I'm doing a stats assignment and for one of the questions I need to make a judgement of whether there is skew and kurtosis from a p-p plot in SPSS. I've been over the lecture, and we were told to look for "snaking" around the line for skew, and points "hanging" off the line for kurtosis. I'm having trouble telling just from eyeballing it - the examples they used in the lecture haven't made it clear to me how to differentiate accurately between skew and kurtosis. Can you tell me how to make the call exactly? Here's the p-p plot.

enter image description here

  • $\begingroup$ This is for my residuals, by the way. Not sure if that makes a difference. $\endgroup$
    – Simon
    Commented Apr 18, 2016 at 8:03

1 Answer 1


There is some curvature in the upper end of the plot, not in the lower end. That could indicate some skew in the distribution. The effects are small, so difficult to say if what we see is significant. You could simulate an envelope around the plot. Also, have a look at the answers here: Interpreting QQplot - Is there any rule of thumb to decide for non-normality?

I don't think I would try to assess kurtosis from that plot!

  • $\begingroup$ Thanks for that! I have to say I'm not too thrilled with all the judgements based on eyeballing we're being asked to make in this assignment... $\endgroup$
    – Simon
    Commented Apr 18, 2016 at 9:16
  • 2
    $\begingroup$ I think you need to take your dissatisfaction up with your course teachers. I don't find P-P plots easy to interpret; give me a Q-Q plot any time I have to choose one. $\endgroup$
    – Nick Cox
    Commented Apr 18, 2016 at 13:01
  • $\begingroup$ An alternative to qq-plots which are easier to interpret are relative distribution plots. $\endgroup$ Commented Apr 18, 2016 at 13:08
  • 1
    $\begingroup$ I don't have anywhere near enough knowledge to argue with my stats lecturer about the material in the unit. I appreciate the help though. In the end I emailed my lecturer who confirmed the answer was that skew is present but kurtosis is not. $\endgroup$
    – Simon
    Commented Apr 18, 2016 at 17:07

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.