Grade retention and psychoeducational resources: evidence from PISA 2022 using propensity score matching

Abstract

Abstract Background Grade retention is often justified as a compensatory measure that gives low-achieving students additional time to master school content. However, research on grade retention has focused mainly on academic achievement whereas studies of non-academic outcomes have generally examined individual constructs separately. This study examined adjusted cross-sectional associations between grade retention and seven psychoeducational resources among Spanish students participating in PISA 2022. Methods We used data from the Spanish PISA 2022 sample, including 29,775 students aged 15–16 years. Of these students, 5642 were classified as retained (18.95%) and 24,133 as non-retained (81.05%). Retained students were compared with non-retained students who were similar in observed background characteristics. The seven psychoeducational resources were sense of belonging, perceived quality of student-teacher relationships, self-directed learning self-efficacy, perseverance, curiosity, emotional control, and stress resistance. To reduce imbalance in observed background characteristics, we applied propensity score matching across five imputed datasets and then estimated survey-weighted regression models that accounted for the complex sampling structure of PISA. Results Retained students reported lower levels of six of the seven psychoeducational resources analysed than their matched non-retained peers. Grade retention was negatively associated with sense of belonging, student-teacher relationships, self-directed learning self-efficacy, perseverance, curiosity, and emotional control. By contrast, the association between grade retention and stress resistance was not statistically significant. Conclusions These findings suggest that retained students reported lower levels of several psychoeducational resources than observationally similar non-retained students. Although the results should be interpreted as adjusted cross-sectional associations rather than causal effects, they identify relational, learning-related, and emotional domains that may warrant closer attention in future research on the support provided to retained students.

The authors' abstract, as published at the source. Large-scale Assessments in Education, 2026 · DOI ↗