Author:

Keywords:

self-concept, academic self-concept, self-concept in reading, self-concept in mathematics, multilevel, Democratic Republic of the Congo, Effet de composition, Multiniveaux multivarié, Rendement scolaire, Concept de soi, République démocratique du Congo, Social Sciences, Psychology, Multidisciplinary, Psychology, Composition effect, multivariate multilevel, academic performance, Democratic Republic of Congo, SCHOOL COMPOSITION, MATTER, 1701 Psychology, 1702 Cognitive Sciences, Social Psychology, 5201 Applied and developmental psychology, 5203 Clinical and health psychology, 5205 Social and personality psychology

Abstract:

Research on school effectiveness has often been criticized for its excessive focus on the cognitive aspects of education, thus neglecting non-cognitive aspects that are not less important. This study therefore analyzes the impact of individual and class characteristics on academic self-concept. The latter was measured in two dimensions: self-concept in reading and self-concept in mathematics. The study is based on a sample of 4787 fourth form pupils of elementary school attending 80 classes and 49 schools of two towns in the Democratic Republic of the Congo. The data have been collected by means of a background questionnaire on the pupil, a self-description questionnaire, reading and mathematics tests. They have been analyzed through multilevel multivariate analysis. The results show that the self-concept measured at the beginning of the year is a good predictor of the self-concept measured at the end of the school year. The results show also that the cognitive level of the pupil has a positive impact on academic self-concept. At the class level, only the teacher’s professional experience and the average grade obtained at school scale significantly affect self-concept in reading. The limits and prospects of the study are also discussed in this article. ++++++++ Cette étude examine la possibilité que la composition cognitive et non cognitive de la classe soient corrélées au rendement et au concept de soi et que cette corrélation varie suivant le niveau des connaissances et du concept de soi que les élèves de 4ème année primaire ont au début de l’année scolaire. L’étude est basée sur un échantillon de 4639 élèves répartis dans 80 classes de deux villes de la République démocratique du Congo (RDC). Grâce à l’analyse multiniveaux multivariée, elle montre que l’effet de composition cognitive et l’effet de composition non cognitive de la classe opèrent faiblement dans les écoles de la RDC. La taille de l’effet de composition cognitive est plus importante que celle de l’effet de composition non cognitive. En mathématiques, la composition cognitive de la classe profite plus aux élèves les plus faibles alors qu’en lecture elle avantage uniquement les élèves les plus forts. Par contre, la composition non cognitive profite plus aux élèves ayant au début de l’année une faible estime de soi en lecture. L’étude recommande la valorisation des aspects non cognitifs de l’apprentissage au même titre que les aspects cognitifs et l’application des pratiques enseignantes qui recourent aux interactions entre élèves. L’étude suggère enfin que la recherche approfondisse l’investigation des facteurs responsables de l’effet de composition cognitive et non cognitive de la classe.