Student self-efficacy, behavioral engagement, and emotional engagement are key factors for academic achievement. Research has yet to identify the developmental cascades linking these four constructs. Three theoretical frameworks, i.e., Self-Efficacy Theory, the Self-System Model of Motivational Development, and Expectancy-Value Theory, suggest different nexus. Following 671 students (51.8% girls) from their 4th to 6th grade, this study aims to assess competing hypotheses from these three frameworks in math. Three cross-lag models were tested to test each theoretical framework. A fourth and final model was tested to include the significant paths from the previous models. Mediation paths were also tested. Results mainly support assumptions from Self-Efficacy Theory, that is student self-efficacy and academic achievement are mutually associated from 4th to 6th grades. Some of the propositions of Expectancy-Value Theory were also supported. Self-efficacy was associated with later emotional engagement and academic achievement. However, emotional engagement in 5th grade was negatively associated with achievement in 6th grade and was not associated with behavioral engagement. Assumptions from the Self-System Model were not supported by the data. Testing the fourth model revealed an unexpected developmental cascade: 5th-grade self-efficacy mediated the association between 4th-grade achievement and 6th-grade emotional engagement. This last finding may have great implications for young adolescents as emotional engagement is an indicator of student well-being and intrinsic value of learning. Implications for theory validation and intervention targets for adolescents are discussed.
Keywords: Academic achievement; Expectancy-Value Theory; Self-Efficacy Theory; Self-System Model of Motivational Development; Self-efficacy; Student engagement.