For f-tests(ANOVA design) before an experiment



所有跟贴·加跟贴·新语丝读书论坛http://www.xys.org/cgi-bin/mainpage.pl

送交者: cornbug 于 2005-8-21, 09:17:51:

回答: gee, I should go to bed soon 由 xj 于 2005-8-21, 01:07:22:

to estimate the sampling size for a power greater than 0.8 and alpha = 0.05.I remember that power is a measure of the probability of making a type II error if the null hypothesis cannot be rejected.I may be wrong.

The trick is that U don't have enough preliminary data for an power analysis.In some cases, post hoc is necessary to readjust the size while you are doing the experiment. Again, as you said before it is better to have an authorized power stat package to do it so that every one will believe you in your estimated smapling size.

I am doing the former in your last question.

Thank you very much for your help



所有跟贴:


加跟贴

笔名: 密码(可选项): 注册笔名请按这里

标题:

内容(可选项):

URL(可选项):
URL标题(可选项):
图像(可选项):


所有跟贴·加跟贴·新语丝读书论坛http://www.xys.org/cgi-bin/mainpage.pl