Abstract
The study investigated the determination of aberrant examinees' responses of 2018 and 2019 May/June Biology multiple choice test questions of the National Business and Technical Certificate Examination (NABTCE) in Nigeria. Six research questions were raised, and two hypotheses were formulated and tested at 0.05 level of significance. This research adopted the survey research design of the expost facto. The population for the study was 57357 Biology students responses which consisted of 28121 and 29254 for 2018 and 2019 students' scored responses of the NABTCE respectively. The sample size for the study was 13533. This comprised of 7086 and
6447 examinees' responses for the 2018 and 2019 NABTCE Biology. A multistage sampling technique was adopted for the study, in which a stratified sampling technique, simple random sampling technique and a cluster sampling was applied for effective selection. The instrument used was the NABTCE May/June Biology multiple choice test questions for 2018 and 2019 made up of fifty items with four options each. The instruments were assumed to be valid and reliable as it was a test constructed by a public examination body. The Perfit which is a non parametric Item Response Theory (IRT) statistic was used to determine aberrant examinees' responses. Examinees' responses whose HT coefficients were less than or equal to .25 and .15 were flagged as aberrant for the 2018 and 2019 respectively. Sample with percentages of aberrancy of ≤ 5% for low level, moderate level 6% - 15% and high level 16% - 25%. The research questions one and two were answered using frequency counts and percentages, research questions three and four were answered using Cronbach Alpha statistics while hypotheses one and two were tested using Chi-Square test of goodness of fit. The findings of this study revealed that the levels of the aberrant examinees' responses in the 2018 and 2019 NABTCE Biology multiple
choice test examination were moderate. It was also revealed that there was a slight insignificant improvement in the internal consistency reliabilities of the NABTCE Biology for 2018 and 2019 multiple choice test after screening the data of aberrant
examinees. There was also a significant difference in the frequency of occurrence between the aberrant examinees and the non aberrant examinees for the NABTCE Biology for 2018 and 2019 multiple choice test. Based on these findings, the researcher recommended amongst others, that the National Business and Technical Examination Board should continue in their good work for constructing Biology multiple choice items having low aberrant examinees' responses. Psychometricians, test researchers and counsellors should be aware that the impact of aberrancy may be strong enough to cause misclassification of ability or construct.