This is the answer that I've been looking for in 3 days long, the SPSS for my CV! Thank you sir for this great video. Now I can continue writing the results in my thesis happily.
Hello, I calculated the "CVar" for each individual as you suggested, and I ran an anova. Now, do I have to look at the significance of the ANOVA test, or run a Levene's on that ANOVA? From the video this point is not clear. Thanks!
Thank you for your video; and I have a question: What ranges of CV% corresponding to narrow variation and that at which I can see my system show wide variation?
I really have searched alot for a good post-hoc test, but Levenes F' is just for normal populations that have the same variance, so it was impractical for absolutely new data sets where no assumption could have been made for distributions. I found the Games-Howell-test. This can also deal with non-normal distributions AND differences in group variances. Just in the case there is somebody else searching for it.....
statisticians do not advocate applying Levene's test for testing for homogeneity of variance. Fit the model, generate the plot of residuals versus fitted values. If you have too few observations to be able to reliably assess homogeneity of variance from a residual plot, then Levene's test will be useless anyway!
This is the answer that I've been looking for in 3 days long, the SPSS for my CV! Thank you sir for this great video. Now I can continue writing the results in my thesis happily.
Great.
Hello,
I calculated the "CVar" for each individual as you suggested, and I ran an anova. Now, do I have to look at the significance of the ANOVA test, or run a Levene's on that ANOVA? From the video this point is not clear. Thanks!
dude you are awesome
so original video
you are a real teacher
Thank you for your video; and I have a question: What ranges of CV% corresponding to narrow variation and that at which I can see my system show wide variation?
I really have searched alot for a good post-hoc test, but Levenes F' is just for normal populations that have the same variance, so it was impractical for absolutely new data sets where no assumption could have been made for distributions. I found the Games-Howell-test. This can also deal with non-normal distributions AND differences in group variances. Just in the case there is somebody else searching for it.....
I wondered how many people clicked part 2 after watching part 1...
I know I did =D
Thank you very much
Very nice.
Coefficient of variation can take only values which values
I clicked part 2!
good job!
Nice
statisticians do not advocate applying Levene's test for testing for homogeneity of variance. Fit the model, generate the plot of residuals versus fitted values. If you have too few observations to be able to reliably assess homogeneity of variance from a residual plot, then Levene's test will be useless anyway!
I think the first one, but just wanted to be sure at the 0.001 level
3277
Obviously you did. Congratulations. -.-