = .990; TLI = .986; RMSEA = .054 (90% CI [.048, .059]).
  • All parameter estimates were statistically significant (p < .001).
  • The eight standardized (StdYX) factor loadings for the two latent variables (self- esteem/h2, body-weight image/h3) ranged from .835 to .927.
  • StdYX values for the in- terfactor correlations between latent variables were .261 between academic achieve- ment and self-esteem, .176 between academic achievement and school sports, .479 between self-esteem and body-weight image, .164 between self-esteem and school sports, and .129 between body-weight image and school sports.
  • Squared multiple correlations for the eight item indicators ranged from .698 to .859.
• Structural Model
  • The full structural model also demonstrated acceptable fit: $x^2$ (83, N = 3,186) = 270.831, p < .001; CFI = .990; TLI = .986; RMSEA = .027 (90% CI [.023, .030]).
  • Overall, the model variables explained 18.3% of the variation ($R^2$) in academic achievement, 26.6% of self-esteem, 5.7% of body-weight image, and 4.2% of school sports.
  • StdYX values for the associations between model variables were all positive and statistically significant: .208 from self-esteem to academic achievement, .186 from school sports to academic achievement, .454 from body-weight image to self-esteem, .131 from school sports to self-esteem, and .125 from school sports to body-weight image
  • Sobel’s test also found two significant StdYX indirect effects from school sports to academic achievement: .012 via self-esteem and body-weight image (p = .001), and .027 via self-esteem only (p < .001).
• Alternative Model
  • Results from the extracurricular activities alternative model also demonstrated acceptable fit: $x^2$ (84, N = 3,186) = 267.396, p = .000; CFI = .991; TLI = .987; RMSEA = .026 (90% CI [.023, .030]).
  • The pathways among model variables were all statistically significant (p < .05): .241 from self-esteem to academic achievement, .160 from extracurricular activities to academic achievement, .474 from body- weight image to self-esteem, .070 from extracurricular activities to self-esteem, and - .061 from extracurricular activities to body-weight image.

Discussion

• All three research questions were answered affirmatively: Students who participated in school sports had higher grades, and self-esteem and positive body-weight perceptions significantly influenced this relationship.
• The model demonstrates that school sports participation likely has a partially mediated effect on achievement.

Limitations

• Cross-sectional data limits definitive causal relationship conclusions.
• Lack of temporal hierarchy prohibits true test of mediation.
• Sample is relatively homogenous, limiting generalizability.
• Child self-reports may be unreliable.
• The SSP did not permit measurement of cognitive ability or socioeconomic status.
• The study also relied on a limited, global measure of self-esteem when a more domain- specific measure (i.e., self-esteem for physical activity) would have been ideal.
• There are other potential models with other mediators that could have fit the data equally well.

Implications

• Youth participation in physical activity is arguably a necessity for short- and long-term health.
• This study provides further support for examining how school-based sports activities can be leveraged to promote key outcomes such as grades.
• School sports participation may provide direct and indirect benefits to students’ grades and may be able to marginally increase their students’ grades.
• Increases in students’ positive perceptions of their body weight and increases in global self-esteem warrant further investigation within the growing field of school physical-activity research.
• Comparisons with an alternative model using nonsports activities suggest that school sports may be distinct from the other extracurricular activities that occur in school, and suggest that extracurricular activities may promote smaller increases in achievement scores when compared with sports.

Conclusions

• There is a great need for more longitudinal research and experimental trials.
• Future research should seek to examine how issues related to type, duration, size, and dosage of school-based forms of physical activity impact achievement, as there is likely to be great variation within and between school- based sports programs.
• There is a need to examine whether models of activity and achievement are salient for other, more diverse samples of students and across multiple groups of students.
• Lastly, future inquiries should investigate to what extent, if any, school sports influence students’ microlevel interpersonal relationships, such as their attachment to school members.