The relations among academic motivation, self-concept, aspirations and choices: Integrating expectancy-value and academic self-concept theory

The fact that many talented and capable students opt out of the STEM (science, technology, engineering, and mathematics) pipeline and that women remain underrepresented in STEM fields are international phenomena and a matter of considerable concern amongst policymakers. Expectancy-value theory (EVT) (Eccles, 2009) is one of the major frameworks for studying achievement motivation, and has been widely used to tackle this issue. Previous EVT research has demonstrated that students’ expectancy and value beliefs for specific academic subjects are important precursors of achievement-related behaviours (Eccles, 2009; Wigfield, Tonks, & Klauda, 2009). Despite the fact that research on task values has increased, it still lags far behind research on expectancy-related beliefs (e.g., academic self-concept [ASC]) (Wigfield et al., 2009), which is known to represent an important determinant of diverse educational outcomes (Marsh, 2007). This thesis integrated EVT and ASC and extended prior work by closely investigating: (a) the unique contributions of ASC and multiple value components in predicting diverse achievement-related outcomes, particularly during post-high school transition; (b) the multiplicative relation between ASC and value beliefs (i.e., ASC-by-value interaction) that was a core assumption of the original EVT but seems to have disappeared from modern EVT (Nagengast et al., 2011); (c) how well the theoretical models posited in ASC theory (e.g., multidimensional and domain specific selfconcept model, reciprocal effect model [REM], internal/external frame-of-reference [I/E] model with its extension to dimensional comparison theory [DCT]) generalise to different value beliefs; and (d) how the social and cultural factors (e.g., gendered socialisation, socioeconomic status [SES]) shape individual and gender differences in educational and career pathways. This thesis explored new perspectives on EVT and ASC theory through five empirical studies relying on advanced methodologies and using data from large and representative national/international samples. Studies 1 and 4 respectively drew on Grade 8 students from Hong Kong (N = 13,621) and four OECD countries (N = 18,047), based on the International Trends in International Mathematics and Science Study. Studies 2 and 3 respectively drew on longitudinal data from representative samples of U.S. (N = 2,213) and Australian (N = 10,370) students during post-secondary school transition. Finally, study 5 was based on a sample of German 9th-grade students (N = 1,978)