Teachers and schools both play important roles in students' education. Yet, the unique contribution of each to students' growth has rarely been explored. In this dissertation, a Bayesian multilevel model was applied in each of Grades 3 to 5, with students' growth estimated across three seasonal (fall, winter, spring) administrations of a mathematics assessment. Variance in students' within-year growth was then partitioned into student-, classroom-, and school-level components. The expected differences in students' growth between classrooms and schools were treated as indicators of the teacher or school "effect" on students' mathematics growth. Results provided evidence that meaningful differences in students' growth lies both between classrooms within schools, and between schools.
The distribution of teacher effects between schools was also examined through the lens of access and equity with systematic sorting of teachers to schools leading to disproportional student access to classrooms where the average growth was above the norm. Further, previous research has documented persistent and compounding teacher effects over time. Systematic teacher sorting results in students' having differential probabilities of being enrolled in multiple "high" or "low" growth classrooms in a row. While clear evidence of teacher sorting was found, the demographic composition of schools did not relate to the sorting, contrary to previous research. The persistence of teacher and school effects was also examined from a previously unexplored angle by examining the effect of students' previous teacher(s) on their subsequent rate of within-year growth during the school year. These effects were found to be small and teacher effects overall were found to decay quite rapidly.
Identifer | oai:union.ndltd.org:uoregon.edu/oai:scholarsbank.uoregon.edu:1794/19191 |
Date | 18 August 2015 |
Creators | Anderson, Daniel |
Contributors | Stevens, Joseph |
Publisher | University of Oregon |
Source Sets | University of Oregon |
Language | en_US |
Detected Language | English |
Type | Electronic Thesis or Dissertation |
Rights | Creative Commons BY-NC-ND 4.0-US |
Page generated in 0.0014 seconds