Date of Conferral

2020

Degree

Doctor of Psychology (Psy.D.)

School

Psychology

Advisor

David Kriska

Abstract

Teacher effectiveness is a key driver of student achievement but persistently difficult to measure. Although precollege student surveys are one cost-effective alternative to traditional observation measures, little empirical research has been conducted on their factor structure. The purpose of this study was to investigate the validity of the Tripod Survey, a widely adopted precollege survey, as a measure of teacher effectiveness. The dataset was the Measures of Effective Teaching (Year 1, N = 1,024 Grade 9 classroom sections, 20,500 students; Year 2, N = 488 Grade 9 classroom sections, 8,658 students). The dynamic model of educational effectiveness guided the study. Multilevel confirmatory and exploratory factor methods were used to evaluate the factor structure of the Tripod Survey at the student- and classroom-levels to determine the degree of construct isomorphism. None of the hypothesized Tripod specifications adequately fit the data at the classroom level and only marginally at the student level. Several alternative bifactor specifications also did not meet minimum requirements for model fit at the classroom level. Additionally, the negatively worded Tripod items appeared to be challenging for students to interpret. These findings suggest that the 36 items composing the Tripod instrument do not capture effective teaching as hypothesized by the Tripod authors. A reduced-form of the Tripod, a two-factor model with 11 items, fit the data well, at both levels. Thus, the full Tripod Survey did not capture a shared perception of students about effective teaching and should be used cautiously to differentiate teachers at the classroom level. This study contributes to positive social change by providing educational leaders with more robust information about the validity of teacher evaluation tools, which can lead to improved learning opportunities for students.

Download

Share

 
COinS