

Path Analysis with different groups a... 

Message/Author 

Orig posted on Sunday, September 02, 2012  9:07 pm



Hi, i want to run a path analysis and i got one dataset, for three groups: treatment_1, treatment_2, control the three groups only differ in an exogenous shock _________________________________________ the data contain the groups in form of two dummy variables as follows: treat12 = 0 (treatment 1) treat12 = 1 (treatment 2) t1co = 0 (treatment 1) t1co = 1 (controlgroup) _________________________________________ Treatment 2 is in the focus of the analysis. What i already got is a fitted model for the "Treatment 2Analysis". My problem is: I want to know, if the effect of treatment 2 generated in a variable, differs from treatment 1, but i donīt know how to do this, when i just got these two dummies. example: treat2 > x > y to account for treat1 : treat1 > y but i dont know how to "build" the variable: treat1 


It looks like you do your dummy variables incorrectly. Try: DEFINE: tx1 = 0; IF (treat1 EQ 1) THEN tx1=1; tx2 = 0; IF (treat2 EQ 1) THEN tx2=1; MODEL: y ON tx1 tx2; The difference in the slopes is the difference in the treatment effects. 

Back to top 

