Recommended Statistical Test

@ntzori Said

Hello,

Here is the issue that I'm dealing with.

We are manufacturing components ofr one of our clients.
Each batch contains 500 to 3000 components.
We take a random sample from each batch and perform inspection. Our sample size is somewhere between 10 to 20 samples. The inspection is detsractive, so the samples are not returned to the batch.
The client also takes samples for inspection.
The client shares the inspection results with us.
All the results are variables.

I would like to perform statistical analysis to try and reveal differences between the two groups of inspection data.
I would like to be able to draw conclusion regarding differences bewteen test methods; data resolutio; etc.

What type of statistical test would you recommend?

Thanks

@Pantagruel Said

I was also thinking T-test.

your sample size is small.

@Pantagruel Said
you cant state statistical significance without a few hundred people in your sample.

@Pantagruel Said

let me put it to you like this... would YOU trust any kind of important life decision on a study which was conducted on only a few dozen people?

Quote:
lets say it was important, like the cancer rates of las cruces new mexico. and they only contacted twenty people. would you base any actual life decision on such a study?