Reliability and Validity of the Strengths and Difficulties Questionnaire in Abu Dhabi Kindergartens: A Rasch Analysis

This study examined the psychometric properties of the teacher version of the Strengths and Difficulties Questionnaire (SDQ) using Rasch analysis in a sample of 326 kindergarten children hailing from one of the emirates in the United Arab Emirates, Abu Dhabi. We followed the following procedures to evaluate the psychometric properties of the SDQ. First, we examined the stability and accuracy of the SDQ using Linacre’s (2004) six criteria with respect to category frequencies, average measures, threshold estimates, category probability curves, and category fit. Second, to establish the unidimensionality of the SDQ, a Rasch principal components analysis (PCA) of the standardized residuals was conducted (see Linacre, 1998). Third, we examined the item and person fit statistics to detect measurement disturbances. Fourth, the indices for person and item separation as well as for person and item reliabilities were examined. Finally, differential item functioning (DIF) analyses were conducted to investigate the invariance of the items across gender. The rating scale diagnostics revealed sufficient number of observations for each of the response categories, and the average measures increased monotonically. An examination of the point-biserial correlations revealed that none of them were negative or zero, suggesting item polarity. The Rasch unidimensionality coefficient was found to be substantially greater than the minimum value of .50, suggesting unidimensionality of the SDQ. The summary statistics for person and item fit suggested that the overall fit for persons and items were within the acceptable range of .06 to 1.4 (see Linacre, 2015). Moreover, person and item separation as well as person and item reliability indices demonstrated that SDQ had adequate internal consistency reliability. Finally, the Rasch-Welch and the Mantel tests, as well as the DIF contrast suggested the absence of DIF or invariance of the items across gender. In sum, the results of the Rasch analysis provided support for the appropriateness of the SDQ as a measure of behavioral adjustment in Abu Dhabi kindergarten children, thereby, to a certain extent, contributing to the cross-cultural generalizability of the teacher version of the SDQ.