hello. If we were to split the data into three groups instead of two, how will the Brown-Forsythe test be conducted? Will the program do the test on the two extreme groups? Or somehow do the test on all three?
Hi Leah, I believe the answer to your question is all three, but to clarify even better: this is a test againt the null hypothesis that the MAD of any number of groups are not statistically significatly different from each other (in much the same way that an ANOVA is a test on the means between groups). So the test result for 3 groups won't explicitly tell you which group(s) differ from each ofher, only that there is at least one difference between the three groups.
Struggling in my predictive modeling class right now and this is a huge help. Appreciate it.
hello. If we were to split the data into three groups instead of two, how will the Brown-Forsythe test be conducted? Will the program do the test on the two extreme groups? Or somehow do the test on all three?
Hi Leah, I believe the answer to your question is all three, but to clarify even better: this is a test againt the null hypothesis that the MAD of any number of groups are not statistically significatly different from each other (in much the same way that an ANOVA is a test on the means between groups). So the test result for 3 groups won't explicitly tell you which group(s) differ from each ofher, only that there is at least one difference between the three groups.