Sunday, October 17, 2010

Confounding Variables in A Study

I like this post from Dr. Mike.  He examines a study done by Dr. Ornish in 1990.  The study provided a generally (but not perfect) good outcome for the study group.  However, as Dr. Mike explains, you cannot determine causality in a study with four different interventions.  We can ask if the smoking cessation, the meditation, the diet, or the exercise help the study group, but without follow on studies, we'll never know.

Strangely, Dr. Ornish wants to claim otherwise.

No comments:

Post a Comment