A test is unlawful when it has a statistically significant negative impact, unless the employer can show that it is consistent with business necessity
It may cover ambiguous qualities like personality and work style, or basics like reading and typing skills-but if it's pre-employment test, you have the possibility of failing out legally. Tests can be unintentionally discriminatory-tipping the scales against women and minorities in the most subtle ways. If they do, you're in trouble-no matter how good your intentions. Howard Simon, a partner at San Francisco employment law firm Landels Ripley & Diamond, offers assessment guidelines for pre-employment testing.
What kinds of pre-employment tests can come under fire?
There are basically three kinds of tests: the content test, which would be like a typing test; the criterion test, which is general aptitude for the position in question, the kind of broad range of questions you'd ask somebody to see if he or she were qualified to be a police officer; the third is the construct test, which tests for the more ephemeral qualities like leadership or personality characteristics. The law covers any of these pre-employment tests.
When are tests unlawful?
When they have a statistically significant negative impact, in that the test works to exclude a protected category, like women and minorities, at a significant rate. If that's the case, the test is unlawful-unless the employer can show the test is job-related for the position in question and consistent with business necessity.
What is considered a significant negative impact?
Disparate impact in the most simple form is what's called "the Rule of 80."You take the group with the highest rate of candidate selection, and if the selection rate for protected individuals is less than 80 percent of that, that's disparate impact. A lot of these cases are fought on the issue of whether that's an appropriate measure, but that's the measure the EEOC uses, so it's certainly a good place to start.
What should employers do to protect themselves?
Under current law, if your test is shown to have a disparate impact, it's unlawful unless you can show the test is job-related and consistent with business necessity. Demonstrating job-relatedness is where you get into validity studies. These studies show that the test is a valid predictor of success on the job. Most people buy commercially prepared tests, so an employer needs to ask the provider of the test for their validity studies.
When can an employer rely on a vendor's validity study?
Three criteria must be met. The validation study must have been done in a workplace with similar job behaviors and a labor market similar to your workplace, and a study must be done to ensure its fairness regarding race and gender within the labor market. If you meet these three criteria, you can use somebody else's data, but if you miss any one of those three criteria, it would be very risky to rely on someone else's data.
Can you cite examples of the criteria?
Meeting the first two criteria means making sure the study's workplace and your workplace match well. …