Detecting the Differential Item Functioning of Numerical Ability Test in the Gulf Multiple Mental Abilities Scale by Mental-Haenszel and Likelihood Ratio Test

The current study aimed to examine differential item functioning (DIF) of a numerical ability test for Gulf state school students. This examined the DIF items by gender and country using two analysis methods; Likelihood Ratio Test (LRT) Mantel-Haenszel (MH). sample size was 2689 individuals throughout grades 5 6, researchers used MH with SPSS LRT BILOG-MG. classification stability coefficient kappa (?) compare how well methods agreed DIF. Regarding gender, yielded precise results 30% items, in terms country, 43.3%. Also, analyzing found that 26.7% exhibited based on country. For there strong concordance (0.925) between method LRT. agreement also high (0.683). suggested investigating causes items' performance comparing types, paper-and-pencil computer-based.