Regression to the Mean & Giving Praise

Hello to all the beautiful people out there following the MTSU History Blog.  After a long and fun day of working on my thesis, I've been relaxing and reading The Economic Naturalist by Robert H. Frank.

The book has been everything I expected and more.  Granted, my expectations were only so high considering that I bought it for $2 at McKays in Nashville.  I expected to find silver.  But I struck gold.  It was buried deep within the book (or it was on page 146, whichever sounds better).



Frank answers "Why do managers tend to overestimate the efficacy of blame and underestimate the efficacy of praise?"  As a former teacher and coach, this question was extremely relevant.  I know the best literature available says be positive, give X (significantly large) number of compliments for every X (significantly small) number of criticisms, etc.  Still, there were times when criticism seemed to work really, really well.

So, why is that?  According to Frank, it has to do with regression to the mean.  People vary in their performance.  Sometimes they do great.  Sometimes they do poorly.  Sometimes they do average.  The person that does poorly one week will generally do average or even great the next.  Poor performance is often met with criticism.  That criticism is too often seen as the cause of improved performance.  The reverse is also true.  Someone who performs well one week is likely to do worse the next.  That's one of the reasons why compliments sometimes seem to do little good, or even worse, harm.  So, if your gut tells you to be negative, don't listen to it.  Hand out some praise!

If I butchered that last paragraph, please forgive me.  Given how bright this audience is, I'm sure you got the idea :)

Matt Bailey