Social and Dimensional Comparison Processes in the Formation of Ability Self-Concepts in Elementary School

• In Event: Cross-Domain Effects Across Math- and Language Arts-Related Expectancy-Value Beliefs and Their Implications: A Developmental Perspective

Abstract

Purpose/Theoretical Framework
This study tests the reciprocal internal/external frame-of-reference model (RI/EM; Möller, Retelsdorf, Köller, & Marsh, 2011) from a developmental perspective, focusing in particular on elementary school children. According to the RI/EM, students judge their ability self-concept (ASC) in a domain (e.g., math) by comparing their own domain-specific performance with the performance of others in this domain (social comparison), with their own prior performance in this domain (temporal comparison), and with their own performance in other non-corresponding domains (e.g., English; dimensional comparison).
The RI/EM has been used to explain the formation of domain-specific ASCs of secondary school students over time (e.g., Möller, Zimmermann, & Köller, 2014). However, according to developmental theories of ability conceptions (Dweck, 2002; Harter, 2012), the use of social, temporal, and dimensional comparisons to evaluate own abilities may differ depending on students’ developmental stage. For instance, dimensional comparisons require the ability to integrate opposing attributes in one’s self-representations. Whereas 8- to 10-year-olds increasingly acknowledge that their abilities across domains can differ (“I am good at English and bad at Math”), younger children may show an all-or-none thinking (“I’m good and therefore can’t be bad”; Harter, 2012). Comparative processes described in the RI/EM might therefore change across the elementary school years. Indeed, cross-sectional studies revealed no dimensional comparison effects on ASCs in mathematical and verbal domains prior to Grade 3 (e.g., Ehm, Lindberg, & Hasselhorn, 2014). To our knowledge, no study has investigated dimensional comparisons in elementary school longitudinally. We hypothesize that—in contrast to social and temporal comparisons—dimensional comparisons will not occur before the middle of Grade 3.

Methods
German elementary school students (N=542) were followed from the end of Grade 2 (t1: M = 7.95 years, SD = 0.58) to the end of Grade 4 (t7: M = 9.93 years, SD = 0.72). Self-reported ASCs and teacher-reported grades were assessed in math (t1-t7) and German (t2-t7).

Results and Discussion
Latent cross-lagged models revealed that grades positively predicted changes in ASCs within domains already at the end of Grade 2/beginning of Grade 3. ASCs predicted change in grades in the corresponding domain from the beginning (in German) and the middle of Grade 3 (in math). However, neither significant (negative) cross-lagged contrast effects of grades on ASCs in the non-corresponding domain, nor other substantial longitudinal cross-domain contrast effects were found.
The positive within-domain relations across time suggest that second graders may engage in social and temporal achievement comparisons. By contrast, we found no evidence of longitudinal dimensional comparison effects in elementary school. This might be adaptive because elementary level children need to acquire fundamental knowledge in different domains, and the need to specialize within a given domain may not be as salient at this stage. Upper secondary school students, on the other hand, usually have to choose between different courses and different educational and career paths. In this case, dimensional comparisons can be advantageous because they support specialization (Möller & Marsh, 2013).
Taken together, developmental differences must be considered to better understand the formation of students’ ASCs.

Authors

• Anne Franziska Weidinger, Technical University Dortmund

• Ricarda Steinmayr, Tu Dortmund

• Birgit Spinath, Heidelberg University