Transformation Models as Predictive of Comparability Indices of Continuous Assessment in Nigeria.

Abstract

This study investigated the predictive strength of Transformation models of continuous assessment scores in English Language, Mathematics, Integrated Science and Social Studies in Secondary Schools in Nigeria. The study employed survey and cross-sectional design. The sample consisted of 2,520 JSS III students selected from 36 secondary schools in 18 Local Government Areas based on multistage sampling technique. Data were collected directly from the Ministries of Education, Continuous Assessment Units with a proforma. The data collected were subjected to inferential analysis using regression. The results showed that the overall best predictor of comparability indices was True score followed by Predictive true score, T-score, derived true score while Z-score was the least. Based on the findings, it was recommended that secondary school teachers should be trained on the procedure to transform continuous assessment scores. True score should be used to standardize continuous assessment scores before finally used by examining bodies and continuous assessment units of Ministries of Education.

Keywords: comparability indices, continuous assessment, predictive true score, true score, derived true score