e., n = 549 for those who did not miss the dependent variable), thereby avoiding biased standard errors. 32 There were 346 (63.02%)

participants who met the PA recommendation (i.e., 150 min of the PA per week), and those who met the PA recommendation reported more weekly exercise on the LTEQ than those who did not (mean = 61.47, SD = 78.28 vs. mean = 34.61, SD = 56.74, Protease Inhibitor Library clinical trial respectively; t (549) = 4.52, p < 0.001, Cohen's d = 1.50). In spite of the caveats noted above about the LTEQ in this sample, the magnitude of this finding offers concurrent validity related evidence in support of the binary approach employed in this study (i.e., MPAR vs. does not MPAR). We also explored the bivariate correlation among the able, worth, enabling, and reinforcing factors. The results indicated that for able, the correlation between self-efficacy and perceived competence was 0.38. For worth, the correlation between enjoyment and

attitude was 0.59. For enabling factors, the correlations among accessibility, knowledge, language barrier, and skill and fitness were between −0.14 and 0.72. For the reinforcing factors, the correlation between role modeling and peer support was 0.36 (Table 3). The moderate-to-high internal consistency of each scale and low-to-moderate correlations among different scales supports the convergent and discriminant validities of the scales employed. We attempted to identify the factors that best predicted the odds of MPAR among Chinese international students. Tables 4 and 5 show the odds ratio of the logistic nested regression comparing the five nested models. The model Ruxolitinib clinical trial comparison results indicate that adding able factors (Model 2) significantly increased the odds of MPAR prediction, compared to the base model (Model 1). Adding worth factors (Model 3) significantly increased the prediction of the odds, compared to Model 2. Adding enabling (Model 4) and reinforcing factors (Model 5) did not significantly increase the model prediction, compared to Model 3. Therefore,

Model 3 was the final model for MPAR. In Model 3, sex significantly influenced the odds of MPAR. The odds of males meeting the PA recommendation was 1.49 times greater than the odds of females Org 27569 meeting them (p < 0.001). Being one SD higher on BMI increased the odds of MPAR by 1.25 times (p < 0.05). Being one standard deviation higher on competence and efficacy increased the odds of MPAR by 1.95 and 1.68 times, respectively (both p < 0.001). Although the direct effects of the enabling and reinforcing factors on PA lacked statistical significance, the indirect effects of the enabling and reinforcing factors on MPAR through able and worth may still exist. We used the user written command “binary_mediation” in STATA to examine each mediation effect. The results showed that there were no direct effects of the enabling and reinforcing factors on MPAR (all p > 0.