On 24 Nov 2002 22:29:08 GMT, [EMAIL PROTECTED] (Radford Neal) wrote: > > >> But what if the sample standard deviation that you computed to do your > >> t test is substantially less than what you think sigma is? Is it really > >> "conservative" to ignore this, and quote a result that is based on what > >> you have reason to believe is an optimistic estimate of the accuracy of > >> the observations? > > In article <[EMAIL PROTECTED]>, > Rich Ulrich <[EMAIL PROTECTED]> wrote: > > >Use the improved estimate and z -- That's what I do informally, > >and small variance is reason enough to dis-believe some data. > >I don't know that I have seen anyone publish using z that way. > > I think you may not have understood the scenario I was imagining. I'm > supposing that the sample standard deviation is lower than what you > think sigma is. Perhaps you might improve your estimate of sigma with > this extra data, but your estimate wouldn't change to match the sample > standard deviation, if you had good reason for your original belief. > Using the sample standard deviation in a z test would be even worse > than just doing the t test, giving an even more mis-leading p-value. >
oops, I should have been more clear. My "improved estimate" was intended to indicate, "the external estimate/ guess". If I really don't like the observed SD, I don't think I will be much happier with the Bayesian use of it -- unless I weight it far less than (I think) Bayesians are prone to weight it. -- Rich Ulrich, [EMAIL PROTECTED] http://www.pitt.edu/~wpilib/index.html . . ================================================================= Instructions for joining and leaving this list, remarks about the problem of INAPPROPRIATE MESSAGES, and archives are available at: . http://jse.stat.ncsu.edu/ . =================================================================
