Utilizing ordinal alpha reliability on modification of Mark H. Davis’s empathy rate instrument (case study in Department of Mathematics FMIPA UI)

Abstract

Most psychology research requires the existence of an instrument to obtain the required information. An instrument using Likert scale needs to know its reliability and validity. However, the use of an instrument is not necessarily suitable for particular population. This can be seen from the fulfilment of reliability, construct validity, item validity, and content validity of the instrument. Unreliable and invalid instruments due to cultural differences in the population. Therefore, a modification of the instrument is needed such that better information can be obtained to the relevant population. Because in this study using ordinal scale data, then the reliability will be measured by the coefficient ordinal alpha with polychoric correlation. In this study, modifications will be made to Mark H. Davis’ rate empathy instrument consisting of 45 indicators into an instrument consisting of only 15 indicators. Modified measuring tools proved reliable, have valid content, valid constructs, and valid items. The methods to be used are factor analysis, Spearman correlation, polychoric correlation, and ordinal alpha reliability.