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
- /
- Proc Glm

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

08-03-2017 05:05 PM

Hello,

I am doing multiple linear regression modeling and I was wondering if there is a way to report variance explained by each predictor aside from the variance explained by the entire model. Any help is much appreciated. Also, if there is another linear modeling procedure that can do this and handle multiple predictors that would be helpful too.

Best,

BIg Jer

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

Posted in reply to bigjer

08-03-2017 06:07 PM - edited 08-03-2017 06:14 PM

Variance explained by each predictor ... kind of a meaningless quantity unless you have an orthogonal designed experiment ... which it doesn't sound like you have

Nevertheless, it is the sum of squares for a model term divided by the degrees of freedom for the model term ... again totally meaningless in this situation ... feel free to ignore this value ... or in other words, the question cannot be answered for multiple linear regression unless you have an orthogonal experiment ... you can do the calculations, but you can't assign meaning to the numbers you calculate.

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

Posted in reply to bigjer

08-03-2017 07:15 PM

Perhaps you could rephrase the question to indicate something about what question you are trying to answer with these values.