Test Item Validity: Item Response Theory (IRT) Perspective Nigeria

Abstract:

Tests as measuring instruments are designed for a variety of purposes, and validity is often evaluated only in terms of purpose. Validity refers to whether or not the test measures what it claims to measure. With item response theory (IRT) the internal validity of a test is assessed in terms of the statistical fit of each model. If the fit statistic of an item is acceptable, then the item is valid. A study aimed at determining the validity of the test items in the Biology examination conducted by the National Examination Council (NECO) and the West African Examination Council (WAEC) from the year 2000 – 2002 using the fit statistic technique confirmed the validity of the items, as the items had fit-statistic that fitted the two-parameter model of IRT. The study sample comprised 1,800 Senior Secondary School Year Two students in all the three education zones of Benue State, Nigeria. The multi-stage stratified sampling technique was used to obtain the study sample while the Maximum Likelihood Technique of the Bilog-MG Statistical Analysis Package and t-test were used for analysis. Results showed that, using the Two-parameter model, items for the year 2000 showed no significant difference just as the validity of items of the Biology Examinations conducted by the two examination bodies in Nigeria (NECO and WAEC). The result has established that the fit statistic (validity) of the two examination bodies (NECO and WAEC) is not statistically different, based on IRT.