Comments on: Unexpectedly, the theoretically best reject-region of T-test is bounded. http://lixiaoxu.lxxm.com/unexpectedlythetheoreticallybestreject-domainoft-testisbounded/ Teaching notes of szpku dot lixiaoxu at gmail dot com Thu, 09 Sep 2010 13:03:10 +0000 http://wordpress.org/?v=2.7 hourly 1 By: liutom http://lixiaoxu.lxxm.com/unexpectedlythetheoreticallybestreject-domainoft-testisbounded/comment-page-1/#comment-91 liutom Mon, 18 Aug 2008 04:37:43 +0000 http://lixiaoxu.lxxm.com/2008/05/11/unexpectedlythetheoreticallybestreject-domainoft-testisbounded/#comment-91 厉害 厉害

]]> By: lixiaoxu http://lixiaoxu.lxxm.com/unexpectedlythetheoreticallybestreject-domainoft-testisbounded/comment-page-1/#comment-84 lixiaoxu Sun, 11 May 2008 01:33:37 +0000 http://lixiaoxu.lxxm.com/2008/05/11/unexpectedlythetheoreticallybestreject-domainoft-testisbounded/#comment-84 The complication dwells in that the alternative hypothesis is single. If the alternative are multiple, that is, [tex] \mu>0[/tex], the theoretically best reject-region is still [tex] T > criterion[/tex]. Multiple alternatives usually involve the bias problem. The best reject-region for one single alternative may be inappropriate (biased) for the other single alternative. The appropriate RD set for multiple alternatives is a strict subset of the appropriate RD set for a single alternative. The complication dwells in that the alternative hypothesis is single. If the alternative are multiple, that is, \mu>0, the theoretically best reject-region is still T > criterion.

Multiple alternatives usually involve the bias problem. The best reject-region for one single alternative may be inappropriate (biased) for the other single alternative. The appropriate RD set for multiple alternatives is a strict subset of the appropriate RD set for a single alternative.

]]>