MID is based on Standard deviation, consequently, I should not use it in non-normal distribution. Am I right? If yes, which test may replace MID?
Hello Matheus,
There are a number of options for minimal clinically important difference, some of which would make sense with non-symmetric distributions. See this article for an enumeration: Article Approaches for estimating minimal clinically important diffe...
If you feel obliged to stay with some variant of standardized differences (e.g., involving the SD), you could always invoke Chebyshev's inequality, as that defines limits for any shape distribution (however, the resulting intervals would be larger).
Good luck with your work.
if you have nnd then you have to remove outliers to ensure normalcy so that you may simplify your work.
Outliers can be deleted for normality distribution and provide a more reliability.
- Badges
- Science topic
- Similar topics
- Mathematics
- Statistics
- Applied Statistics
- Survey Design
- Questionnaire