Education Opinion

Psychometric vs. Legal Discrimination

By Walt Gardner — February 20, 2013 3 min read
  • Save to favorites
  • Print

The case of Ricci v. DeStefano that was decided by the U.S. Supreme Court in 2009 has special relevance for today’s accountability movement. Although the case was correctly covered by the media as a dispute over the use of test scores as the basis for promotion in the workplace, it’s instructive to recall the high court’s ruling 38 years earlier in a case involving the use of test scores as the basis for hiring. Both stem from the same federal law, and reveal a fundamental misunderstanding of the principles of testing.

In 1971, in Griggs v. Duke Power, the court held that even a race-neutral employee test is illegal if more minorities than white applicants for a job fail, unless an employer can prove it has a “business necessity” for using the test. It made clear that testing for very specific skills is legal, but that testing as a means of evaluating candidates for their overall trainability is not. That’s an important distinction given short shrift in the coverage of Ricci v. DeStefano.

Since then, employers have been understandably reluctant to administer tests at all, no matter how devoid they are of items having even the remotest possibility of being challenged in court, nor how valid they are for a particular position. Instead, they’ve increasingly relied on educational credentials, usually in the form of a college degree.

The rationale is that possession of a degree signals that the holder possesses persistence, discipline and intelligence. For many jobs, these traits are seen as a surrogate for the particular knowledge and skills an employer seeks. But they don’t allow comparisons to be satisfactorily made for specific openings. In fact, a bachelor’s degree can mean many different things today, even from accredited schools. Moreover, employers still run the risk of being forced to prove that a college degree has anything at all to do with satisfying the “business necessity” requirement.

Charles Murray in Real Education recognized the reductio ad absurdum of this trend when he wrote: “The advantage conferred by the BA often has nothing to do with the content of the education. The employer does not value what the student learned, just that the student has a degree.” In short, the requirement for a college degree is an end run around the law.

This brings the issue back to Title VII of the Civil Rights Act of 1964, which forms the basis for both Ricci v. DeStefano and Griggs v. Duke Power. It requires that any tests used must not result in disparate outcomes between racial and ethnic groups. While this mandate has great appeal to those who believe in equity in the workplace, it is an impossibility from a psychometric viewpoint. The truth is no matter how well constructed, no test can ever guarantee scores that reflect similar natural patterns of distribution among all groups. Try as they might to engineer results to comply with the court’s order through the use of race-neutral instruments that measure what they purport, designers will always fall short because some test takers have higher innate intelligence, and some prepare themselves better than others. Misunderstanding arises when group averages and individual performance within the group are equated.

Teachers have long known that the fairest way to assess students is by providing all of them with the same opportunity in class to develop the broad body of skills and knowledge to be measured on a test, and then to administer a test that samples as many of these standards as time allows. Teachers are not responsible for conditions beyond their control in the form of their students’ innate intelligence, their socioeconomic backgrounds, nor their willingness to work hard. By the same token, employers need to be granted the same leeway to develop tests that measure what they believe are needed for success on the job. They know this best from experience. They, too, are not responsible for factors beyond their control in test outcomes.

In Ricci v. DeStefano, all candidates vying for promotion to lieutenant and captain in the New Haven Fire Department were given the same time to prepare for the test, and were given a list of the same books. But only 19 whites and two Hispanics qualified. No blacks did. When the New Haven civil service board refused to certify the results, firefighters who passed sued on the basis of discrimination. The U.S. Supreme Court by a 5-4 vote held that New Haven’s decision violated Title VII of the Civil Rights Act of 1964. Although the decision was based strictly on law, it also has implications for education by acknowledging the reality of testing.

The opinions expressed in Walt Gardner’s Reality Check are strictly those of the author(s) and do not reflect the opinions or endorsement of Editorial Projects in Education, or any of its publications.