50 years ago today, July 2, 1964, President Lyndon Johnson signed the Civil Rights Act into law. The new Act prohibited segregation and discrimination in all public places including hotels, restaurants, theaters, sports arenas, and parks. It also made it illegal for employers to make employment decisions on the basis of race, color, religion, sex or national origin and established the Equal Employment Opportunity Commission (EEOC) to enforce the law. The Civil Rights Act was first proposed by President John F. Kennedy on June 11th, 1963.
Many recent news posts reveal that some employers are beginning to use SAT scores as “factors” in the hiring process. Examples include:
While arguments abound regarding the benefits of using SAT scores as indicators of future job success (especially for jobs dealing with vast volumes of complex information), is this a safe practice for employers from a Title VII (test discrimination) perspective? Another important question is: “Do SAT scores provide more useful information on applicant quality than other measures?” Let’s take these questions one at a time…
First, the Title VII liability issue. When it comes to test discrimination liability, Title VII is quite straight forward: If an employer uses a “practice, procedure, or test” (or “PPT”) that results in success rates that are significantly different between groups, such PPT is actionable (by either plaintiff or federal regulatory groups) under Title VII. This means that the employer can be taken into court and required to “make a demonstration” that their challenged PPT is “job related for the position in question and consistent with business necessity” (1991 Civil Rights Act, Section 703[k][A][i]). Because SAT scores might not directly map to concrete and observable skills and abilities that are needed for many jobs, making such a “job relatedness demonstration” would rely on either “construct” or “criterion-related” validity defense under the current federal testing regulations (the “Uniform Guidelines,” 29 CFR 1607).
Both of these techniques would require the employer to prove that SAT scores were statistically significantly correlated with meaningful aspects of job performance. In other words, SAT scores would have to directly translate to job performance in a statistically meaningful way. While conducting such a study might prove this out, the employer is still taking a substantial risk!
Note that such a case would also depend on just how the SAT scores were used—whether they were used in a fixed way or if they were only “soft considerations” combined with many other data points. Further, employers would be challenged to defend SAT scores if they were used by different hiring managers in different ways.
Second, do SAT scores provide more useful information regarding applicant quality than other measures? The short answer is “not likely.” This is because human performance is made up of a myriad of different factors, including smarts, personality traits, drive, experience, discipline, background, motivation, and the list goes on. Further, each of these performance ingredients is needed in different levels, and in some cases different ways, for various positions. That’s why the best selection process is always one that “wraps” various PPTs around the specific job by profiling the job using a thorough job analysis. Only then can the employer be sure that the most important success factors are being properly measured and weighted in a selection process.
Our firm has developed hundreds of custom PPTs for hundreds of employers. We’ve also defended several tests in litigation settings. If I had to choose between a quick and easy measure like SAT scores (perhaps coupled with an interview) and a robust selection system that considered several different competencies and weighted them according to the specific job needs, I would choose the latter every time.