Sex-based differential item functioning in the Wechsler Intelligence Scale for Children–fifth edition: implications for educational assessment in Jordan
Nizar Allabadi, Mohammad Al-Karamneh, Jehad M. Al-Anati, Laith Jumah, Omar Alorani
Abstract
This study aimed to investigate differential item functioning in the Arabic version of the Wechsler Intelligence Scale for Children–fifth edition (WISC-V) in the Jordanian context by sex, using logistic regression and probability ratio testing. A descriptive approach was employed, and 15 subtests of the WISC-V scale were administered to a sample of 468 children, including normal children, gifted, learning-disabled, and mildly intellectually disabled children, selected using stratified random sampling. The children’s ages ranged from 6 to 6.11 years, and the age groups were defined as 11 age categories, each encompassing one calendar year. The logistic regression analysis revealed 14 items with sex-based differential performance in the cube design, memory capacity, information, visual memory, arithmetic, and shape weighting tests. The differential performance was uniform in 10 items and non-uniform in 4. The probability ratio test results showed sex-based differential performance in 10 items of the information test, including arithmetic and shape weighting. Differential performance was uniform in 3 items and non-uniform in 7 items. The overall agreement rate between the detection methods (likelihood-ratio test, logistic regression) was 97.4%. The study recommends employing multiple methods for detecting differential item functioning to enhance the accuracy of interpretation, as well as training psychologists to use test results with caution and to consider potential sex-related differences in specific items.