Entire businesses have been built around how to build self-confidence. I agree that self-confidence can get you far in life and in your career; in fact, I’ve seen incompetentpeople succeed just on the basis of a high sense of self-esteem. They ooze confidence and can trick others into believing that confidence equals ability.

That brings me to a topic that has fascinated me for a long time: misplaced self-confidence. As someone who battled low self-esteem in the past, I’ve always found it fascinating (and even enviable) how some people think their skills are better than they are. I’ve always wondered if it’s just a highly honed defense mechanism or a form of delusion.

So you can imagine my interest in a report from Cornell in regard to self-evaluations used in the workplace as part of an overall employee work appraisal.

Researchers at Cornell found that the worst performers (in a variety of categories) often rated themselves and their performance, in most cases, far above average. But get this: Top performers rated themselves lower than their performance merited.

Here’s what the researchers reasoned:

The reasoning for these behaviors is fascinating. Poor performers lack the skills to perform–which are the same skills required to evaluate their performance. They don’t understand that they don’t understand, and so believe their abilities compare positively to their peers.

On the other hand, Top performers incorrectly assume that their competence is shared among their peers–leading them to rank themselves lower than they deserve.

Given this information, why in the world would any company use a self-evaluation as part of an employee’s review? To me, it would only lead to confusion. A manager looking at the low-rated self-review of someone she considers to be a high performer might think there’s something she isn’t aware of and change her own opinion of that employee. And the poor performer who ranks himself high may cause a manager to question her observations about him.

I’d like to hear from managers out there or even HR folks who can defend the use of employee self-evaluations. In the meantime, here’s the report from Cornell.