Saturday, September 23, 2006

Degree...required?


One of the most common forms of screening down candidates is a college degree requirement. I've heard more than one hiring manager say, when faced with a mountain of applications, "let's just see who has a degree."

This is a classic example of what I call "lazy assessment." Yes, in some cases a specialized degree is required (e.g., nursing, attorney). But let's focus for a moment at the vast majority of situations where a college degree isn't really required, it's just a quick and dirty way of funneling down candidates.

What might be some issues?

1) You could inadvertently be leaving out sizable numbers of folks from different backgrounds. For example, looking at
national statistics , requiring even a high school diploma is likely to have adverse impact against people of Hispanic descent, and requiring a bachelor's degree is likely to have adverse impact against African Americans. This will of course vary depending on the job and the relevant labor market, which is why it's so important to track your applicant flow statistics.

2) If you do end up having adverse impact in your process, and someone calls you on it (read: lawsuit), can you defend this decision? Do you have data showing a significant link between the degree requirement and job performance? Do you even want to go there?

3) Education, in general, does not predict job performance very well. In the most rigorous study of this issue to date,
Schmidt & Hunter (1998) found a correlation of .10 between years of education and job performance (see sidebar for neutered version of this article). This was exactly the same value they found for how well measures of interest predicted job performance. This means asking people how interested in a job they are is about as good as choosing people based on how much education they have--and neither is very good.

4) Education is a very blunt measure of competencies. Many times hiring managers assume if someone has a degree, they must have decent writing skills, analytical skills, etc. Wrong. All it means is someone sat their duff in a chair long enough to be awarded a degree. Without knowing the school attended, GPA, and courses, you know almost nothing about what someone got out of their degree.

5) Falsifying degrees is common. So you may leave out the person who hasn't had a chance to get a degree, or is in the process, in favor of someone who happens to know where to get a fake degree. Nicely done.

What about substitutions (e.g., college degree or X years of relevant experience)? Two points: (1) according to the same study, years of experience in general does not predict performance well either--again, blunt measure of competencies here folks; and (2) make sure you have a rationale for the number of years. Don't go with 2 years because that sounds about right--get subject matter expert input.

So what do we do in situations where there are large numbers of applicants and we need some way to screen down? First, make sure your job description is very clear about required KSAs/competencies and what you expect them to show you day one--this, along with realistic job previews, is probably the most cost effective way you can ensure truly qualified and motivated people are applying. Two, written tests (job knowledge, biodata, etc.) are easy to administer to large groups and have a good track record. Three, on-line tests are even easier to administer; if unproctored, just re-test folks when they show up so you know it was really that person taking the test. Four, if you're using an on-line application process, allow folks to easily select what qualifications they have (e.g., checkboxes) and use this to quickly cut your numbers down.

Those are just some ideas. The point is don't just automatically throw a degree requirement out there without thinking about unintended consequences and injecting a little creativity into things. We can, and should, do better.

No comments: