This study examines the item validity of a checklist on teaching performance fitness developed to identify unsuitable elements in the early process of teaching performance. For this purpose, the researcher collected data from 253 administrator-level t...
This study examines the item validity of a checklist on teaching performance fitness developed to identify unsuitable elements in the early process of teaching performance. For this purpose, the researcher collected data from 253 administrator-level teachers at kindergartens, elementary, middle, and high schools, as well as nontraditional schools nationwide by presenting them with virtual risk teacher cases and asking them to complete a checklist. The study used a final checklist consisting of five subfactors (antisocial aggression, maladjustment due to mental issues, interpersonal conflict resulting from victim mentality, neglect of duty, and sexual misconduct). The checklist contains a total of 30 items, with six categories under each subfactors.
The collected responses were analyzed using item response theory (IRT) with SPSS and R. After reviewing each item for response distribution and inter-item correlation, the researcher examined basic hypotheses such as uni-dimensionality for an IRT application and local independence. For analysis models, the study applied a dichotomous model for dichotomous items and Samejima's generalized partial credit model (GPCM) for polytomous items. Based on these factors, the study calculated parameters such as item discrimination (a) and difficulty (threshold) and additionally reviewed the possibilities of response categories.
The findings show that most items had a good or high level of discrimination, which means that each of the subfactors can effectively sort out intended risk teachers. One item from each indicator failed to reflect teaching performance fitness fully or discriminate it; thus, they were eliminated. For the six items that functioned as dichotomous, the original 3-point scale was revised into a 2-point scale.
These results suggest that a checklist can be utilized as a useful assessment tool to detect the problematic behavior of teachers and signals those that are unfit for duty early. By testing items for validity based on IRT, the study can also contribute to the development and standardization of an assessment tool for teaching performance fitness at the national level. These findings may serve as important materials for educational institutions to reinforce their teacher management system and protect their students.