Collect data for your specific adverse impact calculation. This data is, in short, a list of individuals applying for a position with a given employer. You should be able to clearly divide the groups so ethnic, gender or age differences are clear. The values needed are the applicants from each group to be analyzed and whether they successfully gained employment with the organization.
Calculate the proportions of successful applicants from each group. For each group, divide the number of applicants who successfully gained employment by the total number of applicants.
Establish the highest proportion calculated as your benchmark. For example, 46 percent of the applicants in the 18-to-30 age group were determined to have been hired by the company. Since it is the highest percentage, use this number in future calculations.
Calculate adverse impact. Divide all of the other proportions by the benchmark. For example, the 31-to-45 and 46-to-60 age groups had 41-percent and 21-percent acceptance, respectively. Divide these numbers by 46 percent to yield rounded figures of 0.89 and 0.46.
Compare the numbers calculated to 0.80 (80 percent). According to Dan Biddle in "Adverse Impact And Test Validation: A Practitioner's Guide to Valid and Defensible Employment Testing," the U.S. Department of Labor states any numbers less than 0.80 serve as evidence of discrimination. For the example, the 0.46 calculated for the 46-to-60 age group is significantly lower than 0.80, showing an organizational tendency to discriminate against older applicants.