0
$\begingroup$

this question follows on from a question I asked yesterday Testing whether the variance of a data set is significant which you guys were really helpful in making me realise my intentions for the data were impossible. So now I'm trying to figure out what I'm doing with them and wondered if you guys can help me out again. Sorry if any of this sounds stupid- I'm still very new to this stats stuff.

In short I've taken length measurements for different breeds of dogs in two different countries. I've got 30 breeds for each country. For 20 of these breeds at each country I took the measurements of 6 individuals to get a mean that I'm now using for comparison and for the other 10 I took the measurements of 20 individuals to get the mean. Originally I did this because I intended to show that the variance within the groups was not significant and so the means I was using were a good indication of the overall population and that a sample size of 6 was enough to do this. However, it's been pointed out to me that without knowing more about the overall population I can't make this claim and therefore significant is just an abstract term.

So now I don't know what to do with these differences in sample sizes between breeds- is there anything or have I just pointlessly taken different sample sizes? I was thinking of testing the variance between all the samples using a levene's and if it was not significant this would show that it didn't matter if I used 6 or 20 samples because they had the same variance. But then I'm not sure if that's nonsense I've made in my head in an attempt to get something from my data.

So if anyone can spread any light on this that'd be grand and really, really appreciated.

$\endgroup$
  • $\begingroup$ I have edited in a link to your previous question. Can you edit your post to clarify what your scientific question is? Why do you have different countries, different breeds (presumably the same ones in each country?), and so on? $\endgroup$ – mdewey Mar 3 '17 at 16:38
0
$\begingroup$

I suspect you wish to conduct an Analysis of Variance with uneven sample sizes so you will use Type II Sums of Squares. For more information you could review this.

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.