Evaluation of West African Examinations Council's Physics Essay Test Using Partial Credit Item Response Theory Model
Fidelis O. Nnadi, Rose C. Anamezie
The poor state of secondary school students' achievement relative to policy expectation in Physics triggered the study. At the level of instrument quality, the scoring pattern in West African Senior School Certificate Examination's (WASSCE) Physics essay m arking scheme, difficulty -parameter level of the items and differential item functioning were sought. Instrumentation research design was adopted. One hundred and sixty senior secondary III Physics students consisting of 63 female and 97 male students form ed the sample. The 2018 WASSCE Physics essay with its marking scheme formed the instrument used to collect data. The internal consistency reliability of the instrument sought using Kuder-Richardson's formula 21 had value of 0.81. Three research questions and one null hypothesis guided the study. Research questions were answered using frequencies, mean infit/outfit statistic, mean item difficulty and mean standard error. The null hypothesis was tested at.05 level of significance using Mantel -Haenszel's Chi-square probability value. The result indicated that fractional-polarization scoring was adopted for two items. 25 in 28 items fit the partial credit model. The items were moderately difficult and did not exhibit gender-based differential item functioning. It is recommended that West African Examinations Council should remove fractional -dichotomization scoring and breakup marks allocated to an item into more than two groups for all items in future Physics essay marking scheme.