A closer look at the measurement of dispositional reasoning: dimensionality and invariance across assessor groups
Despite the growing interest in dispositional reasoning as a construct and determinant of good raters ("good judges"), its measurement still requires attention. We address two measurement issues in the present study.
First, this study tests a hierarchical model as a more parsimonious account for dispositional reasoning than component- or general factor-models that were examined in earlier studies. So, this provides a more comprehensive test of the different measurement models underlying dispositional reasoning data.
Second, we assess the measurement invariance of dispositional reasoning measure scores across two different populations of assessors that are often trained and used in workplace assessments, namely psychology students (N = 161) and managers (N = 160).
Results showed that dispositional reasoning is well represented as componential in nature, with a higher-order construct underlying three lower-order components. A comparison of managers and psychology students through measurement invariance analysis showed relatively similar factor structures underlying dispositional reasoning scores across these groups, but metric invariance could be only partially established.
|Keywords||dispositional reasoning, judgment, interviews, accuracy, hierarchical CFA|
|Journal||International Journal of Selection and Assessment|
de Kock, F.S, Lievens, F, & Born, M.Ph. (2017). A closer look at the measurement of dispositional reasoning: dimensionality and invariance across assessor groups. International Journal of Selection and Assessment, in press. Retrieved from http://hdl.handle.net/1765/100023