Initial Concept Planning
This study used data from the TIMSS-R 1999. The methodology used in the TIMSS-R (1999) study is presented in this section. TIMSS participants and procedures are described. Measures were developed of autonomy support, intrinsic motivation, external regulation, introjected regulation, and math self-concept.
Drafts & Revisions
The model-fitting process suggested that Model 2 best fit the data (χ2 (13) = 92.170, CFI=0.953, TLI=0.951, RMSEA=0.049, SRMR=0.044). The model with the lowest AIC is preferred (Kline, 1998). Model 2 had the lowest AIC. On this basis, we selected Model 2 as the best model of all the models and based further analyses on that model. Model 2 is called the trimmed full model.
The path diagram that shows the fully-unstandardized parameter estimates for Model 2 appears in Figure 5. These findings support the model. First, intrinsic motivation positively affected math performance (β=7.865), whereas external regulation negatively affected math performance (β= -15.196), consistent with the first hypothesis (H1). Math self-concept positively affected math performance (β=11.399), consistent with the second hypothesis (H2). Autonomy support positively affected math self-concept (β=0.072) and math performance (β=1.474), consistent with the third hypothesis (H3). Math self-concept significantly affected math performance through the mediator of intrinsic motivation, consistent with the fourth hypothesis (H4). Finally, autonomy support in the classroom significantly affected math performance through the mediator of math self-concept, consistent with the fifth hypothesis (H5).