Multidimensional scaling of noncognitive assessments: Application to the Situational Test of Emotional Understanding
The goal of this study was to use multidimensional scaling (MDS) to provide specific feedback to respondents on the Situational Test of Emotional Understanding (STEU). The STEU and the Mayer-Salovey-Caruso Emotional Intelligence Test (MSCEIT) were administered to 44 Fordham University Law School students and seven emotional understanding experts. Two of the seven experts were assigned to the low emotional understanding group where they were instructed to complete the assessments as someone with low emotional understanding. The remaining five experts were assigned to the high emotional understanding group where they were instructed to complete the assessments as someone with high emotional understanding. Thirty of the 44 law students returned to receive their STEU and MSCEIT score reports. They were asked to compare the specific feedback generated from the MDS on the STEU report with the general feedback on the MSCEIT. ^ The study findings revealed that there is no support for the validity of a non-metric MDS model application to the STEU, meaning the distances between the emotions does not appear to reflect the respondent's emotional understanding ability. This was demonstrated by three major findings. The two experts in the low emotional understanding group both had poor STEU and MSCEIT conventional numerical scores, yet had very different scaling results on the STEU. The emotions in one expert's configuration were relatively scattered, while the other expert's configuration had three clusters. The correlations between the Euclidean distances of the configurations for the experts were high regardless of whether they were in the high or low emotional understanding group. There was also no relationship between students who received high or low MSCEIT scores with the correlations to experts in the high or low emotional understanding groups. However, moderate correlations between the conventional numerical MSCEIT and STEU scores were found, which were consistent with previous research. Finally, out of the thirty students who returned to receive their score reports, 50% preferred the specific feedback of the STEU score report in comparison to the generic feedback of the MSCEIT score report. Although there appears to be no validity in the application of a non-metric MDS model to the STEU, it was suggested that favorable findings might result with the application of an individual differences MDS model. If validity support was found with the application of a different MDS model, then MDS could be extended beyond the STEU to other noncognitive assessments and beyond law students to other higher education populations.^
Statistics|Psychology, Psychometrics|Psychology, Cognitive
Allen, Veleka, "Multidimensional scaling of noncognitive assessments: Application to the Situational Test of Emotional Understanding" (2012). ETD Collection for Fordham University. AAI3560051.