ESTIMATING WEST AFRICAN EXAMINATION COUNCIL MATHEMATICS THEORY TEST ITEMS OVER MULTIPLE FACETS USING GENERALIZABILITY THEORY
1Catherine U Ene, Sylvanus C. Ohagwu, Basil C. Oguguo, Mercy Ngozi Nwoye, Francis Elochukwu Ikeh, Felicia Chinyere Ugwu, Janehilda Oluchi Agugoesi
The study investigated the application of generalizability theory in estimating West Africa Examination Council (WAEC) mathematics test scores over multiple facets. The study employed random effects, two-facets fully crossed design for a G-study and D-study. A sample of 898 senior secondary three (SS3) students was used for the study. Four research questions guided the study. Mathematics Achievement Essay Test (MAET) was used to collect data for the study.The instrument was validated by three experts in Science Education Department (one in Mathematics Education Unit and two in Measurement and Evaluation Unit), Faculty of Education, University of Nigeria Nsukka, Enugu State, Nigeria. The reliability of the instrument was established using Kendal coefficient of concordance (W) and reliability coefficient of 0.91 was obtained. A computer program EduG version 6.1-e based on the Analysis of Variance (ANOVA) and Generalizability Theory (GT) was used to carry out the Generalizability analysis. The resultrevealed that largest contribution of error variance was from the students followed by student -by-question-by-rater factor. The third largest variance component was student-by-question interaction followed by variance of the rater. The variances due to question, student-by-rater and question-by-rater interaction were zero each.
Generalizability Theory, Mathematics, West African Examination Council