Test the indirect effect

[Matthew_Xu]

Hello mates! 

I am now doing a research with an unbalanced panel dataset.  The model I apply is as follow:

Y=a*X+b*X^2+c*other control variables.

However, I think X can also affect Y via other channel, which is : X-->Z-->Y (this is called indirect effect). And i am now confused on what kind of procedure or codes that can help me to identify whether this indirect effect exists or not?

Since i am very fresh towards SAS, hoping the one who can help me can provide some explanation on the code used.

Much appreciate!

[koyelghosh]

I think, may be, what you are referring to as Z being a confounding variable (see wikipedia and this article). Most likely you are aware of how to take care of confounding variables, else this article can be a good starting point for knowing ways to deal with a confounding variable.

You might have to use different methods to find if the indirect effect/confounding variable, Z, exists. I am no expert in SAS but I will not be surprised if there is not a well defined single PROC to tell you with certainty about the same (if Z is a confounding variable/indirect effect).