Search

In the present study, we investigated how students’ expectancies and values can be predicted by their achievements in multiple domains. Our major aim was to extend previous findings on dimensional comparison processes for expectancies to task values while systematically comparing multiple value facets defined in expectancy-value theory. We assessed the expectancies, values, and achievements of N = 857 students in Grades 5–12 from two German academic track schools in five academic domains. The results for students’ expectancies largely supported the predictions that were derived from dimensional comparison theory: We found strong evidence for negative cross-domain paths between achievements and expectancies in “far” domains such as math and languages, indicating contrast effects. There were also some positive cross-domain paths between achievements and expectancies in “near” domains such as math and physics, indicating assimilation effects. We also found similar patterns of cross-domain paths for students’ values. However, the results varied substantially across the nine value facets under investigation. We found the strongest evidence for dimensional comparison processes for the value facets most closely related to expectancy (e.g., intrinsic value and cost facets), whereas we found only a little evidence for dimensional comparison processes for the facets of utility value.

According to the Dimensional Comparison Theory (DCT), individuals’ self-concepts of abilities are influenced not only by external sources of evaluation within the same domain, but also internal comparisons of abilities across different domains, resulting in negative contrast effects for self-concepts in dissimilar domains and positive assimilation effects for self-concepts in similar domains. These dimensional comparisons have been primarily tested with academic domains to date, yet social-emotional learning is an important complement to academic learning. The present study sought to extend the DCT to Emotional Intelligence (EI) self-concepts in a sample of 1,069 Canadian children and adolescents, aged 9-18 years. Using parent appraisals as an external source of evaluation, and the Bar-On Emotional Quotient Inventory: Youth Version as a measure of EI self-concept, four EI domains were tested for dimensional comparisons: Intrapersonal, Interpersonal, Stress Management, and Adaptability. Age and gender were tested as moderators. Structural equation modelling revealed contrast effects between Interpersonal and Adaptability domains, and assimilation effects between Intrapersonal, Interpersonal, and Stress Management domains. Effects were only significant in the younger age groups (9-12 and 13-15 years), suggesting a weaker overall effect of dimensional comparisons on EI self-concepts of older adolescents (16-18 years). Assimilation effects were more prevalent in boys, and their EI self-concepts were less differentiated than those of girls, particularly at the younger ages. The results generally support the application of the DCT to EI domains, raising important considerations for educational programs and opening up new possibilities for future research in the field.

Introduction. With regard to the interal/external frame of reference model of academic self-concept development the present study aimed at testing its theoretical assumptions within the verbal domain – namely to analyze the relations between achievement and self-concept in the native language German (L1) and the foreign language English (L2). Due to inconsistent re-search findings in this field the focus of analyses was on identifying differential patterns of dimensional comparison processes as they might depend on students’ achievement level in both subject areas and on students’ gender.Method. Data were gathered in a sample of 256 ninth grade students from German innercity grammar schools. For measuring students’ self-concepts two subject-specific scales were used. For measuring their achievement in each language subject latest grades as well as teacher competency ratings were used. Relations among constructs were analyzed by means of a structural equation modeling approach including multiple group comparisons between the female and male subsample.Results. No dimensional comparison effects could be demonstrated. Neither negative contrasting nor positive assimilation effects across both language subjects occured. Furthermore, multiple group comparisons of structural equation analyses lent strong support for the invariance model. Thus, despite significantly apparent achievement and self-concept differences between female and male learners in the L1 German gender did not moderate the relations between achievement and self-concept variables across both language subjects.Discussion. In the female as well as in the male subsample self-concept formation in both language subjects appears to be mainly affected by social comparison effects. This finding not only confirms previous research results in the field, but also contributes to clarify the role of gender in the internal/external frame of reference model – concerning the issue of dimensional comparison within the verbal domain, at least.

The present study focused on students’ academic enjoyment as predicted by achievement in multiple academic domains. Assumptions were based on Marsh’s internal/external (I/E) frame of reference model and Pekrun’s control-value theory of achievement emotions, and were tested in a sample of 1380 German students from grades 5 to 10. Students’ academic enjoyment, self-concept, and achievement were assessed in relation to mathematics and verbal language classes. In line with assumptions of the I/E model, mathematics performance assessed in the previous academic year positively predicted enjoyment in mathematics classes, and negatively predicted enjoyment in language classes. Language class performance positively predicted enjoyment in language classes, and negatively predicted enjoyment in mathematics classes. Corroborating assumptions derived from Pekrun’s control-value theory, achievement/enjoyment relations were mediated by academic self-concepts. Despite stereotypic gender differences in mean values, linkages between constructs were invariant across genders.