Grade retention and academic self-concept: A multilevel analysis on the effects of schools’ retention-composition

• In Event: New perspectives in grade retention effectiveness research: State of the art in Flanders

Abstract

Previous studies on the impact of grade retention on repeaters’ academic self-concept suffer from inconclusive findings (e.g., Lamote et al., 2014; Trautwein et al., 2006; Xia & Kirby, 2009). At the same time, being retained is associated with a significant loss of social connections (Demanet & Van Houtte, 2016). We argue that involving sense of belonging might elucidate the aforementioned ambiguity surrounding the effects of grade retention on repeaters’ academic self-concept. Therefore, our study examines whether grade retention impacts repeaters’ academic self-concept and, if so, whether this relationship is mediated by repeaters’ sense of belonging. Moreover, we aim to contextualize retention research by accounting for schools’ retention composition. Based on reference group theory, the effect of grade retention composition upon repeaters’ academic self-concept is expected to be twofold. Normative reference grouping leads to the assumption that repeaters in high retention composition schools will exhibit lower levels of academic self-concept, because retainees’ values are more likely to spread across all students within the same school. Comparative reference grouping might lead to a moderation effect of retention composition on the relationship between grade retention and repeaters’ academic self-concept, in which non-cognitive effects on grade retention are more pronounced in schools with a low amount of retainees, due to repeaters’ isolated position. We performed multilevel analyses on ISCY-data, consisting of 2,354 students in 30 secondary schools in Ghent (Flanders), using the software program MLwiN. Results indicated that there was a significant negative association between grade retention and academic self-concept ( = -0.80; p < .01). This association, however, disappeared when students’ sense of belonging was added to the model ( = -0.18; p > .05), while the effect of sense of belonging on academic self-concept itself was significant ( = 6.18; p < .001). This indicated that sense of belonging mediates the relationship between grade retention and academic self-concept. Being retained leads to a lower sense of belonging, which in turn leads to a lower academic self-concept. Furthermore, students in high retention composition schools had a significantly lower academic self-concept ( = -0.07; p < .001), which indicated that high rates of retention might have negative spillover effects on students that have never been retained themselves. Finally, we found no moderation effect of grade retention school composition upon the relationship between grade retention and academic self-concept ( = 0.00; p > .05). Therefore, this relationship does not seem to be affected by the number of retainees within a given context. Overall, we argue that educational policy makers should be cautious when applying grade retention on a large scale. The findings of this study do not warrant the high economical and personal costs of being retained. Furthermore, we encourage stakeholders to create a supportive and welcoming school environment (O’Keeffe, 2013). The results of this study show that stimulating the sense of belonging of retainees might be a way to prevent a decrease in academic self-concept.

Authors

• Timo Van Canegem, Ghent University

• Jannick Demanet, Ghent University

• Mieke Van Houtte, Ghent University