Abstract
This study examined the discriminating parameters of essay test items in the National Examinations Council (NECO) Senior School Certificate Examination (SSCE) Mathematics papers for the 2022 and 2023 June/July examinations using an Item Response Theory (IRT) polytomous model. The study also determined the number of Mathematics essay items that satisfied the required IRT statistical conditions. A descriptive survey research design was adopted. The population comprised all Senior Secondary School III (SSS III) students who registered for the NECO SSCE in Ethiope West Local Government Area, Delta State, Nigeria. A sample of 1,000 SSS III Mathematics students was selected from ten public and ten private secondary schools using simple and stratified random sampling techniques. The instrument for data collection consisted of the 2022 and 2023 NECO SSCE Mathematics essay past question papers. Item parameters were estimated using Xcalibre 4.1 IRT parameter estimation software developed by Assessment Systems Corporation (ASC). The generated parameters were analyzed using frequency counts, percentages, mean, and standard deviation to answer the research questions, while the formulated hypothesis was tested using an independent samples t-test at the 0.05 level of significance. The findings revealed that some of the Mathematics essay test items for both 2022 and 2023 did not satisfy the IRT statistical conditions. The results also indicated that there was no significant difference in the discriminating parameters of the NECO SSCE Mathematics essay test items between the two years when analyzed using the IRT polytomous model. Based on these findings, it was recommended that examination bodies responsible for the development and validation of achievement tests should ensure rigorous psychometric validation of test items to align them with the ability levels of examinees.
Keywords: Application, IRT Polytomous Model, NECO SSCE, Mathematics, Essay Test Items, Discriminating Parameters