MEASUREMENT INVARIANCE OF TURKISH “CENTRAL EXAM FOR SECONDARY EDUCATION” BY SPECIFIC LEARNING DISABILITY

Abstract

Ensuring measurement invariance for students with disabilities is critical for fair measurement in large-scale testing. Specific learning disability constitutes the largest group among disability groups. In this study, it was aimed to examine the measurement invariance of the Turkish Central Exam for Secondary Education according to whether or not students have a specific learning disability. 994 students diagnosed with specific learning disability formed the focus group, whilst 1,000 students without any disability constituted the reference group. Mantel Haenszel and Lord’s chi-square methods were used to determine whether or not the items in each subtest showed Differential Item Functioning (DIF). In addition, by applying Multigroup Confirmatory Factor Analysis, the configural invariance, metric invariance, scalar invariance, and strict invariance of the subtests were examined. The study’s findings proved that 34 of the 90-item test indicated DIF according to both methods. Eleven items show moderate DIF and five show high DIF. Metric invariance is not provided in all subtests, with factor loadings in all subtests varied between the groups.