International audienceThe present study examined the hypothesis that parental self-efficacy and parental involvement in children’s education mediate the link between family socioeconomic status and children’s academic achievement. We used a multidimensional approach and a structural equation model combining all the study variables to test our hypothesis on both mothers and fathers. Our research examined 203 parents and their children attending the first or second year (CE1 or CE2) of elementary school in seven different schools. This is the first study of its type to be carried out on a sample of French families. Results for the mothers support our main hypothesis of a mediating effect of parental self-efficacy and parental involvement on children’s academic achievement. However, our hypothesis was not supported in the case of the fathers and we noted several differences between mothers and fathers

Résumé Cette recherche a porté sur 179 parents et leurs enfants scolarisés en CE1 (deuxième année de l’enseignement élémentaire en France) dans 8 écoles différentes. Elle examine les liens qui existent entre le sentiment de compétence parentale, la participation parentale au travail scolaire et les performances scolaires des enfants. Un modèle explicatif des performances scolaires a été construit mettant en relation les différentes variables de l’étude. La conformité des données au modèle théorique a été testée par des analyses structurelles effectuées à l’aide des logiciels LISREL (version 8.50). Les résultats montrent l’adéquation des données au modèle, qui explique environ 39 % de la variance des performances scolaires. Des liens significatifs entre les différentes variables du modèle sont mis en évidence, notamment entre le sentiment de compétence parentale, la participation parentale au travail scolaire et les performances scolaires de l’enfant.

No abstract

International audienceTo what extent can teacher–student dyadic interactions modify the hierarchy of student performances within a single class? To answer this insufficiently researched question, the authors conducted two parallelstudies involving 33 Grade 5 classes in France (759 students) and 15 Grade 5 classes in Luxembourg (243 students). Interactions were observed during whole-class lessons. Posttest scores were analyzed using multilevel models controlling for five level-1 variables and two level-2 variables. The authors did not find any effect of dyadic interactions on relative student performance in mathematics or in language (French or German), in France or in Luxembourg. This result is interpreted in terms of both the public character of dyadic interactions in whole-class settings and the class management functions of these interactions

No abstract