Recent research in reducing adverse impact in personnel selection has focused on the use of various weighting schemes to balance levels of adverse impact and the validity of selection processes. De Corte Lievens & Sackett (2007) suggested the use of the normal boundary intersection method to create a number of weights that optimize adverse impact and criterion validity. This study seeks to improve the efficacy of this solution by looking at specific types of performance, namely task and contextual performance. It will investigate whether a focus on contextual performance will improve the trade-off by requiring smaller losses in validity for greater gains in adverse impact. This study utilized data from 272 applicants for exempt positions at a multinational financial institution. The two sets of Paraeto optimal composite were developed, one based on contextual performance and the other based on task performance. Results were analyzed based on levels of adverse impact and validity of weights generated using each method. Results indicate that reducing adverse impact required a greater validity trade-off for task performance than contextual performance. Application of this method would allow for greater reductions to adverse impact than the original method while retaining a validity coefficient of 95% of the maximum achieved with regression weighting. Though this method would limit practitioners to selecting based on contextual performance, the use of minimal cut-off scores on task predictors or job experience could allow employers to incorporate task measures while further reducing adverse impact
| Identifer | oai:union.ndltd.org:ucf.edu/oai:stars.library.ucf.edu:etd-2687 |
| Date | 01 January 2010 |
| Creators | Tsang, Howin |
| Publisher | STARS |
| Source Sets | University of Central Florida |
| Language | English |
| Detected Language | English |
| Type | text |
| Format | application/pdf |
| Source | Electronic Theses and Dissertations |
Page generated in 0.0026 seconds