Hello Sandra.

There is never a minimum that you can form a group but you should be able to prove there is a significance,

or better if you can reference it with another research article would be great

Thank you

You need to sit with Biostatistician and discuss with him/ her what you want to do. (project details) before that you need to go through literature review and search similar studies, their methodology, sample size and techniques and specific outcomes (results) of those studies. it is important to know estimated positive outcomes of your study objective inorder to calculate sample size. (you may use WHO sample Size Calcultor with help of your Biostatistition)

Generally when we divide the study population in more groups via stratification our sample size increases so that we can justify results with atleast 95% confidence and margin of error

Hi Sandra,

There are multiple factors that determine and affect the sample size in each group. These factors include:

1. Your level of significance, alpha error (0.05, 0.01, etc)

2. Your test power (Beta error)

3. Your population standard deviation of the outcome variables

4. How big the difference.

SigmaStat has the option of calculating the sample size, but you will need the above info that I mentioned.