Abstract:
Economists and educational researchers discovered that noncognitive factors such as in-terests, self-concepts, and conscientiousness are both interesting and relevant because of the evidence that these types of constructs can be used to predict important human-capital out-comes (e.g., school functioning; see Kautz, Heckman, Diris, ter Weel, & Borghans, 2014). Given the importance of these factors, the discussion quickly turned to how these factors can be enhanced and fostered. It was recently claimed that a construct’s malleability is one pre-condition for changing or enhancing it (e.g., through an intervention; Bailey, Duncan, Odgers, & Yu, 2017). However, the group of noncognitive factors reflects an inclusive conglomerate of variables, including personality traits and motivational resources. The inclusive grouping of concepts belies the often stark theoretical and conceptual distinctions that divide these constructs when they are used in research. For instance, the Big Five personality traits origi-nate from trait theories, whereas constructs such as self-concepts and interests come from theoretical models that take a social cognitive perspective. Whereas traits are often defined as stable and cross-situationally consistent (McCrae & Costa, 2008b), social cognitive con-structs are conceptualized as less stable, relevant to very specific contexts, and derived almost exclusively from experience (Bandura, 2001b; Eccles & Wigfield, 2002). In addition, be-cause of their different origins, it is uncommon to find both types of constructs included in the same study (Roberts, 2009). This has created an asymmetry in the understanding of how changeable and malleable (e.g., stable or context-sensitive) the two classes of constructs are and has also impeded the ability to explore synergies when considering them together.
The current dissertation has two superordinate aims. The first aim was to test two basic assumptions about the malleability (i.e., time-consistency and context-sensitivity) of traits and social cognitive constructs. For this purpose, the Big Five personality traits (consci-entiousness, neuroticism, extraversion, openness, and agreeableness) were contrasted against several social cognitive variables (i.e., interest, self-concept, and effort) with regard to their temporal stability (Study 1) and their context-sensitivity (Study 2). The second aim of the dissertation was to examine potential synergies when considering variables from both per-spectives (conscientiousness as a trait and interest as a social cognitive construct) in predicting an achievement-related outcome, namely academic effort (Study 3).
This dissertation benefited from a large-scale longitudinal study called “Tradition and Innovation in Educational Systems” (TRAIN; Jonkmann, Rose, & Trautwein, 2013), which is hosted at the Hector Research Institute of Education Sciences and Psychology at the Univer-sity of Tübingen, Germany. TRAIN is a school achievement study that encompasses four time points (from Grades 5 to 8). The study comprises N = 3,876 students in 136 classes in 99 schools from two federal states (Baden-Württemberg and Saxony) of Germany. Among other variables, the TRAIN study includes the Big Five personality traits and a variety of so-cial cognitive variables (e.g., interest, self-concept), which provided the opportunity to test assumptions about the malleability of the different classes of constructs in the same study and moreover, it meant that potential synergies between the classes of constructs could be ex-plored in a longitudinal setting.
In the first study (Social Cognitive Constructs are Just as Stable as the Big Five Be-tween Grades 5 and 8), several social cognitive variables (i.e., self-concept, interest, and aca-demic effort) and personality traits (i.e., the Big Five personality traits) were contrasted with respect to their temporal continuity and change over time. Three research questions were of particular interest. First, how stable are the constructs over time? Second, to what degree can the stability of each class of constructs be attributed to unchanging components, and how much of the instability can be attributed to state components? Third, to get a better idea of whether individuals change more or less on each class of variables, the average level of indi-vidual differences in change (i.e., to what extent students show increases or decreases in the constructs) were compared between the classes of constructs. The results indicated that there were no marked differences between personality traits and social cognitive constructs across multiple indicators of stability or changeability.
The second study (The Effects of Getting a New Teacher on the Consistency of Per-sonality) focused on examining the effect of getting a new teacher on consistency in students’ psychological functioning (i.e., personality traits and social cognitive variables) in two longi-tudinal studies (TRAIN and PISA-E; for the description of the PISA-E study, see Study 2 in Chapter 4). By using quasi-experimental designs, two indices of consistency (i.e., test-retest correlations and changes in variance components over time) were compared between students who got a new teacher and those who did not on a variety of social cognitive and personality constructs. The results showed no differences in the test-retest correlations for the math-related social cognitive constructs of interest, effort, self-concept, anxiety, and the Big Five personality traits (except extraversion). Significantly lower test-retest correlations were found for some of the German- and English-related social cognitive constructs and for self-regulation in math for the group of students who got a new teacher. Finally, regarding the changes in variance (over time), there were found no systematic differences between groups in the TRAIN and PISA-E study.
The third study (The Development of Students’ Academic Effort: The Unique and Combined Effects of Conscientiousness and Individual Interest) tested the unique and com-bined effects of conscientiousness and individual interest (as representatives of each perspec-tive) on the development of academic effort in the school subjects of math, German, and English in the TRAIN study. Three research questions were of particular interest. First, the development of academic effort across three school subjects was examined. In a second step, the unique effects of conscientiousness and individual interest in predicting changes in aca-demic effort over time were tested. Third, to test the interactive relation between conscien-tiousness and individual interest in a longitudinal setting, the proposed interaction between conscientiousness and individual interest was included in the analyses. Results showed that academic effort significantly decreased over time across the three school subjects. However, both conscientiousness and individual interest significantly and positively predicted change in academic effort such that when conscientiousness and individual interest were higher, the decrease in academic effort was smaller. In addition, conscientiousness and individual inter-est interacted in a compensatory manner such that individual interest was less important for the academic effort of students who were high on conscientiousness. The results held for most prospective paths across three waves of longitudinal data.
The findings of the three studies are summarized and discussed in light of a broader research context. Implications for practice and future research are derived.