Quantcast
Channel: Statalist
Viewing all articles
Browse latest Browse all 73277

Prtest in svy mode

$
0
0
Hi

I have a question about testing proportions in STATA. See table below.

Basically, I am comparing rates of treatment benefit in different conditions. They are not mutually exclusive, as an individual may have benefited from treatment both for a neuro and MSK condition.

I am reporting percentages.


I am in survey mode, and am having trouble figuring out how to best test that the proportions' differences are statistically significant.

I am most interested in showing that MSK is different than the others.

1) What is the best statistical test for this and how to do it?

2) Is there any way to do a global test of proportions (an F-test I suppose)?

3) Is there any way to directly compare musculoskeletal to the next highest category such as:


ttest musculoskeletal = miscellaneous

Thanks so much.

Ezra

Total N = 2,616,637 Used treatment for a specific problem and did not benefit a great deal Used treatment for a specific problem and benefited a great deal
Estimated US population size 1,456,420 (55.7) 1,160,217 (44.3)
Musculoskeletal 32.35 67.65
Miscellaneous 40.12 59.88
ID 44.62 55.38
Allergy 51.17 48.83
Neurological 52.11 47.89
Psychiatric 55.64 44.36
Gastrointestinal 59.89 40.11
Endocrine 68.92 31.08
Gynecological 74.71 25.29
Respiratory 77.76 22.24
Dental 79.95 20.05
Hematological (N= 0)
Cardiovascular (N=0)


Thanks
Ezra

Viewing all articles
Browse latest Browse all 73277

Trending Articles



<script src="https://jsc.adskeeper.com/r/s/rssing.com.1596347.js" async> </script>