tag:blogger.com,1999:blog-5418994588621362010.post8058450733226755955..comments2023-11-15T10:08:56.796-08:00Comments on the blog formerly known as The Statistical Mechanic: how little do we knowWolfganghttp://www.blogger.com/profile/07086991199438418163noreply@blogger.comBlogger3125tag:blogger.com,1999:blog-5418994588621362010.post-81061829252396991552011-02-20T11:15:35.071-08:002011-02-20T11:15:35.071-08:00A brief look at this abstract suggests that the un...A brief look at <a href="http://www.sciencedirect.com/science?_ob=ArticleURL&_udi=B6V0M-3YMW4WM-C&_user=10&_coverDate=11%2F01%2F1999&_rdoc=1&_fmt=high&_orig=search&_origin=search&_sort=d&_docanchor=&view=c&_searchStrId=1648797231&_rerunOrigin=google&_acct=C000050221&_version=1&_urlVersion=0&_userid=10&md5=593a70c4493182dae633cc93359f5944&searchtype=a" rel="nofollow">this abstract</a> suggests that the uninformed prior causes indeed a bit of a problem.wolfganghttp://tsm2.blogspot.comnoreply@blogger.comtag:blogger.com,1999:blog-5418994588621362010.post-28816116616890575122011-02-20T11:06:26.276-08:002011-02-20T11:06:26.276-08:00I find it not too surprising that frequentists and...I find it not too surprising that frequentists and Bayesians disagree (by the way, how does a uninformed prior look like in this case?)<br><br>But I find it very surprising that the 'obvious' solution is <a href="http://en.wikipedia.org/wiki/Behrens%E2%80%93Fisher_problem#Welch.27s_approximate_t_solution" rel="nofollow">not really correct</a>.wolfganghttp://tsm2.blogspot.comnoreply@blogger.comtag:blogger.com,1999:blog-5418994588621362010.post-85833070867977215682011-02-20T09:41:39.862-08:002011-02-20T09:41:39.862-08:00All you really need to know about this problem:&qu...All you really need to know about this problem:<br>"Solutions to the Behrens–Fisher problem have been presented which make use of either a classical or a Bayesian inference point of view and either solution would be notionally invalid judged from the other point of view."<br><br>...a footnote to a philosophical problem.Chrisnoreply@blogger.com