Choosing the most suitable personality questions in the measurement of personality dimensions: combining the latent trait theory and network data analysis

The word personality refers to the uniqueness, individuality and subjectivity of the subject being studied. The measurement of such a dynamic and complex concept is considered a fundamental challenge in the field of methodology for the measurement of psychological constructs. The aim of this research is to present a new method in two different parts of personality questionnaire question analysis: a) personality questionnaire question dimensions obtained from the implementation of questionnaires on independent samples through correspondence analysis and b) question prioritization using from the network data analysis method based on the importance of questions in each dimension. To achieve these goals, 32 personality questionnaires - which cover most of the application areas of personality questionnaires - were implemented on 82,988 volunteers via web-based forms. Correspondence analysis results show that personality has two dominant dimensions that explain about 75% of personality variance. The results of network data analysis show that the important questions in different indexes are not necessarily the same and the selection of questions based on a specific index should be based on the meaning of that index, however, according to the correlation structure of the priority of questions in the index network, a general index was defined based on which questions were prioritized in two dimensions of personality. The result of the present research led to the presentation of an algorithm for selecting personality questions in personality dimensions.