Accession Number : AD1033653


Title :   Cross Validated Temperament Scale Validities Computed Using Profile Similarity Metrics


Descriptive Note : Conference Paper


Corporate Author : ARMY RESEARCH INST FOR THE BEHAVIORAL AND SOCIAL SCIENCES FORT BELVOIR VA FORT BELVOIR United States


Personal Author(s) : Legree, Peter J ; Kilcullen,Robert N ; Young,Mark C


Full Text : https://apps.dtic.mil/dtic/tr/fulltext/u2/1033653.pdf


Report Date : 27 Apr 2017


Pagination or Media Count : 22


Abstract : Personality and temperament scales are used in employment settings to predict performance because they are valid and have minimal adverse impact. This project investigated the use of profile similarity metrics (PSMs) in place of conventional distance-based indices to develop scale and composite scores for a battery of temperament scales. Using a sample of 5,191 ROTC cadets, we computed the following PSMs for six temperament scales: the shape of each respondents rating profile relative to the key, rx,k; the difference in elevation between each respondent rating profile and the key, (Xmean Kmean); and profile rating scatter, sdx2. We then used regression procedures to develop optimally weighted PSM-based scores for each temperament scale and for the battery. Using a second sample of 5,720 ROTC cadets, we cross-validated the PSM scale and composite scores. Analyses documented that the cross-validated PSM scores maintained higher criterion validities for five of the six temperament scales. Furthermore, the cross validated battery composite based on the PSM scores had higher validity than the corresponding composite based on conventional scores (r = .41 vs. r = .32). These results demonstrated that PSMs can be used to increase scale validity of temperament scales against important performance criteria. Presented at the SIOP Conference held in Orlando, FL., April 27 2017 - April 29 2017.


Descriptors :   behavior and behavior mechanisms , surveys , bivariate analysis , motivation , personnel selection , military research , personality , physical fitness , training , statistics , algorithms , social sciences , regression analysis


Subject Categories : Personnel Management and Labor Relations


Distribution Statement : APPROVED FOR PUBLIC RELEASE