A Validation of the Assessment Practices Inventory Modified (APIM) Scale using Rasch Measurement Analysis

Abstract

There are many instruments that have been designed to measure assessment practices skills, but very few have been validated for their soundness and consistency in measuring lecturers’ assessment practices skills. This study was undertaken to examine the psychometric properties of the Assessment Practices Inventory Modified (APIM) scale, and its soundness in measuring assessment practices skills among university lecturers. A quantitative survey research design was adopted for this study. The 50-item APIM scale on a five-point Likert scale was administered to a sample of 321 lecturers randomly selected from six universities in Uganda. The data collected was analysed using WINSTEPS Rasch Measurement Modelling Program for both Classical Test Theory (CTT) and Item Response Theory (IRT) to test the psychometric properties of the APIM scale. From the results of both the CTT (Cronbach’s alpha and the point bi-serial coefficients) and IRT (category probability curve, item and persons’ reliabilities, item characteristic curve, item difficulty, fit statistics, and principal component analysis) in this study, the APIM scale was found to have adequate psychometric properties in measuring assessment practices skills among university lecturers. The APIM scale was also found to be invariant to gender of the university lecturers. In conclusion, the APIM scale has been found to be sound and consistent in measuring university lecturers’ assessment practices skills. This study has pronounced a sound and consistent instrument in measuring assessment practices skills among university lecturers in Uganda, and has provided universities in Uganda with a valid and reliable instrument which will measure assessment practices skills of their lecturers. The results of this study have highlighted that the APIM scale can universally measure assessment practices skills among university lecturers