?
Teacher expectations and mathematics growth in primary school: a multilevel moderation analysis
This study examines the association between teacher expectations and mathematics growth among first-grade students in Russia. The sample included 1,155 students and 48 teachers from public schools in two Siberian regions. Students’ mathematics performance was assessed using START, with vertical test alignment used to measure individual growth. Multilevel regression models estimated on 20 multiply imputed datasets were used to examine whether initial mathematics performance moderates the association between teacher expectations and growth. The results indicate that teacher expectations are significantly and positively associated with students’ mathematics growth after controlling for initial mathematics and reading performance, parental perceptions, gender, cultural capital, maternal education, teacher experience, and class size. However, the association did not vary significantly as a function of students’ initial performance level, providing no support for the susceptibility hypothesis and instead being more consistent with a universal effect. A simulation-based power analysis confirmed the design had adequate power to detect a moderate interaction had one existed. Given the correlational design, these finding are discussed as associations rather than causal effects, with implications for future innervation research on fostering teacher expectations in early primary education.