RE: Odds Ratios

[Martin J. Bourgeois]

Title: AAP Blame Game Again

I don't like odds ratios as effect sizes (and I'm not sure what they mean in a logistic regression), but I agree with Karl that an odds ratio of 1.09 seems quite small (and the confidence intervals for 1- and 3-year olds went down to 1.03 and 1.02, respectively). And no p-values?  Given this and the correlational (albeit longitudinal) design of the study, why is it getting so much media attention? And what's with the kids watching 12-16 hours of TV per day? I wonder if these outliers contributed to the effect? Is it appropriate to do logistic regressions when the data are so severely skewed?   Marty Bourgeois University of Wyoming -----Original Message----- From: Karl L. Wuensch [mailto:[EMAIL PROTECTED] Sent: Mon 4/5/2004 7:53 PM To: Teaching in the Psychological Sciences Cc: Subject: Odds Ratios     The odds of having attention problems among those who did watch a lot of TV during infancy divided by the odds of having attention problems among those who did not watch a lot of TV during infancy = 1.09.  This does not impress me in terms of size of the effect.  You can find an introductory lesson on logistic regression at http://core.ecu.edu/psyc/wuenschk/MV/MultReg/Logistic-SPSS.doc .   ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Karl L. Wuensch, Department of Psychology, East Carolina University, Greenville NC  27858-4353 Voice:  252-328-4102     Fax:  252-328-6283 [EMAIL PROTECTED] http://core.ecu.edu/psyc/wuenschk/klw.htm ----- Original Message ----- From: Wallace E. Dixon, Jr. To: Teaching in the Psychological Sciences Sent: Monday, April 05, 2004 7:14 PM Subject: AAP Blame Game Again Dear Colleagues,     All over the news today is a claim that, by implication, watching TV in infancy basically causes attention problems in children.  Of course, the AAP doesn't say this, but that doesn't matter since the press is splattering ITS understanding of the claim all over the place, and this is the message people will take home with them.     In reading over the article to find the source of the claim (found at http://www.aap.org/advocacy/releases/apr04studies.htm) and the strength of the relationship, I found that the authors use a logistic regression statistic that I don't understand.     Having never used logistic regression before, I'm not familiar with the concept of an "odds ratio," at least not in the sense that it's used here. On p. 712, it is shown that the odds ratio (of WHAT I cannot tell) is 1.09. Apparently, this statistic shows the strength of the relation between tv-watching in infancy and later attention problems.  Can anyone explain this statistic to me?     In their position expressed today, the American Academy of Pediatrics may be even more guilty of scientific misconduct than their previous claim that watching TV before age 2 is bad and should be avoided. Wally Dixon --- You are currently subscribed to tips as: [EMAIL PROTECTED] To unsubscribe send a blank email to [EMAIL PROTECTED] --- You are currently subscribed to tips as: [EMAIL PROTECTED] To unsubscribe send a blank email to [EMAIL PROTECTED] b‹š­ç.®·§¶\¬¹»®&ÞvÚ-Š›±ªÜ+Þ¦è®èº{.nÇ+‰·¬zwZnV§‘é[h•æ¯zسŸ;Ñfœ²éß‹-nêàyÛ