I've got myself confused over which one I should use.
If I have several groups of data which I'm comparing, do I calculate SD or SE+95% limits to compare them?
I thought SE shows you the accuracy of your mean, so if I were to go out and repeat the experiment then my mean should fall between the CLs, whereas SD shows the diversity of that particular data sample......the example I've been given uses SD.
Which do I use?
... and the ones that won't