I am working on a mediated model where M mediates the relationship between X and Y and I have one control variable. The data I am using is from 3 different countries with different sample sizes (46, 64 and 29, total n=139). I created country dummy variables where C1 is 1 for the first country and 0 for the other two, C2 is 1 for the second country and 0 for the first and last and included both in my model. When I ran my analysis (using PLS-SEM in SmartPLS), I got significant results for the main and mediated effects.

enter image description here

My question is, can I conclude that my results can be generalized across countries? and does the significant effects of C1 and C2 on Y matter?

Any further analysis I should consider?


  • $\begingroup$ What is the scientific question you're trying to answer? Right now, all that the C1 and C2 dummies are doing is reducing residual variability--your model does NOT, for example, allow for different mediation effects by country. $\endgroup$ – Sheridan Grant Mar 11 '20 at 5:40
  • $\begingroup$ Thanks for your comment @SheridanGrant. I tested the model first using one sample and the results were good (main effect and moderation are significant). Now I want to test if the same relationships apply to other countries. My point is to generalize my results. $\endgroup$ – Sara Mar 11 '20 at 8:20
  • $\begingroup$ Then you need interaction terms, between C and M (and possibly more). Could be a complicated analysis. $\endgroup$ – Sheridan Grant Mar 12 '20 at 7:09

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.