Tag Archives: Item Bias

Examining the Item Bias of Mathematics Examinations Constructed by WAEC and NECO in Nigeria (Published)

This study examined the item bias of mathematics examination items constructed by West African Examination Council (WAEC) and National Examination Council (NECO) in Nigeria on the basis of probability of success of examinees of comparable ability. The study employed descriptive research design of the survey type.  A sample of 600 Senior Secondary School three (SSS III) students was randomly selected from 12 Senior Secondary Schools in Ekiti State, Nigeria using multistage sampling technique.  Two Instruments which included 50 mathematics test items randomly selected from past WAEC questions and 50 mathematics test items randomly selected from past NECO questions were used to collect data for this study. Two hypotheses were formulated and tested at 0.05 probability level. Two- parameter Logistic Bilog MG software statistical analysis was used to generate Item Response Theory (IRT) item characteristic curves (ICCs) for each NECO and WAEC Examination items to examine whether the items exhibit item bias or not . The results show that 44 (88%) of NECO items exhibit item bias while 35 (70%) of the WAEC items exhibit item bias. Hence, it was recommended that examination bodies should commit itself to eliminating or reducing biased items in National Examinations. 

Keywords: IRT and ICC Curves, Item Bias, NECO, WAEC

Detecting Item Bias in Rivers State JSSCE Business Studies Using Item Response Theory (IRT) Approach (Published)

This study identified biased test items in the Rivers state JSSCE Business Studies conducted in 2009 using Item Response Theory Approach. The Sample comprises 4000 JSS III students in public and private schools that sat for the JSSCE Business Studies Examination in 2009 in Rivers state which were selected through stratified proportionate random sampling. The instruments for data collection were the test items in JSSCE Business Studies conducted in 2009 and a result collection sheet. Two null hypotheses drawn from two research questions were tested at 0.50 probability of correct response. The assumptions of unidimensionality and local independence were checked using factor analysis. The students’ responses to the test items were calibrated with an IRT statistical software named ex-calibre 4.2.2 developed by Assessment Systems Corporation. The results revealed that the test items met the assumptions of unidimensionality and local independence. It also revealed that there is school type bias in the Rivers State JSSCE Business Studies test items for 2009. Based on the findings, recommendations were made, one of them being that business studies curriculum should be restructured so as to accommodate the new technological advances which may be one of the reasons for the presence of school type bias in the Business Studies test conducted in 2009 in Rivers State.

Keywords: Item Bias, Item Characteristic Curve, Item Response Theory, Unidimensionality