Independent variables

[yael]

Hello,

I need your advice,

I have 3 groups in my research: Firms that founded on 1930 (FF - dummy variable), Firms that founded on 1970 (FNF - dummy variable), and other firms.

I also have the continous variables that describe the largest ownership in the relevant firm, FOWN for FF firm and FNFOWN for FNF firms.

If I want to check the ownership impact on performance, should I run regression as:

proc reg data=sasuser.sasfile;

model performance = FOWN FNFOWN ;

run;

Or should I create a new variable named, for example, OWN, that reflect that largest owners in each firm but the code will be:

proc reg data=sasuser.sasfile;

model performance = OWN*FF OWN*FNF ; (I will create a new variable for each duplicated variables)

run;

Thanks for your help!!!!

[PeterClemmensen]

I am assuming that your dummy variables are coded as 0/1 as PROC REG does not support a class statement.

I would recommed your first choice. You can also specify multiple model statements like this

proc reg data=sasuser.sasfile;
    model performance = FOWN;
    model performance = FNFOWN;
run;

[yael]

Thanks for your answer

But my question is what is the difference between the two options I mentioned?