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
- :
- Method for calculation of degrees of freedom?

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

Highlighted

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

Oct 18, 2013 1:23 AM
(9995 views)

Hi!

I was wondering if anyone knows how JMP calculate degrees of freedom? I think SAS is using the Satterhwaite approximation, but I cannot find any information about if JMP use the same?

Thanks!

1 ACCEPTED SOLUTION

Accepted Solutions

Highlighted

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

Hi Sandra,

From Chapter 7 of Fitting Linear Models:

Degrees of Freedom

The degrees of freedom for tests involving only linear combinations of fixed effect parameters are calculated using the first‐order Kenward and Roger correction. So JMP’s results for these tests match PROC MIXED using the DDFM=KENWARDROGER(FIRSTORDER) option. If there are BLUPs in the linear combination, JMP uses a Satterthwaite approximation to get the degrees of freedom. The results then follow a pattern similar to what is described for standard errors in the preceding paragraph.

For more details about the Kackar‐Harville correction and the Kenward‐Roger DF approach, see Kenward and Roger (1997). The Satterthwaite method is described in detail in the SAS PROC MIXED documentation (SAS/STAT 9.2 User’s Guide, Chapter 56).

-Jeff

-Jeff

4 REPLIES 4

Highlighted
##

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

Re: Method for calculation of degrees of freedom?

Hi Sandra,

Which JMP platform are you looking at?

If it's Fit Model, which personality are you using?

-Jeff

-Jeff

Highlighted
##

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

Re: Method for calculation of degrees of freedom?

I am working in JMP 9.0.

I just open my data table, then go to analyze - fit model. In the box that opens, I specify the different parameters:

1) the dependent variable in Y below "Pick Role Variables"

2) and other variables below "construct model effects". These are both continuous and nominals, one is random and there are at least 2 interactions included in the original model.

The other settings I use are:

Personality: Standrad Least Squares

Emphasis: Minimal Report

Method: REML (Recommended)

Unbound Variance Component are also used.

Thanks in advance!

/Sandra

Highlighted

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

Hi Sandra,

From Chapter 7 of Fitting Linear Models:

Degrees of Freedom

The degrees of freedom for tests involving only linear combinations of fixed effect parameters are calculated using the first‐order Kenward and Roger correction. So JMP’s results for these tests match PROC MIXED using the DDFM=KENWARDROGER(FIRSTORDER) option. If there are BLUPs in the linear combination, JMP uses a Satterthwaite approximation to get the degrees of freedom. The results then follow a pattern similar to what is described for standard errors in the preceding paragraph.

For more details about the Kackar‐Harville correction and the Kenward‐Roger DF approach, see Kenward and Roger (1997). The Satterthwaite method is described in detail in the SAS PROC MIXED documentation (SAS/STAT 9.2 User’s Guide, Chapter 56).

-Jeff

-Jeff

Highlighted
##

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

Re: Method for calculation of degrees of freedom?

Ok, thank you for you help!

/Sandra