LISTSERV at the University of Georgia
Menubar Imagemap
Home Browse Manage Request Manuals Register
Previous messageNext messagePrevious in topicNext in topicPrevious by same authorNext by same authorPrevious page (October 2002)Back to main SPSSX-L pageJoin or leave SPSSX-L (or change settings)ReplyPost a new messageSearchProportional fontNon-proportional font
Date:         Wed, 16 Oct 2002 17:28:56 -0600
Reply-To:     Judy Brown <Judy.Brown@state.co.us>
Sender:       "SPSSX(r) Discussion" <SPSSX-L@LISTSERV.UGA.EDU>
From:         Judy Brown <Judy.Brown@state.co.us>
Subject:      interpreting kurtosis and skewness
Content-Type: text/plain; charset=Windows-874

as you know, kurtosis is a measure that, generally speaking, indicates how peaked or flat a distribution of scores is relative to a normal distribution. One source says that the kurtosis of a normal distribution is 3, but it goes on to say it is zero (since excess kurtosis standarized distribution is kurtosis minus 3)http://www.itl.nist.gov/div898/handbook/eda/section3/eda35b.htm. So, I don't know if the typical kurtosis value that results from SPSS analysis is expecting a value close to zero for a normal distribution or a value near 3. Do you know?

Regarding interpreting these statistics, guidelines provided by Dr. Brown at Univ. of Hawaii http://www.jalt.org/test/bro_1.htm suggests you calculate the range depicted by plus or minus 2 standard errors of the skewness and if your skewness result is within that range then you can assume a normal distribution (or "skewness was within the expected range of chance fluctuations in that statistic,"). The same method was recommended for kurtosis. Do you know of other guidelines for determining if the kurtosis or skewness results are or are not indicative of a normal distribution?

Thanks for your thoughts on the matter.


Back to: Top of message | Previous page | Main SPSSX-L page