Grade retention research by means of marginal structural models

Grade retention has been a major issue in the field of educational effectiveness research. However, the field has not yet succeeded to find consistent, strong empirical evidence supporting or discouraging the practice. Given the cost and prevalence of the practice, it is an important societal and economic challenge to understand its causal effects on short and long-term outcomes. The current state of the art underlines the need for a comprehensive model and accurate methodology to assess the causal effects of grade retention over time to overcome limitations in previous research. This study addresses this need and contributes to the field of educational research by introducing the method of marginal structural models (MSMs) in the context of grade retention. MSMs are new to the field of educational research and offer several advantages compared to standard regression methods; MSMs allow valid adjustment for time-varying variables that confound the relationship between a (time-varying) treatment and the outcome. In other words, MSMs make it possible to make a fair comparison between children who are retained in different grades and promoted children. Using MSMs, this study investigates the effects of grade retention with regard to short and long term math achievement and wellbeing.status: publishe