Abstract

 study assessed effect of language manipulation on differential item functioning ofAgricultural Science multiple choice test items used by National Examination Council in the Senior School Certificate Examination (SSCE) 1013-2015. A post test only control group of quasi experimental design was tmploved. The population consisted of 3,238 senior secondary school students in class three (3) in public secondary schools in Imo State who studied Agricultural Science as a certificate subject. A sample size of 100 students was used for the study got by three stage proportionate stratified random sampling techniques. Three research questions and three hypotheses were asked and formulated respectively. The instruments of data collection were: Agricultural Science paper 3 objective test used by NECO in 2013-2015 examinations; Agricultural Science form A test made up of 30 test items that were identified as differentially functioning; Agricultural Science form B test which was made up of the 30 items in form B, but presented in simplified language format. The fourth instrument was student's socio-economic status questionnaire that was used to collect students’ biographical information. Scheuneman modified chi-square statistic was used to analyze the research questions. Dependent t-test statistic was used to test the null hypotheses at 0.05 level of significance. The study revealed that Agricultural Science multiple choice questions used by NECO in the SSCE 2013-2015 contain test items that significantly function differently for students based on socioeconomic status, geographical locations and gender of testees. It was also found that simplifying the language ofAgricultural Science test items brought about significant reduction in differential item functioning. Based on the findings it was recommended that test item writers for public examinations should be encouraged to keep simple English language used to reduce the tendency ofintroducing additional tasks into the measurement ofAgricultural Science ability’.