Normality of Data

Six Sigma – iSixSigma Forums General Forums General Normality of Data

Viewing 2 posts - 1 through 2 (of 2 total)
  • Author
  • #53505


    Hi This is vijay & i would like to know
    1. How can i check whether my data is normal or not without using any statisctical software i.e. by manual calculation & drawing the graph how can i check the normality of Data.


    Robert Butler

    Your question is far too general and the answer to any particular piece would be far too long for a forum of this type.

    You can plot your data on arithmetic probability paper and visually check to see how well the plot matches the ideal line. The problem here is that for small samples you can have plots that, visually, don’t seem to match the ideal line at all but, when tested formally, do not result in differences from normal that are significant.

    I’d recommend doing some reading.

    To give yourself an appreciation of normal probability plots and sample sizes look at appendix 3a pp.33 to 43 of Fitting Equations to Data 2nd Edition – Daniel and Wood.

    To give yourself some understanding of the calculations associated with a test (please note I said “a” and not “the” – there are a number of such tests and they are a function of things such as sample size) get Practical Nonparametric Statistics 2nd Edition – Conover and look at Chapter 6 – Statistics of the Kolmogorov-Smirnov Type. This chapter also includes a discussion of the mechanics of the Shapiro-Wilks test.

Viewing 2 posts - 1 through 2 (of 2 total)

You must be logged in to reply to this topic.