Investigation of the Methods for the Comparison of Factor Patterns of Independent Two Groups: A Simulation Study

Objective: In this study, when a scale being applied to two different groups, methods used to compare factor structures were discussed and the dependencies of these methods on the sample size, number of variables, and number of factors were investigated with simulation. Material and Methods: In this research, different number of variables (10, 20, and 40) and number of factors extracted (2, 5, 10 and 20) were treated for different sample sizes (50, 100, 200 and 500) with a 1000 replicated simulation study. In order to be able to distinguish factors good, the correlations between variables in the same factor were high (r=0.80), the correlations between variables in different factors were kept as low for each group. Results: Root Mean Square Coefficient (RMS) increases as the number of variables increase, when the number of factors is fixed. We can also see decreasing in the values of RMS and increasing in the Pearson correlation coefficients with sample size. Since it is difficult to obtain similar factor structures with excessive factors, correlation coefficient decreases as number of factors increase, when the number of variables is fixed. Conclusion: It was agreed that the sample size should be at least 10 times larger than the number of variables (n ? 10p). Furthermore, the number of variables should be two times larger than the number of factors (p > 2f; p: number of variables, f: number of factors).