Using Horn’s Parallel Analysis Method in Exploratory Factor Analysis for Determining the Number of Factors

In this study, the number of factors obtained from parallel analysis, a method used for determining the number of factors in exploratory factor analysis, was compared to that of the factors obtained from eigenvalue and scree plot—two traditional methods for determining the number of factors—in terms of consistency. Parallel analysis is based on random data generation, which is parallel to the actual data set, using the Monte Carlo Simulation Technique to determine the number of factors and the comparison of eigenvalues of those two data sets. In the study, the actual data employed for factor analysis was gathered from a total of 190 primary school teachers using the Organizational Trust Scale to explore a teacher’s views about organizational trust in primary schools within the scope of another study. The Organizational Trust Scale comprises 22 items under the three factors of “Trust in Leaders,” “Trust in Colleagues,” and “Trust in Shareholders.” A simulative data set with a sample size of 190 and 22 items was simulated in addition to the actual data through an SPSS syntax. The two data sets underwent parallel analysis with the iteration number of 1000. The number of factors was found to be three. This was consistent with the number of factors obtained in the development process of the scale. The number of factors was restricted to three and exploratory factor analysis was re-performed on the actual data. It was concluded that the item-factor distributions obtained as a result of the analyses were consistent with those obtained in the scale development study. Hence, parallel analysis was found to provide consistent results with the construct obtained in the scale development study.