ABSTRACT

Much research has evaluated teacher- or school-level variance in students’ achievement as a proxy for teacher or school effects. These effects have almost exclusively been estimated across years with state-level testing data. In this paper, interim assessment data, collected across 9 time points in the fall, winter, and spring of Grades 3 to 5 were used to model students’ within-year reading and mathematics growth. A multilevel model was used to estimate growth rates within each grade while accounting for the summer transition period, and intercept and slope variance was parsed into student, teacher, and school components. Results indicated substantial variance between both teachers and schools in students’ within-year growth, which complicates interpretations about each level without reference to the other.

摘要

许多研究评估了教师或学校水平在学生成绩方面的差异，以作为教师或学校效果的替代指标。这些影响几乎是使用州级测试数据多年估算出来的。在本文中，使用了从3至5年级的秋季，冬季和春季的9个时间点收集的中期评估数据来模拟学生的年内阅读和数学增长情况。在考虑夏季过渡期的同时，使用多层次模型来估算每个年级的增长率，并且将截距和坡度方差解析为学生，老师和学校的组成部分。结果表明，教师和学校在学生年内增长方面存在很大差异，这使得对每个级别的解释变得复杂，而没有相互参考。