Document Type : Research Paper

Authors

• Reyhane Rahimi ¹
• Aso Mojtahedi ²

¹ Iran

² University student

Abstract

The purpose of this study is to examine the questions of the Likert scale with the Classical Test Theory and the Item Response Theory, then putting together the results of these two approaches and answering the question "Are the results of these two approaches consistent? or do they contradict each other?" . The research method is descriptive and in terms of data, it is a secondary analysis method. The subjects studied in this research included 977 students of the junior high school students, that by removing individuals in the data screening stage, finally, the analysis of extraversion items was based on the information of 783 subjects, the analysis of openness items on the information of 763 subjects, and analysis of conscientiousness items on the information of 784 subjects. The research instruments were 3 subscales of extraversion, openness and conscientiousness of the Neo Personality Test. The results of the statistical analysis showed that the higher the internal consistency of the items, the more accurate and valid the results obtained from the graded response model, and for items with low internal consistency through this model, it should be acted cautiously; Because it may show the threshold of the items inverted or the discrimination coefficient of the questions false negative or positive. All in all, the simultaneous use of different methods of statistical analysis will be of aid in better analysis and obtaining more accurate results.

Keywords

• graded response model
• Item Response Theory
• Classical Test Theory
• internal consistency
• threshold