# Thread: Which test to use? Comparing two samples

1. ## Which test to use? Comparing two samples

Hi all - given stats is the vein of my life (I am a lowly geologist!), i wonder if anyone can help with this (simple im sure) question:

I have measured five different (unrelated) parameters on two different pieces of the same rock. So I have two sets of 5 independent numbers.

I need to test whether the results I have obstained are consistent between the two samples i.e. whether the 5 measurements are NOT signficantly different from each other.

Is there a basic test that could do this? If I am expecting the results to be no different between each rock sample (i.e. the same tests on teo samples of the same rock type should yield the same result) perhaps chi square would be possible, but then I get confused about how I calculate the 'expected' result (just the mean of the two results?!).

Im confused because the tests are not something I can replicate to get lots of numbers, its just one outcome per test. I hope I have explained it well enough

Thanks to anyone that can help!!

2. ## hi

hmm, there are tests like deviance tests, but two samples is not much.
I´m not sure that deviance test would be good here either.

Deviance (statistics) - Wikipedia, the free encyclopedia

I will allow for more experienced people to answer this.