Friday, March 31, 2017

Replication and effect size

There has been questioning about the sturdiness of some psychological research.  The terms "replication" and "effect size" are often used.  If I do a study and you can get similar results when you do a similar study, a common term is "robust".  Sturdy studies, robust studies give similar results even when the original experiment was not completely duplicated.

Clearly, complete duplication is not really possible.  It may sound picky, but the moment when some experiment was conducted will not occur in time again.  And, of course, we don't expect the very same people to go through the very same procedure again.  Even if they did, they would have gone through that procedure once before, which toughens them or alerts them or fatigues them.  Besides, that is not what we want anyhow.  We want to be able to pick different people and find that the same kind of result happens with them as happened in the original study.

Effect size is a measure of how much difference something makes.  If we say we can capture more flies with honey than with vinegar, the effect size of using honey will give us an idea of how much more effective honey is.

When I began fulltime graduate school, I had been teaching the 5th grade for four years.  For a PhD at that school, I needed a minor.  I couldn't decide between psychology and philosophy and I completed a minor in both subjects.  At first, I assumed that psychology would be more relevant and more helpful but over time, philosophy has been mind-expanding and thought-provoking.  I think it is fair to say that philosophy has helped me ask better questions.  It has never seemed surprising to me that getting really new insights into myself and others is going to be rare.  I find lists of psychological principles supposedly based on research that are often things that poets and philosophers have understood for millennia.  Here is a recent and responsible one:

Popular Posts

Follow @olderkirby