Standard Deviation comparison between 2 dataset

Hello,

I would like to ask if there is any funtion in matlab that support to do comparasion between 2 standard deviation from 2 data like in Minitab. The only one i can find is Vartest2, but it compare Variance instead of Standard deviation.

Assistant_Test_for_Standard_Deviations.pdf (minitab.com)

Thank you all.

[Mathieu NOE]

hello

as far as I can remember variance is the squared value of SD. so if you have a solution that works for V you have a solution for SD

[Truong Dinh]

the final result maybe the same ( reject or not reject) but the P-value is different. I try to do something that similar with minitab function.

Answer 1

[Image Analyst]

Maybe you want the ttest() to do a T-test, or anova() to do the "F statistic" test.

[Truong Dinh]

sorry for late comment, ttest2 is comparing the different bettwen the mean of 2 sample. However i want to compare the different between the Standard Deviation of 2 sample and i want to compare with Mintab result. F statistis test is not using Bonett test and it cannot test Significant Different of 2 sample. I try to write the function in Matlab using Minitab Formular, however it was not return the same value.

nx= numel(X);
ny= numel(Y);
sx = std(X);
sy= std(Y);

gx = (nx-3)/nx;
gy =(ny-3)/ny;
mx = mean(X) /(2*sqrt(nx-4));
my =mean(Y) /(2*sqrt(ny-4));
cx = nx/(nx-z);
cy = ny/(ny-z);
C =cx/cy;
gamma_p =(nx+ny)*(sum((X-mx).^4) + sum((Y-my).^4)) / (((nx-1)*(sx^2)+(ny-1)*(sy^2))^2);
se = sqrt(((gamma_p- gx)/(nx-1))+((gamma_p- gy)/(ny-1)));
%% estimate z
Z_est =log(C*(sx^2/sy^2))/se;
P = normcdf(Z_est);