Small Samples

Another book I'm currently reading is called Thinking, Fast and Slow by Daniel Kahneman who was a winner of the Nobel Prize in Economics.  It's a very insightful book, he touches on so many interesting topics that he really doesn't have space to discuss all the possible implications of the theories.  One chapter talked about small samples.  I think the best way to describe the problem of small samples is by discussing one of the examples he used in the book.

Anybody listening about the education crisis in our country has heard the idea that we need "smaller" class sizes in order to effectively deliver a quality education.  Studies have shown comparing classes that the smaller classes usually tend to group towards the top of achievement.  However, the same studies have also shown that smaller classes also tend to group towards the bottom of achievement also.  What gives?  The reasoning is that in smaller sample sizes, there is an increased chance of outliers effecting the outcomes.  If you have a group of 10 students, 1 or 2 great students can make up for the performance of the others.  Inversely, 1 or 2 bad students can spoil the whole bunch.  It's a very interesting concept leading one to wonder what other biases have been influenced by this phenomena.  He was obviously not advocating that larger classes are better, but that we might be placing too much emphasis on one aspect of the education crisis.

Another topic I want to address in the future is the issue of home school vs traditional schooling.  As a Christian I think the structured environment of home schooling sounds enticing, but do we do kids a disservice in not letting them experience the world as it truly is?  I'm still on the fence about this.  I lean toward traditional schooling since that's how I went to school but there is no denying that negative peer influences can have a lasting impact on kids.