Psychosocial and educational determinants of primary school achievement: grasp of previous research

U ovom preglednom radu iznesen je prikaz relevantnih znanstvenih spoznaja o mogućnostima objašnjenja razine obrazovnog uspjeha učenika osnovnih škola na temelju različitih osobina učenika, obilježja učenikove socijalne okoline te obilježja škole, učitelja i nastavnog procesa. Prikazan je relativni doprinos svakog od razmatranih obilježja u objašnjenju školskog uspjeha kao i ukupan doseg ovih prediktora u zadanom predikcijskom modelu. Na osnovi rezultata brojnih istraživanja pokazalo se da su temeljne osobine učenika, a poglavito njihove kognitivne sposobnosti, najbolji prediktori školskog uspjeha. Znatno manji dio varijabiliteta obrazovnog uspjeha može se pripisati obilježjima obitelji iz koje učenik dolazi i njegove šire socijalne okoline, dok ostatak otpada na osobine učitelja i nastavnog procesa, te osobine škole. Osim navedenih spoznaja, u radu se razmatra upotreba hijerarhijskog linearnog modeliranja kao adekvatnog statističko-metodološkog pristupa u proučavanju ovog fenomena.This review presents the relevant scientific findings on the relation between primary school educational achievement and different student characteristics, elements of their social environment, features of schools, the teaching process and specific teachers’ characteristics. The relative contribution of each of the considered attributes in explaining school success and overall effectiveness of these predictors in a given prediction model are illustrated. Results of numerous studies have shown that the basic students’ characteristics, particularly their cognitive abilities, are the best predictors of school success. A much smaller part in the variability of educational achievement can be attributed to the characteristics of the student’s family and its wider social environment, while the rest can be accounted for by characteristics of teachers, elements of the teaching process and school features. In addition to these insights, the paper discusses the use of Hierarchical Linear Modeling (HML) as an adequate statistical-methodological approach in studying this phenomenon