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
- :
- Comparing Regression Curves

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

Jun 25, 2009 6:51 PM
(1193 views)

I've looked at literature and there are 2 ways to do this.

1. Compare the slopes and see if they are equal; then compare the intercepts and see if they are equal. Using the t-test to determine if the slopes and intercepts are equal. If both the slopes and intercepts are equal the lines are coincident.

2. Create a single model and with a dummy variable (Z) where Z = 1 for the first data set and Z = 0 for the second data set. Which would give y = a + bX +cZ + dXZ. Where a, b,c, d are coefficients, X is the independent variable, and Z is the dummy variable. You then perform a multiple partial F-test to determine if the lines are coincident by comparing the model with the Z and XZ terms to the model without those terms.

Is there an easy way to do either of these tests with the JMP software?

So far I have fit a model with X, Z, and XZ (with the center polynomials checked). Looking at the effects tests the Z has an F ratio of 31 and Prob > F is <0.0001. The X*Z has an F ratio of 0.3178 and Prob > F is 0.5765. I am using 0.05 significance level with 21 and 18 degrees of freedom for the 2 sets of data.

1 ACCEPTED SOLUTION

Accepted Solutions

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

Jun 26, 2009 4:01 AM
(1267 views)

Solution

I think you are on the right track using Analysis of Covariance. Your result indicates that the slopes do not differ significantly (XZ-term) but the levels ("intercept") differ between the two data sets (Z-term).

In JMP you do not need to explicitly create a 0/1 dummy variable to do this. You can just have a nominal variable with the name or ID of the two data sets.

See example under "Analysis of Covariance with Separate Slopes" in Chapter 12 in the manual (JMP Stat Graph Guide, page 250 for JMP 8)

In JMP you do not need to explicitly create a 0/1 dummy variable to do this. You can just have a nominal variable with the name or ID of the two data sets.

See example under "Analysis of Covariance with Separate Slopes" in Chapter 12 in the manual (JMP Stat Graph Guide, page 250 for JMP 8)

2 REPLIES

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

Jun 26, 2009 4:01 AM
(1268 views)

In JMP you do not need to explicitly create a 0/1 dummy variable to do this. You can just have a nominal variable with the name or ID of the two data sets.

See example under "Analysis of Covariance with Separate Slopes" in Chapter 12 in the manual (JMP Stat Graph Guide, page 250 for JMP 8)

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

Jun 26, 2009 6:54 AM
(1007 views)

> the JMP software?

I don't think there is an easier way to do this.