Which Stat test should I use? - (May/28/2010 )

Hi everyone!

I have a very simple question but I'm out of practice and can't find the answer anywhere!

Basically, I have 3 different machines I have tested against each other, to see if they give statisticaly significant similar results.

I have put 18 different types of sample though each machine. Each of these 18 samples is actually a mean of approximately 16 values.

I would like to compare these 18 means for each machine to see if the machines are reading similar enough.
How do I do this...? ANOVA etc seem to average up these means and compare those, but i want the samples kept seperate....

Help!
Thank you in advance!!

Try changing your data sets into the difference between machines