Estimating the psychometric properties of 50 multiple choice mathematics item for senior secondary II students in Ikwerre Local Government area of Rivers State using classical test theory.

Abstract

Test development for assessment is an art that needs accurate estimation of item difficulty, discrimination and chance factor. This study was on estimating psychometric properties of 50 items multiple choice test item in mathematics. Three research questions were stated to guide the findings of the study. The population comprised all senior secondary II students in Ikwerre Local Government Area. Simple random sampling technique was used to select the sample of the study. A total of 900 SS II students were used for the study. A self-constructed multiple choice test in mathematics made up of 50 items was constructed from SSCE Mathematics syllabus, using a table of specification. The reliability of the instrument was determined using Kuder – Richardson formula 20 (KR20) and the reliability coefficient was 0.79. The test was administered to the selected sample and the psychometric properties estimated using Classical Test Theory. Results showed that thirty-five items were to be retained while fifteen items were discarded based on estimate of difficulty index. Two items have high positive discrimination, seventeen items have moderate positive discrimination, while sixteen items are on borderline positive discrimination and fourteen items had zero discrimination and one item has a negative discrimination. The distracter index is forty-two items on negative distracter, thirty items on positive distracter and eight items on zero distracter. Based on these recommendations were made including that test developers should be trained and know how to apply the formula effectively