Subscribe Bookmark RSS Feed

Why is regression output from JMP different from Excel?

diegoc11

Community Trekker

Joined:

Jan 12, 2012

Hey everyone,

I've attached an excel file that has some data (One dependent variable = WTP, and two independent variables).

The second tab has the regression output from excel, which we've used in classes at school all the time. What I don't understand is why regressing the same variable as least squares fit in JMP produces a different adjusted R^2, coefficients, p-values, etc.

Can anyone help me address any settings, etc. that might be taking place?

I'm using JMP 9 and excel's data analysis toolkit in Office 2010.

Thanks,

Diego

1 ACCEPTED SOLUTION

Accepted Solutions
ms

Super User

Joined:

Jun 23, 2011

Solution

I tried to reproduce the analysis in JMP and I got exactly the same numbers as in your file's second tab. Run the saved script of the model I used and look for what differ from the model you tried in JMP.

Fit Model(

          Y( :WTP ),

          Effects( :Name( "Frozen/Fresh" ), :Name( "Wild/Farmed" ) ),

          Keep dialog open( 1 ),

          Personality( "Standard Least Squares" ),

          Emphasis( Effect Leverage ),

          Set Alpha Level( 0.05 ), Run

);

3 REPLIES
ms

Super User

Joined:

Jun 23, 2011

Solution

I tried to reproduce the analysis in JMP and I got exactly the same numbers as in your file's second tab. Run the saved script of the model I used and look for what differ from the model you tried in JMP.

Fit Model(

          Y( :WTP ),

          Effects( :Name( "Frozen/Fresh" ), :Name( "Wild/Farmed" ) ),

          Keep dialog open( 1 ),

          Personality( "Standard Least Squares" ),

          Emphasis( Effect Leverage ),

          Set Alpha Level( 0.05 ), Run

);

diegoc11

Community Trekker

Joined:

Jan 12, 2012

Thanks MS,

It looks like my excel isn't set up correctly.

Cheers mate!,

Diego

larryworley

Community Member

Joined:

Dec 15, 2011

How different are your results? 

Since Excel treats numeric data differently than either JMP and SAS and since it also has its own regression algorithms, I would not be suprised to see differences.  If differences are minor, then I would suspect these might account for the differences.

If the differences are large, I would suspect an issue with Excel and not with your excel setup.

Can you post your JMP results?