Why It's So Hard to Assess Employees
A "fundamental attribution error" often leads HR leaders to assume that poor performers are just bad apples, because it is difficult to consider that circumstances might be contributing to their subpar performances.
By Peter Cappelli
Almost everyone in HR has had the experience of dealing with an executive who has his or her own special technique for selecting applicants. It's usually something that has absolutely no validity -- "I watch what they order as an appetizer" -- and we roll our eyes while listening.
The problem in that situation is that people are often afraid to tell the executive that what he or she is doing is nuts, so the executive keeps doing it. No one bothers to collect evidence about hiring outcomes that would show it doesn't work.
The challenge is greater when it comes to assessing current employees. The hot topic at the moment is performance appraisals, whether they really are valid and whether we should dump them for something else (more on that next month). The next hottest topic is assessing an employee's potential for bigger roles, as more companies are thinking about whether they should try to develop more of their talent from within. These two tasks are difficult because prior research doesn't give us a lot of clear guidance. Performance appraisals are difficult because we rarely have better measures of performance against which to validate our appraisal process. In assessing potential for advancement, we do have measures of performance, but almost no companies use them to validate their assessments.
Here's why this is so unfortunate, because it means that decisions can be heavily biased by "The Fundamental Attribution Error." Isn't that one of the best terms social psychology ever produced? What it means is that we have a tendency to explain or attribute the behaviors of individuals to who they are as people. That might include attributes such as character, intelligence or other dispositional factors. We are also inclined to ignore or at least play down the role that circumstances might play in explaining their behavior.
The classic study demonstrating the fundamental attribution error took place in the mid-1960s when Fidel Castro was very unpopular in the United States. Students here were asked to read statements about Castro, one very positive and the other very critical. It's not surprising they liked the latter better, and they also thought better about the author of that piece than of the author of the former piece.
In the second part of the experiment, a different group of students were given the same two pieces, but this time were told that the two authors were told what to write, one preparing a positive piece, the other a critical piece. In this situation, they still thought the author of the critical piece was a better person than the author of the positive piece, despite knowing that there was no reason to expect that those papers reflected the actual views of the authors. That's the fundamental attribution error.
The most-common example of this error is to ask ourselves what we think about the person who raced past us on the shoulder of the road while traffic was otherwise stopped. Do we assume that driver is a jerk? Yes. Do we consider that he or she might be rushing to an emergency? No.
National cultures seem to influence the extent of the fundamental attribution error, with it being a bigger issue in more individualistic countries such as the United States.
But back to performance appraisals. The fundamental attribution error leads us to assume that poor performers are just bad apples. It is difficult for us to consider that circumstances might be contributing to their poor performance. There may be other factors at work as well, but the fundamental attribution error would lead us to support forced ranking systems, especially rank-and-yank systems that fire the worst performers: We know they are bad apples, so just get rid of them. And if they are bad this year, they will always be bad.
I have been truly puzzled by how sure some people are that the workforce can be divided into "A" players -- as Jack Welch put it -- who are always good, and "C" players, who are always bad. If you ask whether they have actually seen evidence for this -- say, from performance appraisals over time -- the answer is always, "no, but it's obvious."
When assessing potential, the fundamental attribution error can make us biased toward an initial good impression: If we see a good performance right off the bat, we are inclined to think that person has the right stuff, the native ability and disposition to go all the way. That helps create self-fulfilling prophesies so that, in the next period, we give those initial good performers a break, which means they get rated higher than they should. The whole notion that the key to picking candidates for advancement should be based on some idea of who they are as a person rather than what they have been able to accomplish sounds an awful lot like an attribution error.
To make the obvious point, this error causes employers to waste talent, undermine meritocracy and, on a small level, create injustices in the workplace.
Can it be treated? Yes, and the treatment starts with a recognition that the problem exists.
Peter Cappelli is the George W. Taylor Professor of Management and director of the Center for Human Resources at The Wharton School of the University of Pennsylvania in Philadelphia. His latest book is "Will College Pay Off? A Guide to the Most Important Financial Decision You'll Ever Make."