DCI Consulting Blog

“Inside Look” at the 2007 ILG Conference Presentations on Employment Testing

Written by DCI Consulting Group | Sep 7, 2007 6:50:00 PM

by Eric Dunleavy, Ph.D., Sr. Consultant, DCI Consulting Group, Inc.

I attended a number of presentations at the recent 25th Annual Industry Liaison Group (ILG) Conference focused on employment testing. These sessions were targeted toward federal contractor employees who are responsible for developing, monitoring, and evaluating employment testing systems. Employment tests can be a true competitive talent advantage to organizations because, when developed and implemented adequately, these procedures often ensure that the most qualified candidates are selected and that the unqualified candidates are not. Of course, employment tests that have adverse impact must also meet the legal defensibility requirements of the Uniform Guidelines for Employee Selection Procedures (UGESP). In recent years the OFCCP has strengthened their efforts to identify systematic discrimination associated with employment tests, so these presentations were of interest to many conference attendees.

Dr. Lisa Harpe of Peopleclick Inc. and Dr. Dan Biddle of Biddle Consulting Group presented a session entitled ‘Testing the Test – Making Sure Your Employment Tests are Useful and Defensible’. These experts introduced the notion of adverse impact and summarized various validation strategies for employment tests, with particular emphasis on criterion-related and content validity evidence. The presenters then provided useful guidance regarding what organizations should consider before purchasing a test from a third party vendor. For example, what evidence should that vendor be able to provide to you immediately once you have expressed interest? As the presenters suggested, if the vendor doesn’t have validity evidence in the form of a technical report, test reliability information, or a clear understanding of which jobs the test has been validated for, potential test users should walk away. The presentation ended with an entertaining mock audit, where the experts were put ‘on the stand’ and demonstrated how validity evidence would be used to defend a test in the litigation setting.

Fred Azua, Regional Director of OFCCP’s Southwest and Rocky Mountain Regional Office, and Dr. Richard Fischer, OFCCP’s testing expert, presented a session entitled ‘The Importance of Testing Procedures to Achieving a Diverse Workforce’. This presentation focused primarily on the requirements of test users under the UGESP. Early on during the presentation some members of the audience were shocked to find that any procedure used to make an employment decision can be considered a test, including interviews. The presenters then provided useful guidance on what documentation the OFCCP would be interested in looking at during an audit if testing was identified as a problem. The list of documentation included but was not limited to the (1) test development process, (2) validation strategy including the correlation coefficient for criterion and construct validation evidence and the link between test content and job requirements for content validation evidence, (3) job analysis information, (4) standard setting procedure, (5) test reliability, and (6) consideration of less adverse alternatives. This informative session ended with a review of the requirements of the above issues, and stimulated interesting discussion between federal contractor representatives at the conference and the presenters from the OFCCP.

Employment testing is an issue on the forefront of the battle against systemic discrimination. In addition to these ILG presentations, the Employment Opportunity Commission (EEOC) held a public meeting last May on how employment testing relates to current EEOC initiatives, further suggesting that employment testing will continue to be evaluated by enforcement agencies. Given the current enforcement landscape, employers should have a clear understanding of what employment tests are being used in their organizations, what validation evidence exists for each, and whether these tests produce adverse impact.