Make sure your physical tests gauge realistic demands of the job — Business Management Daily: Free Reports on Human Resources, Employment Law, Office Management, Office Communication, Office Technology and Small Business Tax Business Management Daily
  • LinkedIn
  • YouTube
  • Twitter
  • Facebook
  • Google+

Make sure your physical tests gauge realistic demands of the job

Get PDF file

by on
in Human Resources

Any tests you use to screen applicants should relate to the job and you must be prepared to prove that they do. If you can't, and a protected group of workers (e.g., women, minorities) tend to score poorly, you're just asking for a lawsuit. That's particularly true with tests of a candidate's strength, which can often eliminate women unfairly.

To see if your testing is vulnerable, look at your work force's composition. Do fewer women pass your physical endurance test, resulting in a low number of female employees? How does your male/female ratio compare with other employers in the same field? Make sure any test realistically reflects the job's essential functions.

Recent case: The city of Erie, Pa., uses tough physical testing to screen police officer applicants, including a certain number of push-ups and sit-ups within 90 seconds.

More than 87 percent of female applicants failed Erie's test over a 10-year span, compared to 29 percent of male applicants. As a result, just 4 percent of Erie police officers are female. Police departments in Pittsburgh and Philadelphia have about 25 percent women.

The court said these facts prove that Erie's test discriminates. Since Erie couldn't articulate a clear reason for the 90-second test, the court concluded that it had none. (DOJ v. City of Erie, 3rd Cir., 2005)

Final tip: If you use a test administered or designed by a third party, be sure that the administrators explain how it relates to the job and why it's necessary. Before signing on, ask for data on who passes and who fails, and look for possible discriminatory patterns.

Leave a Comment

Previous post:

Next post: