SOME BACKGROUND ABOUT EMPLOYER REQUIREMENTS

There are two types of illegal employment discrimination:

1. Disparate Treatment - Treating an individual, or several individuals, different from other people based on their membership in a protected class under Equal Employment Opportunity laws. Protected classes include such things as Race, Color, National Origin, Religion, Sex, Pregnancy, Mental Disability, Physical Disability, Veteran Status and other groups designated by state laws.
2. Disparate Impact - Using employment selection devices, such as written tests and interviews, that appear reasonable but actually have a biased impact on protected groups. Such devices screen out one or more groups at rates that exceed legally acceptable limits.

Adverse Impact is discrimination normally indicated by statistical analysis of an employer's decisions about job applicants and employees who will be selected for job placement, compensation increases, and so forth. If women are treated significantly different from men in hiring for technical jobs, for example, statistical analysis of placements by sex may show those data indicating illegal levels of discrimination.

Any employer with 15 or more people on the payroll is subject to most federal equal employment opportunity laws. Some state laws have lower headcount requirements before employers are required to abide by similar protections.

The only way you can know if you are illegally discriminating because of an employment practice that seems to be non-discriminatory is to conduct statistical tests on employee movement data. This is a requirement for all federal contractors. It is also a requirement for all employers with 15 or more workers on the payroll according to federal regulations. (41 CFR 60-3.2(A))

Adverse Impact testing is a complex process that most of us are ill-equipped to slog through on our own. If it has been a while since your last statistics class, you may have forgotten what formulas to use for statistical significance and probability testing.

If that's the case, you don't have to be concerned any longer.

The Adverse Impact Toolkit is a statistical program designed to analyze hiring and promotion processes for possible adverse impact. The program operates within a Microsoft® Excel® Workbook to maximize versatility and make data entry simple.

The program does not require advanced statistical or EEO training to interpret the results. After data entry, the program color codes the output to reveal areas of concern ("disparate impact" or "adverse impact" findings). In addition, complex statistical calculations are reduced to simple outputs (e.g., "the odds of the passing rate difference is 1 chance in 47").

TYPES OF ANALYSIS THIS SOFTWARE WILL DO FOR YOU

There are four types of analysis Adverse Impact Toolkit will complete for you.

• Selection Rate Comparison
• Availability Comparison
• Combined Data-Passing Rate Comparison
• Combined Data-Availability Comparisons

Selection Rate Comparison

The Selection Rate Comparison evaluates the passing rates between two groups (e.g., men and women) on a practice, procedure, or test. This is the most conventional type of analysis for making adverse impact determinations, and is modeled after Section 4D of the Uniform Guidelines on Employee Selection Procedures and numerous court cases. The specific tests included are:

• 80% Test: This test compares the passing rate between two groups (e.g., men vs. women) using the 80% test for adverse impact specified in the Guidelines (a "violation" occurs if one group’s passing rate is less than 80% of the group with the highest rate).
• Statistical Significance Test (estimated, two-tail): This test uses a statistical procedure (the Hypergeometric Variance formula) to estimate the statistical probability of the observed passing rate difference between groups.
• Statistical Significance Test (exact, two-tail): The Fisher Exact Test is used to calculate the exact probability of the passing rate difference between two groups. Statistically significant values are calculated in the "exact" form and reveal the precision likelihood of the differences occurring by chance.
• Practical Significance Test: This procedure evaluates whether the statistical significance findings are practically sound, or if they are based on "barely significant" results that would become insignificant if just two more persons from the disadvantaged group would have passed the practice, procedure, or test. The adverse impact test is taken from Waisome v. Port Auth., 948 F.2d 1370, 1376 (2d Cir.1991) and U.S. v. Commonwealth of Virginia, 569 F2d 1300 (CA-4 1978), 454 F Supp. 1077.

Availability Comparison

This program compares one group’s representation (e.g., percentage of women in a job group) to their availability in the qualified labor market. Statistical tests are used to indicate significant areas of under-utilization. Initially used in Hazelwood School District v. United States, 97 S.Ct. 2736 (1977), this adverse impact test has multiple applications in EEO analysis and litigation. The specific tests included are:

• Binomial Probability (estimated, two-tail): This test uses an estimator formula for the binomial probability distribution to evaluate the statistical significance of the group’s utilization rate.
• Binomial Probability (exact, two-tail): This test calculates the exact probability of the data set using precise calculations for the exact likelihood of the group’s utilization rate.

Combined Data-Passing Rate Comparison

This program combines "multiple events" (e.g., passing rates between two groups on a test over several years) or multiple data sets (e.g., divisions or job groups) to analyze adverse impact. The procedure involves the same type of statistical comparison as applied in the Selection Rate Comparison program, but allows for multiple data sets.

• Homogeneity of Odds/Rates Test (two-tail): Because combining multiple events or data sets can sometimes create statistical anomalies, two tests are included in the program to assess whether the multiple events or data sets can be combined for a valid overall analysis. These tests are the Breslow-Day (with Tarone’s Correction) and the Treatment by Strata Interaction Test proposed by Mehrotra. Both tests typically output similar results, and statistically significant findings indicate possible complications with combining all events into a single analysis.
• Statistical Significance Tests (two-tail): These analyses include two versions of the Cochran-Mantel-Haenszel procedure and a test called the Minimum Risk Weights Test to estimate the statistical probability of the observed passing rate difference between groups for the overall, combined data set. All three test typically output similar values, and one version of the Cochran-Mantel-Haenszel procedure typically produces values that very closely mirror the exact probability value.

Combined Data-Availability Comparisons

This program combines "multiple events" (e.g., one group’s representation over several years) or multiple data sets (e.g., divisions or job groups) to investigate statistically significant underutilization. The program includes the Generalized Binomial test to estimate the statistical probability of one group’s underutilization across years/departments/job titles/etc. Color warning indicators show areas of concern.

System Requirements

• Pentium III processor (or higher)
• 512 MB of RAM minimum (1 GB or more recommended)
• 5 MB of free hard drive space
• Super VGA monitor, 800x600 resolution minimum, 1024x768 recommended
• Microsoft Windows 2000/XP
• Microsoft Office 2000/XP/2003
• Excel must be configured with macros enabled

This product allows you to avoid the math. Expert statisticians have already done that work for you. All you have to do is input your data for the employment event(s) you wish to test. It's that simple.

HERE'S HOW THIS PRODUCT IS SOLD ...

You will receive a CD-ROM containing all the formulas and tests discussed above. They will be imbedded in formatted Excel spreadsheets so you have an easy visual display of your input and the output. It is also easy to print your results.

Each CD-ROM carries a license for unlimited use of its content on one computer (PC) at one location (site). You can analyze employee movement data from as many sites as you wish, as long as you do it on the one PC where your Adverse Impact Toolkit is installed. Installing the software on more than one machine will require purchasing additional licenses.