Evaluating the Validity and Reliability of Authentic Learning Instruments using RASCH Model

Abstract

Social science research with a quantitative approach should use the appropriate scale of measurement to interpret a variable correctly. There are two types of measurement scale that is category scale and numerical scale. Category scale is for nominal and ordinal data, while the numerical scale for an interval and ratio data. In mathematical analysis, numerical scale indicates the exact measurement and more precise definitions of certain variables. Category scale is only used to classify a variable. However, Likert scale which is used for ordinal data are often used as a measurement tool in social science research. Therefore, the Rasch model analysis that can change the nominal and ordinal data to interval and ratio data was used in this study. Here, it was used to evaluate the validity and reliability of authentic learning instrument at Malaysian Polytechnics. Validity of instrument was seen by unidimensional analysis through principal component analysis method. Reliability of instrument was seen by summary statistics, variable map and item fit table. The findings show that the authentic learning instrument at Malaysian Polytechnics has high validity and reliability.