I performed a Kruskal-Wallis test to compare 5 groups, in the multiple-comparison Z-value test I found differences between groups, with z-values >2.8070 (Bonferroni test) or >1.9600 (Regular test), but I couldn´t find the power of my results. Does a power analysis it possible? I'm using a NCSS 2007 software to perform such comparisons.
This could be useful
http://www.marinespatialecologylab.org/wp-content/uploads/2010/11/Non-parametric-power-analysis.pdf
Instead of doing a post-hoc power analysis, which is considered bad practice by many researchers, you could use confidence intervals in order to make statements about the meaningfulness of your tests.
You can have a look here: https://www.researchgate.net/post/Is_it_possible_to_calculate_the_power_of_study_retrospectively
One way to do it is to use a Generic Chi Squared Test, sensitivity analysis.
- Badges
- Science method
- Similar topics
- Biological Science
- Histology
- Staining