The thing I hate about small number is you just never know if they're worth anything.
Now, is that a significant difference? 2 is twice 1, after all. Twice!
But if the next number is 183, 1 and 2 look pretty chummy.
And how about 19.8 for n = 4, versus 32.2 for n = 21?
But is it any good?
I hate n being only 4. What if some fifth sample has a value of 40, and my little comparison of means is shot to hell? Sigh.
I knew I should have taken statistics in college.
18 hours ago