Excel provides formulas for most of the common tests and distributions (as described in Built-in Statistical Functions). Excel doesn’t provide formulas for some tests, especially for non-parametric tests. For these we use statistical table, as described in this appendix.

- Wilcoxon Rank-Sum Table
- Mann-Whitney Table
- Wilcoxon Signed-Ranks Table
- Runs Test Table
- Kolmogorov-Smirnov Table
- Shapiro-Wilk Tables
- Studentized Range q Tables
- Spearman’s Rho Table
- Kendall’s Tau Table
- Durbin-Watson Table
- Lilliefors Table
- Dunnett’s Table
- Augmented Dickey-Fuller Table

See Supplemental Functions for a list of the various Real Statistics functions which can be used for the lookup of values in these tables.

Also see Interpolation for a description of how interpolation is done for values between those found in these tables.

This is, by far, my favorite statistics site. Thank you fo much for your efforts and your explanations!

Greetings from the Caribbean.

Jazzy,

Nice to hear. Thanks for your comment and support.

Charles

Hi Charles,

Thanks for providing these amazing resources especially for people who can’t afford to buy Minitab or other statistical software. Your instructions are very simple and well written. Do you any plans in the future to add MSA (Attribute and Continuous) analysis as well?

Hi Jas,

Thank you very much for your kind remarks.

I already provide support for Gage R&R. See

Gage R&R.

Charles

Hello Charles,

I was wondering if there existed a table for Pearson’s correlation coefficients. (ones similar to the spearman and kendall tables)

Also, is it still appropriate to use the one-sample correlation function on excel if the data is binary.

For example, linking gender (Males=0, Females=1) with whether or not they lived past the age of 100 (/=100 = 1) in a sample size of 50 people.

Connie,

Tables for Pearson’s correlation coefficient do exist. I am planning to add such a table to the website.

With binary data the normality assumption for the one-sample correlation test won’t be met. You can use the CORREL function, but the correlation test won’t be valid. You can use the chi-square test for independence if it meets your needs.

Charles

Many many thanks for your wonderful and exhaustive service, which is really amazing. I find your explanation and add on so very helpful. It would be wonderful, if you could kindly add time-series analysis and cluster analysis, as well.

I am glad that you like the site. I already have time-series analysis and cluster analysis in the list of future enhancements. I hope to get to these after the summer.

Charles