Three sets of analysis were conducted. First, we conducted a multilevel CFA using Mplus version 8 (Muthén and Muthén 2018) to analyze if each of the three indicators at the within-person level (i.e., weekly self-regulation including plan, monitor, control and reflect, weekly academic performance, and weekly psychological well-being) is a distinct construct. Results of confirmatory factor analyses with all three within-person level variables as separate constructs showed relatively acceptable fit indices (χ2(406) = 2915.460; CFI = .85; TLI = .83; RMSEA = .07), indicating that the constructs are sufficiently distinct from one another. Moreover, this model was significantly better than the model collapsing plan, monitor, control and reflect into one factor (χ2(465) = 3337.282; CFI = .73; TLI = .71; RMSEA = .09; Δχ2 (59) = 421.822, p < .001), which support our focal variables can be differentiated from each other.

In addition, to justify our multi-level analysis, we examined the between-person and within-person variance components of the week-level constructs by calculating the intraclass correlation coefficient (ICC). The between-person variance of self-regulation, academic performance, and psychological well-being were 31.1%, 39.0%, and 59.5% respectively. We conclude that our variables varied both within and between persons, which warrants an examination of predictor variables at the person and week level.

Finally, to test our hypotheses, we used the MLwiN program (version 2.35) (Rasbash et al. 2000) to conduct a multilevel regression. To avoid multicollinearity and spurious regression, all week-level variables were centered on the person-mean. We started with a null model that included the intercept as the only predictor (see Table Table2,2, Model 1) and then entered the main effects (Model 2). We examined fixed effects of slopes and tested the improvement of each model over the previous one by computing the differences of their log-likelihood statistic −2*log and subjected this difference to a χ2 significance-test. For testing mediating effects (H2), we conducted a bias-corrected bootstrapping analysis by using INDIRECT syntax in SPSS (version 25). Finally, we estimated the moderated mediation relationship with the bootstrapping technique by using the PROCESS syntax in SPSS (Hayes 2017; Rockwood and Hayes 2017).

Conceptual model

Multi-level regression of goal-oriented self-regulation on psychological well-being

* p < .05; ** p < .01; *** p < .001; N = 74 participants and N = 296 data points

Note: The content above has been extracted from a research article, so it may not display correctly.

Please log in to submit your questions online.
Your question will be posted on the Bio-101 website. We will send your questions to the authors of this protocol and Bio-protocol community members who are experienced with this method. you will be informed using the email address associated with your Bio-protocol account.

We use cookies on this site to enhance your user experience. By using our website, you are agreeing to allow the storage of cookies on your computer.