Author:
Adeyemo Emily Oluseyi,Opesemowo Oluwaseyi Aina
Abstract
The study determined the parameter estimate of the Senior School Certificate Mathematics items of June/July 2017 NECO examinations and testlet effect under Multilevel Measurement Modeling with the aim of providing information on the psychometric properties and quality of the items. The research design was an ex-post facto, The examinees response were the multiple – choice items of the National Examinations Council Mathematics paper two for June/July 2017 which consisted the data for the study group. The targeted population consisted of 26,086 senior secondary three examinees who registered for Mathematics Senior School Certificate (NECO) in June/July 2017 in Osun State. A total of 318 private schools and 179 public schools registered for the paper. The results revealed the following items to be good which implied that such items functioned well, these were: items 1, 4, 7, 8, 9, 10, 11, 12, 14, 15, 17, 18, 20, 21, 22, 23, 26, 27, 28, 29, 31, 32, 33, 34, 35, 36, 37, 40, 42, 43, 44, 45, 46, 47, 48, 50, 51, 52, 54, 57, 58, 59, 60 whereas items that were considered to be bad included 2, 3, 5, 6, 8, 13, 16, 19, 24, 25, 30, 38, 39, 41, 49, 53, 55 and 56. Furthermore, investigating an average bundle of item statistics under the measurement framework indicated that the Item Discrimination Means value and Standard Deviation under IRT approach were 1.26 and 0.60 respectively while the Mean value difference was 1.26. Although, item difficulty Mean value and the Standard Deviation were 0.26 and 4.26, respectively, whereas the Mean value difference was 0.26. Similarly, the guessing Mean value and the Standard Deviation were 0.15 and 0.19, respectively, whereas the Mean value difference stood at 0.15. The study concluded that any standardized examination, especially from an examining body in charge of certificate examination, if issues like differential testlet effect is not taking into consideration, it could harm the validity of the items and also alter the ability estimates of the examinees The validity of the test would be strengthened when issues like differential testlet effect is adequately taken care of.