I am just a bit confused about continuity correction! We use continuity correction while aproximating binomial or poison to normal.. But why don't we use continuity correction when we find difference between two proportions(in binomial) or two lamdas(in poisson) . Please reply as soon as possible Thank you! Shikha
But if we different n values.? I mean X1= 25, X2= 20 and n1= 100, n2= 150 I think dont they wil cancel out in this case.
You only apply a continuity correction of ±½ to the X's (regardless of the sample size) and since they'll both be in the same direction they'll cancel out.