What differentiates the elite best-of-the-best from others in their respective fields?
This turns out to be a somewhat tricky question. There is ongoing debate about what the distribution—the actual full range of performance outcomes across people—looks like.
Researchers and industry practitioners who have studied this are usually divided between one of two conclusions:
- Some have found that performance is normally distributed—the average level of performance is also the most frequent, and there is relatively equal variation above and below.
- Others have found skewed distributions—the statistical average is actually higher than the most frequently observed levels of performance, because the highest performers exponentially outpace the performance levels of their colleagues.
Which one of these is true?
They both are, in their own way. For better or worse, it depends on how we define and measure performance.
1. Performance vs. performance outcomes
If we focus on behavior—what people actually do when they are performing—the distributions tend to be normal.
If we instead focus on outcomes—how people are perceived by others, rewarded, or compensated—a skewed distribution is much more likely.
To the extent that elite performers also do many of the same things that others are actually capable of doing, skewed distributions are probably a reflection of untapped talent and potential.
2. Single vs. multiple tasks
Some people are particularly good at a specific type of thing. If we only look at outcomes related to such a thing, it is likely that we will see a skewed distribution with a few elite performers at the top.
If we focus on multiple tasks—the full variety of things people can actually do—the distributions tend to be relatively normal.
Unless a specific focus on one type of task is really justified, it is likely that skewed distributions are actually missing out on other relevant areas of performance. Many people might be just as “elite” as those recognized, only in a variety of different ways.
3. Partial range vs. full range of people
It is often the case that we are simply unable to study the full range of people (and how well they would perform). Much of the issue is due to the fact that we only have information about those who were actually hired or accepted in the first place.
Ideally, at least part of the reason for a skewed distribution is that the people who were not hired or accepted would have also performed worse than those who were (thus, if we actually had the full range of information, we would instead see a more normal distribution).
Unfortunately, many of the methods commonly used to make hiring and acceptance decisions are not very accurate. The more that these decisions are left up to chance, the more likely it is that average performers will be selected at a higher rate (because there are simply more of them). A skewed distribution can just as easily reflect a selection process that tends to overlook potential high performers.
The Candidate Assessment Platform by Cangrade is an easy-to-use online tool that accurately predicts performance on the job. Give us a try if you’re concerned about missed opportunities and untapped potential.
Image credit: Simon Webster