turn on suggestions

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

Showing results for

- JMP User Community
- :
- Discussions
- :
- Discussions
- :
- Need Help - Making forward Stepwise Regression Out...

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
- Get Direct Link
- Email to a Friend
- Report Inappropriate Content

Sep 7, 2009 4:19 AM
(941 views)

Hello,

I am trying to do a forward stepwise regression, only the estimates turn out to be negative, which does not make any sense since the variables I am using cannot get a negative value.

Is there any way to restrict the regression to output only positive values??

Thank you in advance,

Jonathan.

I am trying to do a forward stepwise regression, only the estimates turn out to be negative, which does not make any sense since the variables I am using cannot get a negative value.

Is there any way to restrict the regression to output only positive values??

Thank you in advance,

Jonathan.

1 REPLY

- Mark as New
- Bookmark
- Subscribe
- Subscribe to RSS Feed
- Get Direct Link
- Email to a Friend
- Report Inappropriate Content

Sep 8, 2009 10:02 AM
(926 views)

Do you mean parameter estimates?

The negative parameter estimates are indeed possible, even when your data is positive.

Negative parameter estimates means that as your independent variable increases, the value of the dependent variable decreases.

There is no way (nor is it a good idea) to force the parameter estimates to be positive. For that matter, it isn't considered a good idea (by most statisticians) to do stepwise regression.

The negative parameter estimates are indeed possible, even when your data is positive.

Negative parameter estimates means that as your independent variable increases, the value of the dependent variable decreases.

There is no way (nor is it a good idea) to force the parameter estimates to be positive. For that matter, it isn't considered a good idea (by most statisticians) to do stepwise regression.