Moderating impact

[Omar Shaher]

Dear Statalist,
Actually, I am bit confused between the joint effect and the moderating impact. I mean I saw some people doing the following example:

Cancer = Smoker + Sport + Smoker*Sport.

So the code in STATA will be if am using panel data:

Xtologit Cancer i.Smoker##i.Sport

So, I believe the above is the joint effect.

But, what if I want to see the moderating impact of Sport on the relationship between Cancer and Smoker. How the model should look like and what is the code for this by using the above variables?

Many thanks in advance.

[Rich Goldstein]

the above is the moderating effect (interaction); I'm not sure what you mean by "joint effect" - do you have a cite defining this term or can you do so yourself?

[Omar Shaher]

Thanks very much, Rich for the answer. The joint effect means the combined effect of two independent variables on the dependent variable. I mean when we interact two independent variables, we cannot know which variable is affecting which relationship. I mean based on the above example, we can't know if the sport affects the relationship between smoker and cancer OR smoker affecting the relationship between sport and cancer. I hope this makes sense.

Here are two links about the joint effect:

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3018505/

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2752945/

[Omar Shaher]

I have found the following paper
Sharma, S., Durand, R. M., & Gur-Arie, O. (1981). Identification and analysis of moderator variables. Journal of marketing research, 18(3), 291-300.

After reading the paper, and as per my understanding that there are two types of moderators, the first one is called pure moderator and the second one called quasi moderator. So, I THINK, if I understood this correctly, that the pure moderation happens when the moderator is not signifiancant while the interaction term is significant, while the quasi moderation happens when the moderator and the interaction term are singifcant.
You can see that on page 295.

If you have any other ideas to clarify more, I would grateful.