On Dec 14, 5:26 pm, djh <halitsk...@att.net> wrote: > Thanks for doing those two plots. > > You wrote: > > "The biggest difference I see is that the SE's in set 2 are more > consistent than those in set 1, especially for L > 90 or so. However, > that could be just a sample-size artifact." > > Yes - we agree on the difference we're seeing. > > And I'm therefore hoping you'll tell me there's a way to reliably > decide the sample-size artifact question. > > Is there?
Multiply each SE by the square root of its n, then plot. Also, you might try plotting the points without connecting them, which will make it easier to see holes.