+ Reply to Thread
Results 1 to 2 of 2

Thread: Npar K-S test acceptable for normailty test??

  1. #1
    Points: 834, Level: 15
    Level completed: 34%, Points required for next Level: 66

    Thanked 0 Times in 0 Posts

    Npar K-S test acceptable for normailty test??

    hi just testing parametric assumptions and realised that they are 2 ways to check Kolmogorov-Smirnov Test.
    1) Analyze-Nonparametric Tests-1-Sample K-S… place variables in the Test Variable List… box, click OK
    2) Analze - Descriptive Statistics, then Explore, plots button. Under descriptives click on the Histogram. Click on Normality plots with test.

    im still learin stuff in stats still havnt got the hang of it yet. but can i use the 1-Sample K-S method to test for normailty. i have a sample size of 185, im looking at differnces between 2 groups lets say male n female and have 6 varibles (interval)
    i did read somewhere that 1-Sample K-S is just for 1 sample but if i am testing the sample as a whole for normailty is that acceptable? this method gives me non-sig results so i would able to use parametreic tests
    the other other method gives me 4/6 varibles that are sig.

    i really hope this makes sense!!?? and some1 could shed some light on this ohh and in the easiest way to understand as i am not familar with many stats terms

  2. #2
    Points: 71,220, Level: 100
    Level completed: 0%, Points required for next Level: 0
    User with most referrers
    trinker's Avatar
    Buffalo, NY
    Thanked 928 Times in 809 Posts

    Re: Npar K-S test acceptable for normailty test??

    There are numerous test of normality (depending on your situation). I can't tell what program you're using. And so can't give much insight about your question.

    For me, I usually check normality (though this may not be my sole approach) with a QQ norm plot. This is visual and pretty easy to understand/interpret. I believe most programs can give you this plot fairly easily. Using the QQ-plot is a standard practice for many statisticians.

    Wikipedia QQ-plot Article
    "If you torture the data long enough it will eventually confess."
    -Ronald Harry Coase -

  3. The Following User Says Thank You to trinker For This Useful Post:

    aarong (07-15-2011)

+ Reply to Thread


Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts

Advertise on Talk Stats