Statistical Test for Differences Between Groups with Varying Sizes

[Yann]

A very simple thing; I have to collect data with 4 different group size. Something like;

Group A; 12.4, 17.8, 12.5 ...
Group B; 15.2, 9.5, ...

I simply want to know if the differences between the groups are statistically signifiant. I don't want anybody to explain the me the math, I just want the name of the statistical test.

 

[EnumaElish]

You have to explain what you mean by "differences among the groups." Do you mean group means? Or distributions?

Not knowing the answer, two tests that come to mind are:

Parametric: http://en.wikipedia.org/wiki/T-test#Unequal_sample_sizes
Non-parametric: http://en.wikipedia.org/wiki/Pearson's_chi-square_test

 

[Yann]

Hi,

I want to know if the variations between the group means are statistically significant (which probably mean I'll need to use the size of the sample for each group as well).

 

[EnumaElish]

You can use the t test for pairwise comparisons of means.