Re: Media violence

[Christopher D. Green]

This is a metanalysis of past research I take it? Rosenthal had an article in (I think) _Psychological Methods_ recently about how to convert t-tests into r-equivalent effect size measures (based on a talk I saw him give at APA a couple of years back). It's as reasonablea  thing to do as r is a reasonable statistic. I'd have to know the knid of data they're converting to know for certain, but my suspicion is that r would tend to underestimate effects because it is sensitive to the linearity of the relationship (i.e., it misses non-linear relationships) and it tends to be depressed by restricted ranges (i.e., less than 9 or so *used* scale values). Regards, -- Christopher D. Green Department of Psychology York University Toronto, Ontario M3J 1P3 Canada office: 416-736-5115 ext. 66164 fax: 416-736-5814 e-mail: [EMAIL PROTECTED] WWW: http://www.yorku.ca/christo/ .================================== don allen wrote: Hi Tipsters-   I need some help in deciphering the latest issue of Psychological Science in the Public Interest (Vol. 4(3), 2003) on "The Influence of Media Violence on Youth".  On page 84 the authors state,   "There are several commonly used measures of effect size, any of which can be applied to experimental, correlational and longitudinal types of studies. To provide a common metric for this discussion, we have converted all effect sizes to correlation coefficients (rs)."   I haven't seen this type of transform before. My questions are:   1. How is it done? 2. Is it a reasonable thing to do?   Given my biases on this subject I suspect that it will artificially inflate the effect sizes and make them look as if they were important when they are really trivial but I am willing to be corrected by someone more mathematically knowledgeable than I.   As always, your assistance is greatly appreciated.   -Don.   Don Allen Psychology Dept. Langara College 100 W. 49th Vancouver, B.C. V5Y 2Z6 Canada 604-323-5871 --- 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]