Study of applying network data analysis in item analysis

Abstract: The foundation of network data analysis in psychometrics rests on particular theory, special ontological assumptions and particular methodology. Therefore it is possible to call network analysis as an independent paradigm with given techniques for data gathering and data analysis. This study introduces the analysis of network data as a psychometric-mathematical technique, and its use in questionnaire item analysis. To achieve this goal, data gathered from implementation of a self-made questionnaire on 1000 Tehran's high school students. The questionnaire was made based on occupational-personality Holland theory. Researchers analyzed questionnaire’s items according to conventional methods (classical test theory, item response theory and factor analysis) and proposed method (network data analysis). Comparison of the results shows that a complete conformity between conventional methods final outputs and the outputs from network data analysis. Therefore, psychometricians can use network data analysis independently or alongside conventional methods to analyze questionnaires items. Advantages of such a method are simplicity, accuracy, being virtually and integrity.