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
- /
- SAS Programming
- /
- SAS Procedures
- /
- Will SAS calculate pooled variance?

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
- Highlight
- Email to a Friend
- Report Inappropriate Content

08-11-2009 01:02 PM

I have summary stats (mean, std dev, N) on 2 independent samples from the same population. I want to combine them, so I can compare this population to another one (with proc ttest).

I know the formula for pooled variance, and can thus use a data step to calculate the pooled variance for my two samples, but I'm wondering if there is an existing procedure where SAS will do this for you? Seems like something that would need to be done frequently enough....

Thanks!

I know the formula for pooled variance, and can thus use a data step to calculate the pooled variance for my two samples, but I'm wondering if there is an existing procedure where SAS will do this for you? Seems like something that would need to be done frequently enough....

Thanks!

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

08-11-2009 02:15 PM

You say you have the summary stats, but do you also have the raw data? There are a number of procedures that you can get a pooled variance out of with the raw data, such as GLM.

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

08-11-2009 02:24 PM

No - unfortunately, I don't have the raw data - I only have the summary stats (from a published paper.)

Thanks!

Thanks!

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

08-11-2009 02:48 PM

Well, since you are planning on using TTEST on summarized data, I assume you know how to do that. So, I'd try passing the summary statistics for your two groups through TTEST first. You aren't interested in the statistical comparison, but it will give you the pooled standard deviation (under method='POOLED').

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

08-11-2009 03:03 PM

Ahhh - okay - I'll try that - I've already started calculating the pooled vars in excel, but this will be a good check.

Thank you!

Thank you!

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

08-11-2009 02:39 PM

Then you are stuck with doing it manually or generating simulated datasets with those means and standard deviations and sample size to use in the SAS procedures.