# Compare scores calculated for two groups who overlap?

I have three groups of ten children (A, B and C). They are each have different treatments (the treatments are to do with how they are taught about a certain biological feature). Two scores are calculated: one which uses the results from group A and group C and the other uses the results from group A and group B.

How can I test to see if the differences in the two scores are significantly different?

Each child does not have an individual score/mark.

I'm not sure about using a t-test because the samples are not paired and they are also not independent. Also, I'm not using means so I don't know if that rules out a lot of tests.

• What do you mean by not using means? Because if your $n=2$, then it might be problematic without the raw data. I mean you would need to know at least the median/mean and MAD/std for both populations, because else you can't really assess how far apart they are. – FisherDisinformation May 5 '16 at 13:38