PDA

View Full Version : A statistics question

xyang41
05-26-2011, 04:51 PM
Dear Biomch-L community,

In my studies, I need to test homoscedasticity(or homogeneity of
variances) of different groups. However, commonly used test methods
(Bartlett's test, Levene test and Brown-Forsythe test) are designed to test
effects of only one independent variable on the outcome variability. Are
there any other statistic methods that can be conducted to test the effects
of two independent variables in homoscedasticity analysis? We are looking
for a method analogous to a 2-factor ANOVA, that can be used to test
(compare) the variability in several groups.

Yang Xu

Biomechanics Lab
Department of Kinesiology
The Pennsylvania State University

gwhite43
05-27-2011, 07:34 PM
Starting here might help. Sorry but I'm at home and don't have access to the paper.

http://www.sciencedirect.com/science/article/pii/0167947383900622

xyang41
05-30-2011, 10:35 PM
Thank you very much!
That helps a lot.

Yang Xu

Starting here might help. Sorry but I'm at home and don't have access to the paper.

http://www.sciencedirect.com/science/article/pii/0167947383900622

icegirlclaire
06-05-2011, 11:23 PM
That is what I am looking for. Thanks

Rob Herbert
07-10-2011, 03:51 AM
Dear Biomch-L community,

In my studies, I need to test homoscedasticity(or homogeneity of
variances) of different groups. However, commonly used test methods
(Bartlett's test, Levene test and Brown-Forsythe test) are designed to test
effects of only one independent variable on the outcome variability. Are
there any other statistic methods that can be conducted to test the effects
of two independent variables in homoscedasticity analysis? We are looking
for a method analogous to a 2-factor ANOVA, that can be used to test
(compare) the variability in several groups.

Yang Xu

Biomechanics Lab
Department of Kinesiology
The Pennsylvania State University

Just joined BIOMCH-L and saw your post. You might like to check out this paper:

Herbert RD, Hayen A, Macaskill P, Walter SD. Interval estimation for the difference of two independent variances. Communications in Statistics - Simulation and Computation. 2011; 40: 744-58.

If you don't have access I can send you a pdf copy - just contact me on rherbert@george.org.au.

Rob Herbert