Date of Award
Doctor of Philosophy
Dr. Zoe A. Barley
Dr. Jianping Shen
Dr. Mary Anne Bunda
This study explored the extent to which confirmatory factor analysis (CFA) can be used to address the measurement challenges faced by evaluators engaged in the assessment of change; in particular, the interpretation of self-report survey data collected under quasi-experimental conditions. The psychometric principles behind the instruments used to measure change are built on the assumptions that the constructs of interest remain stable and that error and score magnitude alone may vary. This study examined the complications that arise, with respect to the valid use of change scores, when the constructs of interest reflect conceptual change.
CFA techniques are available enabling structural comparison of the equivalence, or invariance, among factors across groups, situations, and/or time applicable to situations where issues of construct coherence and stability threaten the valid use of survey data. In that factor structure reflects the "mental model" expressed by a group of respondents for a given construct, these techniques more importantly can be utilized to provide as yet untapped evidence of conceptual change, widely theorized to precede behavioral outcomes. This investigation of factorial invariance served as the means to examine the extent to which systemic reform-minded professional development was associated with the structural evolution of teachers’ perceptions with respect to the multi-dimensional nature of classroom practice (traditional, investigative culture, investigative practice factors).
The findings from this study provide evidence that teachers who have participated in reform-minded professional development envision their teaching practice in different ways Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. than teachers that have not yet been reached. Although treatment exposure was not associated with extensive alterations in the measurement structure for any of the three teaching practice factors, these data do provide evidence of conceptual change in the relationships among factors in that higher levels of treatment exposure were found to be associated with reform factors both more distinct from each other and from the traditional practice factor. In addition, interpretation of these results presents dear implications and suggestions for improved evaluation practice and a deeper understanding of the challenge of change.
Phillips, Cynthia C., "Issues of Factorial Invariance Inherent in Conceptual Change: Teachers’ Evolving Perceptions of Classroom Practice" (2000). Dissertations. 1479.