off course, we do sampling to decrease the time , cost, effort but we must choose the sample in way must the results gate from study can be applicant to the target population, so its very important to choose good representative sample with correct sample size,

Dear Paul,

Yes you have presented the question clearly enough. My suggestion is as below,

1. Regarding sample size:- It ll be based on your popolation size. Further, determine the same using sample size determination formula. There are many of it. Here is one,

(S)/1+[(S-1)/population]

Where S= Zsquare* p(1-p)/M square

Here S=sample size for infinite population

Z= Z score

p= population proportion( generally assumed as 50% or .5)

M= Margin of error (generally assumed as 5% or .05)

2. Regarding sampling technique- It ll be better if you determine the technique based on your objective. What i mean is, if you want to analyse ethnicitywise effect, you can do a stratified random sampling. If you want to test gender specific effect, you may go for simple random sampling technique.

However, breaking your 6 samples(say) as 3 male and 3 female will be quota sampling.

Just summarized all thoughts came to mind. Hope it helps.

Have you done a power analysis?

3 participants per group is mostly likely not going to be sufficient.