No more stats questions from me for a while. What are the odds? No wait, no need to answer that, it is just an expression. Variables that uniquely identify margins: locationgroupOn another note, I am a student in Albany. ( 3) 1b.locationgroup - 4.locationgroup = 0 ( 2) 1b.locationgroup - 3.locationgroup = 0 ( 1) 1b.locationgroup - 2.locationgroup = 0 ( 2) 3b.rep78 - 5.rep78 = 0Ĭode: Logistic regression Number of obs = 40 Log likelihood = -18.767412 Pseudo R2 = 0.5574įoreign | Odds Ratio Std. Instead of such a post-hoc test, I recommend that you examine confidence intervals. ![]() Just to state one paradox: if you find that one of the pairwise calculations (say \(p_1\) and \(p_4)\) is significant at p=0.05 (with or without correction for multiple tests), it will turn out that the pre-study power of detecting the observed difference was about 50%. However, since you've already done this analysis, then a post-hoc calculation can be misleading. If you have continuous covariates, then estimate proportions at the covariate means It's probably easier to use Stata's power twoprop than it is use powercal. I suggest that you do the power calculations for each of the pairs of categories (1 vs 2, 1 vs 3, 1 vs 4, 2 vs 3, 2 vs 4, 3 vs 4. However this second calculation requires expert knowledge. test require 1) that you guess response proportions in each category: and 2) that you form what is known as the non-centrality parameter. Power calculations for the overall 3 d.f. You have done the calculation for a one parameter test of the effect of a different variable. This is analogous to the F test in a four-group ANOVA. The overall test is the "all" test Example, which will have 3 degrees of freedom for the null hypothesis of no difference between the four groups. The mtest() option is needed to correct for multiple tests of pairwise differences) Result to be calculated is power in variable: powerageĬode: mtest i.locationgroup.1 = i.locationgroup.2 = i.locationgroup.3 = i.locationgroup.4, mtest(sidak)(See example below. Variables that uniquely identify margins: locationgroup margins locationgroup, at( ageatoperation=generate(26(1)94)) logit whc ageatoperation bmikgm2 i.locationgroup
0 Comments
Leave a Reply. |
Details
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |