turn on suggestions

Auto-suggest helps you quickly narrow down your search results by suggesting possible matches as you type.

Showing results for

Find a Community

- Home
- /
- Analytics
- /
- Stat Procs
- /
- Surveymeans and confidence limits

Topic Options

- Subscribe to RSS Feed
- Mark Topic as New
- Mark Topic as Read
- Float this Topic for Current User
- Bookmark
- Subscribe
- Printer Friendly Page

- Mark as New
- Bookmark
- Subscribe
- Subscribe to RSS Feed
- Permalink
- Email to a Friend
- Report Inappropriate Content

05-06-2011 04:52 AM

I have a question about the estimation about confidence limits and proc surveymeans.

When we estimates the confidence limits, it seams that SAS does not take the df in account? It makes just a point estimate with SD * 2.04.

Is this the correct behavior? or is there a way to make SAS adjust the 2.04 coefficint according to the df and the numbers in the t-tables?

When we estimates the confidence limits, it seams that SAS does not take the df in account? It makes just a point estimate with SD * 2.04.

Is this the correct behavior? or is there a way to make SAS adjust the 2.04 coefficint according to the df and the numbers in the t-tables?

- Mark as New
- Bookmark
- Subscribe
- Subscribe to RSS Feed
- Permalink
- Email to a Friend
- Report Inappropriate Content

Posted in reply to AndersBergquist

05-09-2011 08:55 AM

SURVEYMEANS does take df into account. The df calculation depends on whether you have clusters and/or strata. In the simplest case, with no clusters or strata (which means that each observation is a cluster), df = number of observations - 1. For your example, 2.04 IS the critical t value (with alpha=.05). By the way, CI is SE*t (not SD*t).

You can see the calculated df value by putting df in the procedure statement.

You can see the calculated df value by putting df in the procedure statement.