BETA
This is a BETA experience. You may opt-out by clicking here

More From Forbes

Edit Story

When It Comes To Assessing You, Who Knows You Best?

Following
This article is more than 9 years old.

Who understands you best? Who knows how well you perform? Who can most accurately predict your strengths and weaknesses? Some gurus contend that only you truly know you. They believe the best person to identify what you are good or bad at is you. After all, they argue, don’t you understand your motives, intentions and aspirations better than anyone else?

The temptation to mix personal intentions and actual behavior is one major downfall of any individual’s ability to accurately know themselves. After receiving a 360-degree feedback report, an executive in a workshop jokingly commented, “If only people rated me on my intentions! I would like to be viewed as genuine, kind, respectful, considerate and fair. Unfortunately, people rate me on my actual behavior.”

We thought it would be interesting to test how well most individuals are able to predict their overall capability. For our study, we measured 18,336 leaders on 49 leadership behaviors that we had determined were the most effective predictors of leadership effectiveness. For each leader, we gathered data from at least 10 raters, along with their self-rating. That meant we had ratings from a total of 267,116 managers, peers, direct reports and others (usually former colleagues or someone reporting to one of their subordinates). Based on their ratings of those 49 behaviors we calculated an overall effectiveness index for each leader.

We knew from our past research that this overall rating was an accurate predictor of numerous outcomes such as employee turnover, customer satisfaction, profitability, sales and employee engagement. In our view this overall effectiveness rating is a highly accurate predictor of a person’s leadership behavior.

We next compared these leaders’ overall rating to the ratings from each of the other rater groups. For this study, we had included the leaders’ self-scores in the overall rating, even though typically we exclude self-scores from total scores. That meant that the ratings from managers, peers, direct reports, and others (again, usually former colleagues or someone reporting to one of their subordinates), along with the subject leader’s self-scores all had exactly the same influence on the leader’s overall leadership effectiveness score. The closer each rater group came to the overall rating, the higher the correlation.  The more disparate the scores, the lower the correlation coefficient. *(see bottom note for statistical explanation)

The graph below demonstrates that the leaders’ rating of themselves accurately accounts for only 14% of the variance between their self-scores and the overall effectiveness rating.  The “other” raters, on average, account for 30% of the variance between their ratings and the total.  In other words, the feedback from “others” is basically twice as accurate as the individual’s self-ratings.

Not surprisingly, we tested each of the rater groups individually and found that the direct reports were the best predictors. They predicted 33% of the variance, peers predicted 28%, “others” predicted 27% of the variance and managers predicted 27% of the variance.

Weird and Wonderful

This may strike the reader as odd because many people believe they know themselves well and are the most accurate predictors of their performance. The simple truth is that we do not. Also, while it is tempting to think, “Well, I’m an exception to the rule,” the odds are that you are not.

We often catch ourselves thinking, “If that person could only see what a _________ they are.” The blank can have a variety of responses, ranging from extremely positive:  amazing person, saint, intelligent, highly productive, or creative; to less flattering adjectives such as: idiot, creep, annoying or narcissistic.  There’s one consistency that predicts the direction of this chasm: those who perform the worst tend to over-rate their capabilities, while those who are the best tend to under-rate themselves.

The wonderful part of this equation, however, is that others are often very willing to share their more accurate feedback with you if you ask. In our view, the best way to get good constructive feedback is by asking for it from others. When a person asks for feedback, it prepares them mentally to receive and accept the feedback as well. Most of us know that others are often prepared to tell us what we want to hear. Instead, we need to ask for feedback in a way that allows others to tell us what we need to hear to progress.

Rather than saying, “What did you think of my presentation?”

Ask, “What is one thing that I could have done to make my presentation much better?”

Inviting others to be honest and straightforward will increase the chances that we will receive helpful feedback.

Be aware that while other people are better predictors than you, they are not perfect. They account for 30% of the variance, but not 100%. In other words, the feedback you receive from one other person may not be totally accurate, and it will not be the same message you would hear from others. However, when we collect feedback from several people, the cumulative results become increasingly accurate.

With each additional person you query, the likelihood of receiving fully accurate feedback increases.  As the Greek aphorism goes, “Know thyself.” If you want to truly “know yourself” the best advice you can receive is to ask multiple others, and ask them to be specific, and with a mindset of wanting to fully absorb and consider the input they share.

* Now, for a brief review of basic statistical analysis for those who don’t live in this world.  We calculated a Pearson ’s Correlation Coefficient (the classic measure of how well two things correlate with each other) for both the self-rating (r = 0.38) compared to the total score, and another for all the other raters collectively (r = 0.547) compared to the total score. By squaring any correlation coefficient you can estimate the amount of the variance that is accounted for by knowing one of the variables. In other words, you can estimate how accurately any rating correlated with the overall rating.