Title:
COMPARISON OF RELIABILITY COEFFICIENTS OF TESTS CONSTRUCTED USING DIFFERENT ORDERINGS OF DIFFICULTY OF TEST ITEMS

Authors:
Kaku Zakka and Romy O. Okoye

Abstract:
The study compared reliability coefficients of tests constructed using different orderings of difficulties of test items for English language and Mathematics in the North-East, Nigeria. Experimental research design was adopted. The population of the study comprised all the 5,403 JSS 3 students of 2021/2022 academic session in the fifteen Unity schools of the zone. The sample used for experimental study consisted of three hundred and seventy-five students obtained through a combination of multi-stage and cluster sampling techniques. Two instruments (one for each of the subjects) were used for the study. Each instrument was prepared in three different formats, titled: Mathematic Achievement Test and English Language Achievement Test. Each format of the instruments was subjected to test-retest reliability. Method of data analysis was Pearson Product Moment Correlation coefficient (PPMC). The reliability estimate that were obtained from the mathematics multiple-choice questions are as follows: FORMAT ETD (0.512), DTE (0.849) and MDI (0.718). The reliability estimate that were obtained from the English language multiple-choice questions are as follows: FORMAT ETD ETD (0.728), DTE (0.669) and MDI (0.877). The study recommends that for English Language, items ordered from easy to difficult format should be used when constructing test items; since it was the one that had the highest reliability coefficient; for Mathematics items format from difficulty to easy should be used since it has the highest reliability coefficient among the test items format.

Keywords:
Reliability, Test item ordering, Order of difficulty, Test items, Test item arrangement.

DOI:
https://doi.org/10.37500/IJESSR.2023.6321