Various criticisms have been leveled against psychological testing. A large proportion of the criticism pivots on the construct validity of test items. This article discusses the procedures to adopts in validating Mathematics test Items. The design of the study is instrumentation. A multi – stage sampling technique was used to acquire a sample size of 200 students for the study. The instrumentation for the study was a self- developed 150 objective mathematics test items. The content validity was examined based on some experts’ judgment on the development of the items. The analysis of data was based on win steps analysis. The result showed that the 86 items not only met the one parameter-latent trait model assumption of measurement construct but also demonstrated good psychometric properties.
Keywords: construct validity, one parameter latent trait theory, unidimentionality and classical test theory.
