Others may differ, but I teach regression to the mean as being a different phenomenon than what you might refer to as capitalizing on the cyclical nature of an event. For example, many people may not go to the doctor until their cold or flu is at its worst which means that, by the time they get a prescription, it is impossible to distinguish the effect of the drug from the natural cycle of the illness.
I explain regression to the mean as being due to the inherent error in any test based on X = T + e (or the score on a test (X) is the sum of the true score (T) plus or minus error (e)). This means that some extreme high scores are based on the true score plus error. In those cases, the obtained score will regress to the mean on second testing because traditional testing theory assumes that the error is as likely to decrease as to increase a score. Therefore, if you choose the extremely high scores, the score on second testing is likely to regress toward the mean (be lower) because the error is unlikely to make such a positive contribution on second testing. At the other extreme, some extreme low scores are based on the true score minus error. In those cases, the obtained score will regress to the mean on second testing (get higher) because the error pulling the score down is not likely to be as strong on second testing. So, if you sample just the extremely low scores, the score on second testing is likely to regress toward the mean (be higher) because the error is unlikely to make such a negative contribution to the total score on second testing. Do others make this distinction when discussing these effects? Rick Dr. Rick Froman, Chair Division of Humanities and Social Sciences Box 3055 x7295 [email protected] http://tinyurl.com/DrFroman Proverbs 14:15 "A simple man believes anything, but a prudent man gives thought to his steps." -----Original Message----- From: Frantz, Sue [mailto:[email protected]] Sent: Monday, February 09, 2009 11:15 AM To: Teaching in the Psychological Sciences (TIPS) Subject: RE: [tips] Bogus treatments For regression to the mean, how about vitamin C to cure colds? (http://www.mayoclinic.com/health/cold-remedies/ID00036) Everyone starts out with a cold, a day or two into it they start drinking lots of orange juice, and a few days later they feel better... which likely would have been the case anyway, but they credit the orange juice. --- To make changes to your subscription contact: Bill Southerly ([email protected])
