Dear Ruben, That depends on the amount of skewness. If the mode is located at the beginning or the end of the scale, there's no transformation that can possibly accommodate for that. In that case, I would suggest to dichotomize the variable and to perform a different analysis. If not, I think you can perform some kind of log-transformation, construct the confidence interval of this transformed variable, and transform the lower- and upperbound back. Good luck! Best regards, Joost van Ginkel Joost R. Van Ginkel, PhD Leiden University Faculty of Social and Behavioural Sciences Data Theory Group PO Box 9555 2300 RB Leiden The Netherlands Tel: +31-(0)71-527 3620 Fax: +31-(0)71-527 1721 ________________________________ From: SPSSX(r) Discussion [mailto:SPSSX-L@LISTSERV.UGA.EDU] On Behalf Of Ruben van den Berg Sent: 04 September 2009 11:20 To: SPSSX-L@LISTSERV.UGA.EDU Subject: Confidence interval for extremely skewed metric variable Dear all, I want to estimate a confidence interval for the mean of a metric variable that's extremely skewed to the right. As I (hopefully!) understood, the central limit theorem will make sure that the sampling distribution of the mean will follow a Gaussian distribution (assuming enough observations). However, the skewed distribution causes the standard deviation to be very large compared to the mean value, rendering a very wide confidence interval that's not too informative. Is there any way (e.g. by a transformation or something) to obtain a smaller interval? TIA! Ruben van den Berg ________________________________ See all the ways you can stay connected to friends and family <http://www.microsoft.com/windows/windowslive/default.aspx> ********************************************************************** This email and any files transmitted with it are confidential and intended solely for the use of the individual or entity to whom they are addressed. If you have received this email in error please notify the system manager. **********************************************************************