I have used Kruskal-Wallis and Wilcoxon's rank-sum test for some analysis. For calculating the corresponding effect sizes, I have used:
- epsilon-squared for the Kruskal-Wallis test
- r-squared for Wilcoxon's rank-sum test
I'm not getting any definitive source to interpret the effect size values. Does anybody know of any definitive source?
Hi,
Here are some of them:
Effect sizes for research: a broad practical approach / Robert
J. Grissom, John J. Kim.
The essential guide to effect sizes : statistical power, meta-analysis, and the interpretation of research results / Paul D. Ellis.
Confidence Intervals for Proportions and Related Measures of Effect Size
Robert G. Newcombe
Hello Arka,
Bhogaraju Anand 's reply gives some good suggestions. However, please note that published guidelines are just that: guidelines, and not commandments as to what is or isn't an important effect size.
The specific variables, the specific target population, and the perspective of the consumer of the results all make a difference as to what comprises a noteworthy (vs. an ignorable) effect size. In other words, context matters. Jacob Cohen, author of the widely cited, Statistical power analysis for the behavioral sciences (2nd ed.), 1988, would have been the first to acknowledge that his guidelines (e.g., "small" effect size = r of .10, "medium" effect size = r of .30, "large" effect size = r of .50) are suggestions for when one doesn't have intimate understanding of the context. I can give you an example in which a correlation of +/- .01 is salient, and another example in which a correlation of .80 is ignorable.
Finally, yes, a lot of times people suggest expressing rank-based comparisons in some strength of correlation metric as the effect size.
Good luck with your work.
Dear Arka,
This is nice paper
https://www.frontiersin.org/articles/10.3389/fpsyg.2013.00863/full
Bhogaraju Anand, thanks, I'll check if these are available in our library and go through them.
David Morse, thank you for the detailed answer, I'll keep it in mind while reporting my findings.
Ahmad Al-shallawi, the paper talks about effect sizes for parametric tests, but I was looking for effect sizes for non-parametric tests; thanks anyway.
For anybody facing the same problem, I finally followed the recommendations mentioned in the book SPSS survival guide by Julie Pallant.
Christoph Centner, I did bump into this website during my research. However, the values and interpretations I was getting from this site didn't match the ones that I calculated manually by following the book SPSS survival guide. I decided to go with the book because I thought the book is more citable.
- Badges
- Science topic
- Similar topics
- Biological Science
- Kinesiology
- Running