Abstract
The language of communicating test tasks plays a vital role on the performance of students. It is a factor to be considered in item functioning differently among sub-group of examinees. The study examined the effect of language manipulation on Physics multiple choice items that functioned differentially for students. It also compared the difference in differential item functioning indices before and after manipulation based on gender and school location. The study adopted the pre-test post-test quasi-experimental research design. The population for the study comprised all secondary school students in Osun State. A sample of 720 senior secondary school III Physics students was used for the study using the multi-stage sampling procedure. Two research instruments (adopted and adapted 2017 WAEC Physics items) were used for data collection. The adopted instrument was already a standardized instrument developed by the examination body. However, the reliability estimates of the adapted version yielded a Crombach's Alpha coefficient (α) of 0.89. Data collected was analysed using the Wald-type statistics and related samples Wilcoxon ranked test. The result showed that 100% reduction in DIF indices was observed when item change from negative to positive phrase. Also, 80%, 70% and 75% reduction in DIF indices were observed when items were manipulated by rewording, rephrasing and using simpler synonyms respectively. The result showed that the differences observed in the DIF indices before and after language manipulation based on gender (Z=2.981, p=0.003<0.05) and school location (Z=3.530, p=0.000<0.05) were significant. The study concluded that language manipulation had significant influence on WAEC Physics multiple choice items which are flagged as functioning differentially among Osun State secondary school students. It therefore, recommended that simpler and short sentences should be used in stem of objective test so that performance of candidates can be enhanced.