II. Probability and Statistics for Reliability
A. Basic concepts
5. Non-parametric statistical methods (Apply)
Apply non-parametric statistical methods, including median, Kaplan-Meier, Mann-Whitney, etc., in various situations.
Let’s talk about a few approaches that use ranking.
Additional References
Kruskal-Wallis Test (article)
Spearman Rank Correlation Coefficient (article)
Mann-Whitney U Test (article)
Quick Quiz
I had trouble following the examples here. How did you get the chi-square value? What was S in the equation?
Hi Raye,
The S is the sum of squares, the sum of the (Ri – R-bar) squared for all i. Slides six, I think, in the presentation.
Chi squared for the test statistic is approximated from 12S / nm(n+1) which is about the same as the X^2 alpha, n-1 value, which allows us to make the comparion to the critical value, which is found from the chi-squared table with alpha of 0.1 and df=n-1
hope that helps.
Cheers,
Fred
Fred,
You mentioned in the lesson that you would provide a link to a table that summarizes all of these methods and associated calculations. Could you please provide the link or send the table via email if it is available?
Thanks,
Ray