Turkish adaptation of the digital literacy scale: A rasch analysis

Abstract

The study took a Rasch measurement theory approach to validating the 10-item Digital Literacy Scale (DLS) using the unidimensional rating scale model (RSM). To that end, the study used the data from a sample of online Turkish university students. The study began the Rasch analysis with all 10 items in the scale and, to improve in the local independence assumption, identified and eliminated two items which did not adequately fit the RSM. Under the eight-item DLS, the assumptions of undimensionality and local independence were both satisfied and the fit of all individual items to the RSM was adequate. Next, the psychometric properties of the eight-item DLS were examined including rating scale effectiveness, relative endorsability of the items, differential item functioning (DIF) by each of three demographic variables: (a) gender, (b) connection device, and (c) grade level. Through the analysis, evidence of reliability and validity was identified which generally supports the use of the DLS instrument among the population of online Turkish university students from which the sample was obtained. The study also identified items which demonstrated either misfit to the model or DIF by the demographic variables, and recommends they be further reviewed and revised for future use.