Date of Award

Fall 12-15-2021

Degree Type


Degree Name

Doctor of Philosophy (PhD)



Committee Chair

Kyna Shelley, Ph.D.

Committee Chair School


Committee Member 2

Thomas J. Lipscomb, Ph.D.

Committee Member 2 School


Committee Member 3

Richard S. Mohn, Ph.D.

Committee Member 3 School


Committee Member 4

Randolph C. Arnau, Ph.D.

Committee Member 4 School


Committee Member 5

Peter Paprzycki, Ph.D.

Committee Member 5 School



The main purpose of this research was to examine the relationship model between culturally responsive teaching self-efficacy (CRTSE) and culturally responsive teaching outcome expectancy beliefs (CRTOE) in Thailand. There were two research questions outlined in the study: what is the relationship between preservice teachers’ CRTSE and CRTOE? Also, does the culturally responsive teaching teacher empathy (CRTTE) moderate the relationship of preservice teachers’ CRTSE and CRTOE?

Based on the CRT standards by Siwatu (2005), there were three measures applied in the study: CRTSE, CRTOE, and CRTTE scales. These instruments were piloted with 536 preservice teachers in the Faculty of Education, Prince of Songkla University. The results suggested appropriate modification of the instruments to fit a new research setting. Psychometric examinations revealed that all three scales were an acceptable fit by confirmatory factor analysis. Cronbach’s alpha estimates were high (> .80). Conducting the differential item functioning (DIF), a few items functioned differently between two language groups of participants. However, the effect sizes were small (< .02) so they remained in the instruments.

Following the pilot, 510 preservice teachers were recruited for the study. Five relationship models were tested in order to answer the research questions. Descriptive statistics, regression analysis, moderation analysis, structural equation modeling, and invariance testing were all conducted to examine the models. The results revealed that an independent relationship between CRTSE and CRTOE was statistically significant, r = .747, p < .001. A dependent relationship between CRTSE and CRTOE was also statistically significant, with CRTSE explaining 55.8% of the variation in CRTOE. A dependent relationship between CRTSE and CRTOE with the structural model showed a very good fit: CFI = .99, TLI = .99, RMSEA = .070 (90% CI [.04, 1.00]), SRMR = .02. Considering the moderation effect of CRTTE in the model, the interaction between CRTSE and CRTTE was not statistically significant although effect sizes were different across levels of CRTTE. Finally, the invariance testing showed invariance models [ΔChi-Square (p > .05), ΔRMSEA < .010, and ΔCFI < .005]. Mean CRTSE scores of the medium and high CRTTE groups were significantly higher than the low CRTTE group.


Available for download on Friday, December 15, 2023