A Meta-Analysis of Physical Education Interventions on Students' Physical Activity

Background/Purpose: There have been attempts to promote physical activity (PA) in college students to curb the overreliance on modern technologies that has resulted in sedentary lifestyles. The new shift in lifestyles requires physical education/ health education courses be taught differently in higher education.  Researchers have implemented interventions to improve college students’ PA through physical education/health education courses.  However, there is no information available on the overall intervention effects of such courses. The purpose of the meta-analysis was to examine intervention effects designed to increase PA post-intervention. 

Method: Major databases in physical education, psychology, education, health, and fitness were searched for physical education/health education interventions in PA. Studies were chosen based on reported statistics necessary to compute effect size (ES), were written in English, examined physical activity in college students, incorporated a physical education/health education course intervention and were published from 2000 to 2013. Random-effects analyses were used to synthesize data, and meta-analytic analogues of regression and analysis of variance were utilized to examine potential moderator variables. Moderator variable robustness and publication bias were also tested. 

Analysis/Results: Nine studies satisfied the selection criteria. Some studies had more than one experimental group, and 27 ESs were calculated. Additionally, two types of research design (i.e., pre- and post-test with a control and experimental group, and pre- and post-test without a control group, respectively) were found. The overall interventions had a medium effect on general PA (Hedges g = 0.47, p < .01, n= 27 ESs). The effect size for interventions on MVPA and other PA measures (i.e., such as steps, weekly active days, etc.) was .23 (p < .01, n=13 ESs) and .38 (p < .01, n=14 ESs), respectively, and the difference was significant (p < .01). Contrary to what has been reported in the literature, the interventions on cognitive component only (g =.31, p < .01) yielded a significantly larger ES than those studies that used both cognitive and behavioral interventions (g =.29, p < .01).

Conclusions: Interventions designed to increase college students’ PA were only slightly effective, indicating that more effective interventions are needed if larger ESs are desired. Interestingly, cognitive interventions only generated a larger ES on changing PA levels than those that used both cognitive and behavioral interventions, contrary to results involving other adult populations. More studies are needed to confirm the discrepancy of the PA intervention ESs between college students and other adult populations.