--------------- help for fstar ---------------
F* test -------
. fstar depvar indepvar
The fstar command performs the F* test. The F* test is similar to a one way ANOVA, except that it can be more effective in yielding appropriate Type I error rates under conditions of unequal variances and unequal means across groups.
The standard ANOVA F test assumes that the groups have equal variances, and when this assumption is violated, actual type I error rates can exceed nominal type I error rates. If you suspect this problem, you can download and use the simanova command to perform monte carlo simulations given your variances and sample sizes to compute actual type I error rates given that the null hypotheses is true. If your type I error rates are inflated, you can use the fstar option to see how the F* test performs and if it performs to your satisfaction, you may choose to use the fstar command to perform the F* test on your data.
For more information about the F* test, and the inflation of Type I errors under unequal variances and unequal sample sizes, you can see
Wilcox, R, Charlin, V, Thompson, K. (1986). Communications in Statistical Simulation and Computation. 15(4) 933-943.
If you have comments or suggestions, please email Michael Mitchell at email@example.com .
Michael N. Mitchell Statistical Computing and Consulting UCLA, Academic Technology Services mnmatucla.edu Also see -------- simanova net from https://stats.idre.ucla.edu/stat/stata/ado net install simanova wtest net from https://stats.idre.ucla.edu/stat/stata/ado net install wtest