Benue State University, Makurdi

EMPIRICAL COMPARISON OF CLASSICAL TEST THEORY AND ITEM RESPONSE THEORY USING GEOGRAPHY ACHIEVEMENT TEST IN BENUE STATE

Yeke, C. M., Onah, D. O., Emaikwu, S. O. and Obinne, A. D. E.

Abstract

This study focused on Empirical Comparison of Classical Test Theory and Item Response Theory using Geography Achievement Test in Benue State. In carrying out the study, four research questions were posed and two hypotheses formulated. The study adopted Ex-post-facto research design. The sample size for the study was 581 secondary school Geography students representing 60% of the population for the study. This was drawn from 986 geography students randomly selected from146 secondary schools in Benue North-East Education Zone. A - 50 item Geography achievement test was adopted and used for data collection. The reliability coefficient of 0.93 was obtained using Kuder Richardson 20 formula (K-R 20 ). The instrument was administered to respondents with the aid of research assistants. The research questions posed were answered using percentages, difficulty and discrimination indices with the aid of Bilog-MG statistical package while the hypotheses formulated were tested using independent t-test statistic at 0.05 level of significance. The result revealed that, statistical significance exists between the CTT and IRT in item discrimination and item difficulty indices in favour of IRT. All the hypotheses tested were rejected as they were all statistically Significant in favour of IRT. Based on the findings it was concluded that the two-parameter logistic model can successfully be applied in the determination of item statistics and the items were good for measuring achievement of Geography students with little modification. It is therefore recommended among others that: The item difficulty and item discrimination estimates of students’ achievement test in Geography based on CTT revealed that few of the items had poor difficulty and discrimination indices, the Item difficulty and item discrimination estimates of students’ Geography achievement test (GAT) based on IRT revealed that, majority of the items have better difficulty and discrimination indices to measure students’ ability than CTT.

DOWNLOAD COMPLETE ARTICLE HERE (CLICK)