Pre-Employment Testing and Adverse Impact
Dr. Patel is an I/O psychologist retained by a Nebraska manufacturing company with 380 employees. The company wants to fill 40 maintenance-technician openings and has 240 applicants. Dr. Patel recommends a 45-minute cognitive-ability test and a mechanical-comprehension test. After testing, the company proposes a cut score at the 80th percentile because supervisors believe 'the best candidates score high.' Results show 100 white applicants: 72 pass; 50 Hispanic applicants: 24 pass; 40 Black applicants: 22 pass; 50 women: 30 pass. The company also wants to use the same tests to deny promotions to current employees and to keep the answer keys in an unlocked office. The test publisher's manual reports validity evidence from 1987 using 112 utility workers in another state; no local job analysis has been done. The company asks Dr. Patel to 'just lower the Hispanic cut score' to avoid a disparate-impact claim. Dr. Patel must advise on validation, adverse impact, ADA, test security, and ethics.
1. Calculate whether the proposed cutoff would create adverse impact under the Uniform Guidelines' four-fifths rule, and explain what the numbers mean.
Read the worked answer
Selection rates are: white 72/100 = 72%; Hispanic 24/50 = 48%; Black 22/40 = 55%; women 30/50 = 60%. Under the four-fifths rule, compare each group's rate with the highest rate: Hispanic 48/72 = 0.667; Black 55/72 = 0.764; women 60/72 = 0.833. The Hispanic and Black ratios fall below 0.80, showing adverse impact; the women's ratio does not. This does not prove illegal discrimination, but it triggers a duty to justify the cutoff and consider alternatives. The proposed 80th-percentile cut score is likely to increase adverse impact. Dr. Patel should not create different cut scores by race; that would be unlawful and unethical. Instead, the company needs a job-related, validated selection system.
2. What validation and legal steps should Dr. Patel recommend before using the tests for selection and promotion?
Read the worked answer
Dr. Patel should start with a job analysis for maintenance technician and for promotion to identify essential knowledge, skills, and abilities. Then he should gather local validity evidence-criterion-related validity against supervisor ratings or work samples, or content validity for the mechanical test-and reliability data. The 1987 out-of-state utility-worker study may support transportability only if the jobs and workforce are comparable; it is not enough by itself. Cut scores must be justified by expected job performance and documented. He should evaluate alternative predictors with less adverse impact, such as work samples, structured interviews, or mechanical aptitude plus training. Under Title VII, the employer must show the tests are job-related and consistent with business necessity. Under the ADA, no pre-offer medical exams; testing must be accessible and accommodate disabilities. Test security must be fixed.
3. How should Dr. Patel handle the unlocked answer keys and the request to use the tests for current employees' promotions?
Read the worked answer
The unlocked answer keys are a serious security breach. If applicants or employees can see the keys, scores are compromised and the test's validity is undermined. Dr. Patel should secure all materials immediately, notify the publisher, determine who had access, and use an alternate form or re-test if needed. He should document the breach and remediation. Using the same tests for promotions requires the same validation and adverse-impact analysis, plus attention to incumbent employees and seniority systems. He should not supply keys to supervisors and should train HR on test security. Ethically, APA Standards 9.02, 9.06, and 9.11 require appropriate test use, accurate interpretation, and maintaining test security.