Determining the Factors Affecting the Psychological Distance Between Categories in the Rating Scale

Abstract

In this study, the assumption of the equality of psychological distance between categories of rating scale was tested based on the number of categories and ability distributions. Category parameters were estimated by using generalized partial credit model. The data sets based on the conditions of categories counts and ability distributions were generated by WinGen3 software. The results show that the assumption of the equality of psychological distance between categories of rating scale was not provided in any different ability distribution and different category counts conditions. However, the number of categories influenced the psychological distance between categories, particularly for the 7-point scale. As the number of categories increases, the deviation amount from the conventional category value also increases. Also, endpoints of scales tend to close to middle point of scale when the number of categories is increased. When the converted scale values of the cases with the different ability distribution characteristics were compared, it was seen that the deviation from the conventional category value slightly varied in all the number of categories. However, these differences did not have a systematic order. The degree of violation of the assumption increases as the number of categories increases.

Keywords

• interval equality
• number of categories
• ordinal response scale
• rating scale