1. Improving Measurements in Large-scale Surveys and Using Survey Data to Assess Program Impacts
- Author
-
Zhong, Shujin
- Subjects
Education ,Public health ,Quantitative psychology ,Causal Inference ,College Students' Science Career Intent ,Faculty Mentoring ,Item Response Tree Model ,Matching ,Sensitivity Analysis - Abstract
Using surveys to collect data for evaluating program effectiveness is a common approach in large national multi-pronged program evaluation, and it is widely used in the evaluation of biomedical training programs initiated by the Diversity Program Consortium (DPC). The dissertation consists of three studies. In study one, Measuring Research Mentoring Skills: Revisiting the Faculty Mentoring Competency Assessment and Developing a Short-form to Measure College Faculty-Student Mentoring, I investigated and validated the between-item dimensionality of the Mentoring Competency Assessment (MCA) and created a short form of MCA for future evaluation that was tailored to the DPC population. In study two, An Item Response Tree Modeling Approach for Assessing “Not Applicable” Responses in the Enhance Diversity Study, I used an item response tree (IRTree) modeling approach for assessing “Not Applicable" or "N/A” responses, and took the measurement of faculty mentoring as an example to examine the nature of the “N/A” responses in the MCA scale and investigate within-item dimensionality. In study three, Evaluating the Impact of the BUILD Scholar Program on First Year College Students' Intent to Pursue Science-related Research Careers, I examined the effectiveness of the BUILD scholar program, an affiliated undergraduate diversity training program developed at each BUILD site. I studied the influence of program participation on students’ intent to pursue science-related research careers during students’ initial stage in college.
- Published
- 2021