Use of Differential Item Functioning (DIF) Analysis for Bias Analysis in Test Construction (Published)
The study used differential item functioning (DIF) analysis to determine item bias in a chemistry aptitude test (CAT). A total of 270 items were administered, in paper and pencil format for item analysis purposes. Items were analysed by means of classical item analysis, IRT item analysis and DIF analysis in particular. The sample size was 554 secondary school students from SS1 and SS3. The study focused on selected secondary schools in Yoruba speaking states of the western part of Nigeria, Hausa speaking states in the northern part of Nigeria and Igbo speaking states in the eastern Nigeria. The sample was divided into subgroups which was used to investigate the DIF for the level of Education, gender, language and culture. This was used to indentify items that indicated bias in terms of gender, culture, language or level of education. Items that exceeded a predetermined amount of DIF were discarded from that final item bank, irrespective of which subgroup was being advantaged or disadvantaged. The process and results of the DIF analysis were discussed.
Keywords: Classical Test Theory, Differential Item Functioning, Item Characteristic Curves, Item Response Theory
Determination of Differential Item Functioning By Gender in the National Business and Technical Examinations Board (Nabteb) 2015 Mathematics Multiple Choice Examination (Published)
The purpose of the study was to investigate the differential item functioning (DIF) by gender in National Business and Technical Examinations Board (NABTEB) 2015 Mathematics Multiple Choice Test Items (Dichotomous) Examination in Nigeria. This was conducted by determining the items that functioned differentially by male and female examinees. A survey research design was employed. A sample of 17,815 examinee responses was selected from two states each from two geo-political zones out of 63,584 examinee responses from the six geo-political zones in Nigeria. This comprised 11,873 male and 5,942 female examinees. A 50-item multiple choice Mathematics test item was used to gather data. To detect the items that functioned differentially by male and female examinees, Area Index (Raju) method which is one of the item response theory methods of DIF detection was applied. The results of the analysis revealed that male and female examinees functioned differentially in seventeen items (34%) and no difference in 33 items (66%). Out of the seventeen items, six items were in favour of male students while 11 items were in favour of the female students. Based on the result of the findings, it was then recommended among others that for bias-free items to be produced, examination bodies, test experts and developers should make certain that activities and connotations reflected in the test are relevant to the construct being measured and explore the use of Area Index method of DIF to detect the items that function differentially by gender.
Keywords: Area Index (Raju) method, Bias Item, Differential Item Functioning, Multiple Choice Item, NABTEB