Date: Wed, 23 Feb 2005 14:39:25 +0100
Reply-To: Rouckhout Danny <Danny.Rouckhout@lessius-ho.be>
Sender: "SPSSX(r) Discussion" <SPSSX-L@LISTSERV.UGA.EDU>
From: Rouckhout Danny <Danny.Rouckhout@lessius-ho.be>
Subject: effect size and non normal distribution
Content-Type: text/plain; charset="iso-8859-1"
We recently compared the testscores from two samples (clinical versus non-clinical). Both samples show a very strong deviation from a normal distribution (as was expected with psychopatological characteristics). In addition with the common significance we wanted to calculate the effect size and the confidence interval for the effect size. However, effect size calculations are based on the assumption of a normal distribution. How can we determine a more appropriate effect size and CI. Are there any SPSS syntax files for this problem?
I read a recent article about the problem [Kelley, K.(2005) The effects of nonnormal distributions on confidence intervals around the standardized mean difference: bootstrap and parametric confidence intervals, Educational and Psychological Measurement, ol 65, 1, p51-69] that included a syntax for another statistical package. However, my syntax an math knowledge was inadequate to put it in an SPSS syntax.
Are there any suggestions?
University Hospital Antwerp