October 26, 2011
In my Recruitment and Selection class at DePaul University, I am currently in the process of planning our fourth lesson on measurement in HR selection. We will be covering concepts like criterion and predictors of on-the-job success, measurement scales (nominal, ordinal, interval, and ratio), and validity. Of course, this got me thinking about the GMAT as a selection tool designed to predict academic success in one’s first year of a master’s degree in business administration.1 As I did some digging around, I uncovered some alarming information about misuses of GMAT scores. Namely, employers are using these scores as employment selection tools.
In order for a selection mechanism to be a valuable predictor of success, it must have empirical proof in the form of correlated data as to what attributes it is specifically designed to predict. The GMAT has that data regarding its use by academic institutions to predict academic success. There is absolutely no evidence for the GMAT to be used as a predictor for professional success, much less across an infinite array of occupations.
On the face of it, one can quite easily swallow the rationale: if you are a top scorer on a test designed to discern potentially successful MBA candidates, then it stands to reason that you are the type and quality of individual that a top company would be happy to recruit and, ultimately, select as one of its own. Intuitively, it makes easy sense. Here’s the problem: intuition is not enough.
In the 1971 Supreme Court case Griggs v. Duke Power Co., a very important precedent was set. Duke Power had been using a high school diploma and a written test as selection tools for supervisory positions within the company. Management is on record as saying these tools were used in order to increase the overall quality of the workforce. That’s reasonable, right? I mean, what’s wrong with requiring their incumbent supervisors to have a high school diploma and to successfully complete a written examination? Intuitively, we all get it. However, the Supreme Court ruled against Duke’s employment practices.
There are some significant social issues that are wrapped up in this court case, but those matters are not within the scope of our discussion here. However, the precedent regarding selection tool validity that was born from the ruling is definitively within our focus. In short, the high court said that companies are only allowed to measure the individual for the job, not the individual in the abstract. This means that whatever selection mechanisms an organization employs must be directly and empirically connected to job-related attributes revealed by a thorough job analysis of a given position.
The use of the GMAT as such a selection tool is woefully inappropriate. Sadly, though, because it appears to make such intuitive sense, I suspect the trend in its use by organizations looking to hire from a vast and varied candidate pool will only continue to rise. That is, until someone gets sued.