Ask here about academic and research issues, like designing studies, recruiting participants, choosing statistics, submitting for publication, etc.
Post Reply
Posts: 24
Joined: Sat Aug 15, 2009 12:15 pm


Post by Becky.D » Wed Jun 11, 2014 12:00 pm

Hi there,

I've been going back to basics for my medical statistics exam coming up, and have started over analysing assumptions of normality and have now totally confused myself to before I started revision.

Skewness: -1 negative skew; 0 normal; +1 positive skew
Kurtosis: <3 tall/thin; 3 normal; >3 spread out

There is a range for skewness, were normality could fit between -0.5 and +0.5. However, this doesn't seem true with kurtosis. Is it a case of, either exactly 3 or not, normal or not?

Also, how do you decide on normality if kurtosis and skewness are not equal, where one is normal and the other isn't?

Many thanks & apologies for my baby brain at present.

Posts: 98
Joined: Sat Aug 11, 2012 12:07 pm

Re: Kurtosis

Post by Megzi » Sun Jun 22, 2014 9:16 pm

Hi Becky.D,

I would judge kurtosis from the resultant histogram. From what I know, leptokurtic is characterised by thinness at the point of central tendency, whereas platykurtic is flat. I believe mesokurtic is the characteristic of a normally distributed sample.

Normality can be tested by doing a Kolmogorov- Smirnov test on SPSS (it should be non-significant).

It seems that you wanted a more mathematically underpinning response - however I hope the above is helpful! Also I am aware you may have already complete mentioned exam, which in that case I hope it went well.

Post Reply

Who is online

Users browsing this forum: No registered users and 1 guest