September 28, 2014

LESSONS LEARNED FROM THE OFCCP SETTLEMENT WITH GENERAL ELECTRIC

On September 12th, a General Electric subsidiary in Ohio agreed to pay $537,000 to settle a sex discrimination allegation with OFCCP. The agency alleged that the company used a set of employment tests that produced adverse impact against female applicants to attendant positions and was not validated in accordance with the Uniform Guidelines (41 C.F.R. part 60-3). The employment tests, an off-the-shelf set of tests called WorkKeys, measure a series of cognitive abilities, including applied math, locating information, reading for information, and applied technology. According to OFCCP, test content did not adequately match job content and the test cut score was not related to performance differentiation, and as such the requirements of the Guidelines were not met.

The OFCCP press release also noted that the agency settled another case back in 2011 that focused on WorkKeys.  That 2011 settlement was with Leprino Foods, and the set of tests were alleged to have adverse impact against minority applicants to laborer jobs, which is an allegation that is generally consistent with the personnel selection research literature assessing subgroup differences on cognitive tests. Once again, OFCCP alleged that the validation evidence did not meet the requirements of the Uniform Guidelines.

This settlement is a reminder that OFCCP can allege and litigate unintentional discrimination under a disparate impact theory. In this scenario, any facially neutral step in the selection process may be challenged. The Uniform Guidelines, which were developed in 1978 and are jointly enforced by OFCCP, EEOC, and DOJ, require that employers justify any selection procedure that produces adverse impact by demonstrating that it is “job-related and consistent with business necessity”. This is often accomplished via a validation study that uses scientifically rigorous research methods and shows persuasive results that the tool allows for meaningful inferences about candidates to be made.

This settlement is a reminder for federal contractors to monitor their employment testing programs. Employment tests and other professionally developed selection procedures can be an important competitive advantage to organizations, but they can be challenged. We suggest that contractors keep in mind the following:

  • The Guidelines identify a number of employee selection procedures that could be challenged under a disparate impact theory, including:
    • Job requirements (physical, education, experience)
    • Application forms
    • Interviews
    • Work samples/simulations
    • Paper and pencil tests
    • Performance in training programs or probationary periods.
  • Any facially neutral step in a selection process can be evaluated for adverse impact via differential “pass/fail” results.
  • Be aware of what tests and other selection procedures are being used in your organization;
    • Consider an independent test audit to help you understand what tests are being used, whether they were professionally developed, are psychometrically sound, have been validated for similar jobs and are likely to produce adverse impact.
      • If the answer to any of the above questions is “I don’t know”, then there is likely potential risk.
  • If you are thinking about identifying and implementing tests or selection procedures in your organization, consider conducting a job analysis to identify what work duties are performed in a job and what worker characteristics are needed to perform that job well. Hopefully there are available tests and other selection tools to simulate those duties and/or measure those characteristics.
  • Should any adverse impact be identified, consider formal validation research conducted by an industrial/Organizational Psychologist or other measurement expert.

 

By: Eric Dunleavy, Ph.D., Principal Consultant, and Emilee Tison, Ph.D., Consultant at DCI Consulting Group

Categories:

DCI Alerts

Submit your email address to receive free industry news updates


Type the above number:

You have Successfully Subscribed!

Bitnami