I used 3 construct to measure the single variable (Board diversity). The results of the 2 construct show -sig(as above), the last one show +insignificant. But, put together the overall result indicated +significant (0.0156).
My question is that: is it theoretically and statistically feasible. Thank you
based on not much information, I will hazard a guess. IMO the result says c2 lowers the DV and and you can't say about the last one. The overall says you have had some success predicting the DV. However since this is with almost no information as a reviewer I would reject the paper because you have given so little information. p-values and what I assume are signs on regression coefficients are not the entire story. For examples, see the attached.. Best, D. Booth
I am agree with David Eugene Booth, If the theory related to your code and simulation is not available, it is hard to be review in short time. I think, in such case your supervisor may be the best Doctor to give you some advise.
In GR-forum, mostly you get some general review for your general issue/problem, not for special issues!!! (special issues need sufficient time for investigation).
The main reason for the question is to know if the scenario ( 2 negative sig variables + 1 positive significant) is statistically possible to give overall result of positive significance when we run 3 variables together.
At times some mentors may not have good statistical knowledge of certain software, so I need expert assistance.