DIFFERENTIAL ITEM FUNCTIONING OF 2015/2016 BIOLOGY MULTIPLE-CHOICE ITEMS OF OSUN STATE JOINT PROMOTION EXAMINATION

Bamidele Abiodun FALEYE & Abdulwasiu Adeyemi RASHEED

Abstract

Differential Item Functioning is one of the important and currently used methods to investigate test fairness. This study investigated the test items that function differentially based on gender and school location of the test takers in 2016 Biology multiple-choice items of Osun State Joint Promotion Examination (OSJPE). The research design adopted for this study was ex-post-facto design. The Sample for the study was 2,000 respondents who sat for OSJPE in 2016. The sample was selected using a multi-stage sampling procedure. Three Local Government Areas (LGAs) were selected from the three senatorial zones in Osun State, using a purposive sampling technique (where at least two schools were located at both rural and urban in each of the LGAs). From each Local Government Area, four schools (two from each of rural and urban centres) were selected using a stratified sampling technique, using location as a stratum. The students’ responses of the 40 multiple-choice questions were subjected to the 3-parameter model of Item Response Theory analysis using the Calibrating process of analysis setting Mantel Haenszel as statistics. The result showed that One (item 5) out of the 40 multiple-choice Biology items functioned differentially for male students. Eight items (19, 20, 23, 28, 36, 37, 38 and 40) were biased against rural and urban students. The findings implied that there are incidences of differential item functioning in 2016 OSJPE Biology multiple-choice items. It was recommended that examination bodies should organise training for item developers on the construction of valid, reliable and fair tests especially in the area of DIF. In addition, items flagging DIF should be revised, modified or eliminated from the test.