SAS Data Integration Studio, DataFlux Data Management Studio, SAS/ACCESS, SAS Data Loader for Hadoop and others

How to Compare more than two Regression Slopes in SAS

Reply
New Contributor
Posts: 3

How to Compare more than two Regression Slopes in SAS

 

I am using SAS 9.4 TS level 1MZ.on window version of 6.1.7601. I went to compare more than two regression lines so anyone could help to do this please? or anyone share this syntax would be highly appreciated.

Thanks in advance!

Super User
Posts: 11,343

Re: How to Compare more than two Regression Slopes in SAS

First, which regression procedure are you using? It may helps us to provide some of the code you are running to give us an idea of what may be practical.

 

Second what kind of comparison are you looking to do? Graphs? Differences of slopes? Differences of Intercepts? Differences of R-squares?

New Contributor
Posts: 3

Re: How to Compare more than two Regression Slopes in SAS

At the movement I used simple liner regression. I measured plant height
over different genotypes and treatments in some hour interval in the growth
stage of the plant and I went to see the effect of days on plant height
growth.

Regards to the type of comparison I am looking is Differences of slopes,
Differences of Intercepts but if you give me some more relay I am happy
especially on nonlinear regression models. I do have other parameters which
have nonlinear regresses on dates from planting.
Thanks in advance!
##- Please type your reply above this line. Simple formatting, no
attachments. -##
Respected Advisor
Posts: 4,919

Re: How to Compare more than two Regression Slopes in SAS

Here is an example. The slopes are compared as the interaction term between engine size and origin in the ANOVA table, and the predicted values are compared for a 4-liter engine in the LSMeans output  :

 

proc glm data=sashelp.cars;
class origin;
model horsepower = origin|engineSize;
lsmeans origin / at engineSize=4 pdiff; /* At 4-litre engine size */
run;

(for an illustration, look at the ANCOVA graph)

PG
Ask a Question
Discussion stats
  • 3 replies
  • 219 views
  • 0 likes
  • 3 in conversation