Confidence interval, does a difference in sample size affect comparison of results?

#1
I have a time series and im trying to calculate the confidence interval (95%) for the time taken to experience a +ive or -ive move of X%. I would like to do this for several X's, ie 10%, 30%, 50%.

Am i correct in assuming that i should try and ensure that each level of X% has the same/similar sample size n, in order to do a fair comparison between the different X's, or would it not really affect the results that much?
 
#2
Re: Confidence interval, does a difference in sample size affect comparison of result

A confidence interval for an average allready has the sample size in the formula, does that help?