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
- /
- Statistical test for binary independent variables

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

05-10-2016 02:18 PM

Dear SAS experts,

Could you please tell me the statistical test that suits for binary independent variables and binary dependent variable? Would logistic regression be appropriate?

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

Posted in reply to Babloo

05-10-2016 02:27 PM

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

Posted in reply to Babloo

05-10-2016 02:29 PM

Could you tell us what you are testing for?

A common test is for association. For tests of association, you might want to run a chi-square test by using PROC FREQ.

You can use PROC LOGISTIC if you want to create a model that predicts the probability of the response variable for each level of the explanatory variable.

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

Posted in reply to Babloo

05-11-2016 02:53 AM

One more way is using Decision Tree, if you have it .

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

Posted in reply to Babloo

05-12-2016 03:22 PM

For most purposes, PROC LOGISTIC is the procedure to use. It is definitely recommended, rather than CATMOD, for logistic modeling. This note describes the many possibilities.

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

Posted in reply to Babloo

05-13-2016 08:49 AM

If you want to build a predictive model than proc logistic would be suited. There are several other statistical tests for categorical data .so ultimately depends upon what you are testing for .