Subscribe Bookmark RSS Feed

interpreting report for cross product in regression

shahin

Community Trekker

Joined:

Jun 23, 2011

Hello,

I have seen similar questions to mine here, but I am still a bit confused. I am doing an ordinal logistic regression using the fit model function, to compare the slopes of 3 regressions (for 3 different groups). I wish to see if the slopes of these regression lines are significantly different or not. I have included here the two sections of the report I wish to interpret. Can someone tell me what are the 3 slopes, and how I figure out which ones are significantly different from each other (or not) ?

thanks

Parameter Estimates

Term Estimate Std Error ChiSquare Prob>ChiSq
Intercept[0] 2.14767299 0.0995951 465.01 <.0001
x -0.0122746 0.1072329 0.01 0.9089
Watershed 1 -0.0036705 0.1434278 0 0.9796
Watershed 2 -0.3389146 0.1260122 7.23 0.0072
Watershed 1*x -0.1632537 0.1617892 1.02 0.3129
Watershed 2*x -0.0676399 0.1267653 0.28 0.5936




Effect Likelihood Ratio Tests

Source Nparm DF L-R ChiSquare Prob>ChiSq
x 1 1 0.01309947 0.9089
Watershed 2 2 8.7829368 0.0124
Watershed*x 2 2 2.00804932 0.3664
2 REPLIES
shahin

Community Trekker

Joined:

Jun 23, 2011

sorry, the spacing in the table is terrible
shahin

Community Trekker

Joined:

Jun 23, 2011

Hi,

I've tried to space out the table better, although it's damn hard to copy it here from another program. Hope someone can help. thanks

Parameter Estimates
>
> Term ---------------Estimate-------------- Std Error --------ChiSquare-------Prob>ChiSq
> Intercept ----------2.14767299-------------0.0995951--------465.01-----------0.0001
> x --- ----- ---------0.0122746 --------0.1072329--------------- 0.01----------0.9089
> Watershed 1 ------0.0036705-----------0.1434278 -------------0-------------0.9796
> Watershed 2-----------0.3389146 -----------0.1260122---------7.23-------------- 0.0072
> Watershed 1*x ---------0.1632537-------------0.1617892--------1.02 ------- 0.3129
> Watershed 2*x ------------ -0.0676399------0.1267653----- 0.28----------.5936
>
>
>
>
> Effect Likelihood Ratio Tests
>
> Source ----------Nparm------ DF-----L-RChiSquare-------- Prob>ChiSq
> x --------------- 1------------------1------ ---------0.01309947------------ 0.9089
> Watershed--------2-----------------2 ---------8.782936------------- 0.0124
> Watershed*x -------2----------------2 ------------------2.0080493--------------0.3664