The Big-Fish–Little-Pond-Effect revisited: Do different types of assessments matter?

Abstract

In large-scale assessments such as the PISA study, there is a current shift from traditional paper-and-pencil towards computer-based assessments of students' literacy in reading, math, and science. Thus, research has focused on examining the comparability of test scores and motivational factors across the different assessment types. However, it is currently unclear whether the negative effect of school-average ability on students' individual self-concept (i.e., the Big-Fish–Little-Pond Effect, BFLPE) is also robust against the shift in assessments. In the present investigation, we consequently evaluate the relations between math self-concept, school-average ability, and the type of assessment. Based on the Norwegian PISA 2012 data set (N = 4686), we employed multilevel structural equation modeling and found that (a) the BFLPE existed for both assessments with a slightly higher effect size for the computer-based assessment; (b) math anxiety moderated the BFLPE, whereas self-efficacy, the availability and use of information and communication technology (ICT) did not. Our results support the robustness of the BFLPE across the different assessments of mathematical literacy and point out that computer-based assessments involve similar social comparisons as traditional paper-and-pencil tests.