Frontiers in Education (Nov 2024)

Psychometric properties of an instrument to assess teachers’ practice of differentiated instruction (IATPDI): a confirmatory factor analysis on Bhutanese teachers’ data

  • Tshering Dorji,
  • Pelden Nima

DOI
https://doi.org/10.3389/feduc.2024.1445865
Journal volume & issue
Vol. 9

Abstract

Read online

IntroductionWe developed the IATPDI questionnaire, comprising seven scales for assessing teachers’ implementation of assessment methods, content, process, and product differentiation, familiarity with and use of various differentiated instruction (DI) strategies, factors influencing DI implementation, and resources used to enhance DI efficacy. This study examined the psychometric properties of the initial four scales with 35 items.MethodsThe questionnaire was administered to a sample of 237 Bhutanese teachers (66.2% male, 33.8% female), and confirmatory factor analysis (CFA) was used for psychometric evaluation.ResultsCFA supported the hypothesized four scales (CFI = 0.911, TLI = 0.903, SRMR = 0.052, RMSEA = 0.059, χ2/df = 1.58). Cronbach’s alpha coefficients ranged from 0.83 to 0.92, and Composite reliability (CR) ranged from 0.86 to 0.92, indicating high internal consistency reliability. Inter-factor correlations supported discriminant validity for most factor pairs, but correlations exceeding 0.85 between some pairs suggested potential overlap, prompting further investigation. Average Variance Extracted (AVE) values for assessment, content, process, and product factors were 0.50, 0.41, 0.53, and 0.51, respectively. While AVE for process and product factors surpassed the commonly accepted threshold (0.50) for convergent validity, the assessment factor approached the threshold and the content factor fell below it, indicating the need for further refinement of its indicators. However, all standardized factor loadings were significant (p < 0.05), confirming convergent validity.DiscussionThese results indicate that the proposed four scales of the IATPDI questionnaire are reliable and valid in measuring the intended constructs. Nevertheless, they also highlight the need for further refinement in identified areas to enhance the tool.

Keywords