cancel
Turn on suggestions
Auto-suggest helps you quickly narrow down your search results by suggesting possible matches as you type.
Showing results for 
Show  only  | Search instead for 
Did you mean: 
Check out the JMP® Marketplace featured Capability Explorer add-in
Choose Language Hide Translation Bar
Nima
Nima
Level I

How to compare 4 treatments with unequal number of samples

Created: Nov 11, 2021 12:09 PM  |  Last Modified: Jun 8, 2023 5:41 PM (732 views)

Hi,

I have 4 treatment groups, each with different n (sample number). The experimental design is completely randomized design. What is the best way to analyze the data to make sure unequal variances are taken into consideration.

0 Kudos
2 REPLIES 2
P_Bartell
P_Bartell
Level VIII

Re: How to compare 4 treatments with unequal number of samples

Created: Nov 11, 2021 12:30 PM  |  Last Modified: Nov 11, 2021 9:36 AM (726 views)  |  Posted in reply to message from Nima 11-11-2021

I presume you are primarily interested in testing the hypothesis that the means of the 4 treatment groups are equal? If that's the case the Fit Y by X platform is a good place to start. There are options within the report for the unequal variances scenario. Here's a link to the online JMP documentation related to the unequal variances scenario.

Fit Y by X - Unequal Variances 

0 Kudos
Mark_Bailey
Staff Mark_Bailey
Staff

Re: How to compare 4 treatments with unequal number of samples

Nov 12, 2021 12:52 PM (693 views)  |  Posted in reply to message from Nima 11-11-2021

You started by saying that the sample size varied across the treatment groups, but then you ask how to analyze the data when the variances are not the same. You omitted the explanation of why one difference leads to the other difference in your case.