Journal Information
|
| Research Areas |
| Publication Ethics and Malpractice Statement |
| Guidelines for Authors |
| For Authors |
| Instructions to Authors |
| Copyright forms |
| Submit Manuscript |
| Call for papers |
| Guidelines for Reviewers |
| For Reviewers |
| Review Forms |
| Contacts and Support |
| Support and Contact |
| List of Issues |
| Indexing |
Journal of Emerging Trends in Educational Research and Policy Studies (JETERAPS) (Vol 1 No 1)
| Abstract: Post-test estimates of item parameters of multiple-choice tests are usually ignored in the assessment industry, except when it is necessary for research purposes. The study estimated item parameters using responses of candidates in mathematics multiple-choice items for three years. This is a longitudinal study which adopted the ex-post facto research design. The population comprised all the candidates who sat for mathematics examination during the years under consideration. Sample consisted of 10% of those candidates. Candidates’ result in the Senior School Certificate examination conducted by a public examination body constituted data for the study. Data was analysed using Factor Analysis and X-Caliber IRT software. The results of the study showed variations in the item parameter estimates. Also, some of the items calibrated for three years showed avoidable clerical, concept and construct errors which could only be detected using statistical estimate of the parameters. The study therefore concluded that manual development of marking scheme for multiple-choice items is neither sufficient nor efficient enough to ignore statistical estimates. It is recommended that before scoring, statistical post-test estimates of multiple-choice test parameters ought to be carried out to ascertain the characteristics of the test items. |
| Keywords: Post-test, Item Parameters, Mathematics, Multiple-Choice Tests, Examination |
| Download full paper |
