The Effect of the Number of Response Categories on the Assumptions and Outputs of Item Exploratory and Confirmatory Factor Analyses of Measurement Instruments in Psychological Research

Many researchers have investigated exploratory and confirmatory factor analyses whether in Arabic or English studies. However, no study has investigated the effect of the number of response categories on the assumptions and outputs of item factor analysis, as one of the most advanced methodologies for handling ordinal data. Thus, the present study aims at identifying the effect of the number of response categories (3, 5, 7) on the assumptions and outputs of item factor analysis. Participants were 307 university students. The academic boredom scale (Prepared by Mahsoub Abdelkader Eldowy Hassan, 2015) was used for data collection to benefit from after estimating the internal structure validity evidence for three versions based on the number of response categories using item factor analysis. Results revealed that the assumptions of item exploratory factor analysis were best met when the number of response categories was five or seven rather than three especially in terms of the amount of shared variance that is the basis for the factorability of data. Concerning the outputs of item exploratory factor analysis, when the number of response categories was three, the number of extracted factor increased with more insignificantly items’ loading compared to the five and seven response categories. Concerning the outputs of item confirmatory factor analysis, as the number of response categoriesdecreases, fit indices improve but the model parameter estimates were attenuated and consequently the low probability of rejecting the factorial model which does not underlie the data especially when the number of response categories was three. Some recommendations and suggestions for future research were presented.