Can anyone tell me how I can tell if a distrubution is nonnormal when I am trying to construct a confidence interval.

Beyond the scope. If you are trying to construct a C.I. for the mean of the distribution, then for big enough n X-bat is normally distributed and then just use usual stuff. Otherwise, testing for normality and what to do about non-normal distns is not part of this material.

SO the question would tell us its a nonnormal distrubution?

It would not ask you to construct a C.I. unless the data is approximately normally distributed (e.g. t-Dist’n) They will usually tell you the (population or sample) mean as well as the (population or sample) variance. They might also tell you the number of observations. It is your job to interpret the data and construct the proper C.I.

supersharpshooter Wrote: ------------------------------------------------------- > It would not ask you to construct a C.I. unless > the data is approximately normally distributed > (e.g. t-Dist’n) > > They will usually tell you the (population or > sample) mean as well as the (population or sample) > variance. They might also tell you the number of > observations. > > It is your job to interpret the data and construct > the proper C.I. IF THEY GIVE YOU VARIANCE, DO NOT FORGET TO GET THE SQUARE OF THAT B/C THE CI EQUATION ASKS TO TO INPUT ST DEV, NOT VARIANCE. THANKS ST DEV = SQ RT OF VARIANCE VARIANCE = ST DEV ^ 2