What Happens With Comparison Processes When “the Other” is Very Similar? Academic Self-Concept Formation in Twins

•Empirical test of mirror effects in large sample of monozygotic and dizygotic twins.•Co-twin achievement positively affected twin’s academic self-concept within domains.•Co-twin achievement negatively affected twin’s academic self-concept across domains.•The (mirror) effect tended to be more pronounced for older monozygotic twins. According to the internal/external frame of reference (I/E) model (Marsh, 1986), individuals’ academic self-concept is strongly influenced by comparing their achievement in one domain with their achievement in other domains and with the achievement of others. Research has typically found contrast effects such that high-achieving others have a negative effect on students’ academic self-concept. Yet, what happens if the “other” is somebody very similar to oneself as in the case of monozygotic twins? We postulate and examine the mirror effect, which means that rather than serving as a contrast, the effect of the co-twin’s achievement parallels the effect of a monozygotic twin’s own achievement on academic self-concept. We used data from two school-aged cohorts (11- and 17-year-olds) from a representative sample (N = 4,202) of monozygotic and dizygotic twins in Germany. We regressed twins’ math and German self-concepts on their own and their co-twins’ mathematics and German achievement. Internal and external comparison effects as postulated in the I/E model were replicated for both monozygotic and dizygotic twins across both age groups. In line with our hypothesis, the mirror effect was found in monozygotic twins only: Co-twins’ achievement and twins’ own achievement showed a parallel pattern of positive effects on academic self-concept within each domain and negative effects on academic self-concept between domains, duplicating the I/E pattern. The mirror effect tended to be more pronounced for older monozygotic twins. We argue that the mirror effect is likely caused by high interpersonal similarity and constitutes a rare exception to the broad generalizability of contrast effects as assumed in the I/E model.