Moderators of Dimensional Comparison Effects: A Comprehensive Replication Study Putting Prior Findings on Five Moderators to the Test and Going Beyond

• Published in: Journal of educational psychology Vol. 113; no. 3; pp. 621 - 640
• Main Authors: Wolff, Fabian, Zitzmann, Steffen, Möller, Jens
• Format: Journal Article
• Language: English
• Published: American Psychological Association 01.04.2021
• Subjects: Academic Achievement; Comparative Analysis; Foreign Countries; German; Individual Differences; Mathematics Achievement; Nonverbal Ability; Secondary School Students; Secondary School Teachers; Self Concept; Student Attitudes; Teacher Characteristics; Teacher Competencies; Verbal Ability; Germany
• ISSN: 0022-0663
• DOI: 10.1037/edu0000505

Dimensional comparisons, where students compare their achievements in different subjects, have a significant impact on the formation of students' subject-specific self-concepts. This research examines the influence of five moderators that have been shown in previous research to affect the strength of dimensional comparison effects: (1) the intraindividual difference between students' math and verbal achievements, (2) the absolute level of students' math and verbal achievements, (3) students' belief in the negative interdependence of math and verbal abilities, (4) students' perception of the dissimilarity of math and verbal subjects, and (5) students' perceptions of their math and verbal subject teachers' diagnostic competence. We examined the effects of all moderators within a sample of 1,424 students in Germany by conducting latent moderated structural equation modeling within the framework of the internal/external frame of reference model. Only two moderators showed significant influences on the strength of the dimensional comparison effects: The dimensional comparison effects on students' math and German self-concepts were stronger, the more students believed in the negative interdependence of math and verbal abilities, and the more dissimilar they perceived the subjects math and German to be. These moderating effects also persisted if we integrated both moderators into one model simultaneously. Our findings have important implications for self-concept research and dimensional comparison theory, as well as for educational interventions aiming to foster and regulate students' self-concepts. Moreover, they illustrate the importance of replication studies in educational psychology.