Abstract :
[en] In 2015, urged on by the worrying results of international surveys (Lafontaine et al., 2019 a & b; Quittre et al., 2017), the French-speaking Community of Belgium initiated a structural reform of its educational system (Le Pacte pour un enseignement d’Excellence). One of the original features of this reform is that it is the first (Dachet & Baye, 2020 & 2024) in the French-speaking countries to explicitly refer to the Evidence-Based Education paradigm in one of its founding texts (FW-B, 2015).
Among the many aspects of this structural reform, reducing the use of grade retention is undoubtedly the most “evidence-based” one (Allen et al., 2009; Bright, 2011; Goos et al., 2021; Valbuena et al., 2021). In Belgium, however, grade retention is often viewed positively by teachers and, more generally, in the public opinion. Besides, debates remain lively in the scientific community, with some researchers defending the value of grade repetition because of its perceived usefulness to teachers (Draelants, 2018 & 2019), while others point out the negative or null effects highlighted in the scientific literature (Galand et al., 2019).
At the international level, the debate on the effects of grade retention that might have thought to be closed has revived, since educational economists have shown, on the basis of discontinuity regression designs, some positive effects of grade retention in specific contexts (Özek & Mariano, 2023). At the same time, an analysis of the intrinsic quality of quasi-experimental studies on grade repetition has revealed some worrying findings regarding their risk of bias (Dachet, 2024): existence of confounding factors, low baseline equivalence, post-hoc assignment of students, use of over-aligned measuring instruments, high attrition rate, high differential attrition rate, low statistical power…
To date, previous work aimed at measuring the quality of research dealing with the effects of grade retention has not analyzed the quality of studies following alternative causal designs such as discontinuity regression designs. However, these are precisely the study designs that currently find the most positive effects of grade retention (Goos et al., 2021). Moreover, these are also research designs whose validity is widely debated in the scientific community (Cook, 2002; Huntington-Klein, 2021; Lee, 2016).
In this contribution, we propose an analysis of the quality of all available causal research on the effects of grade retention. We aim to provide an answer to the following question: based on the analysis of the quality of research on repetition, what can be concluded about its effectiveness?