I have two continous explanatory variables. I assume that the interplay (joint effect) of these variables can predict a third, continous response variable.
What statistical test can be used to test whether the two factors and the response have a statistically significant relationship?
The actual project is aiming to find out whether the interplay of group size and cognitive abilities of members in a primate society can predict the centralisation of the social network.
Dear Martin,
Sounds like you are interested in what is commonly called an interaction effect. You can test this in a multiple regression where you input the two continuous explanatory variables and their interaction as predictors and the response variable as dependent variable.
To do this, you need to create the interaction term:
1. demean/center both continous explanatory variables (i.e. compute the mean for each variable and substract the mean from the individual values of each variable).
2. compute the interaction term by simply multiplying the demeaned variables (Xdemeaned x Ydemeaned).
3. Then use the demeaned variables and the interaction term as predictors in a multiple regression as stated in the beginning.
If the interaction effect gets significant, you can run further analyses to understand how the interaction looks like and how it can be interpreted.
This can be done via a so-called simple-slopes analysis - Here is a very useful tutorial on that issue (done in SPSS, but same logic in other packages):
https://www.youtube.com/watch?v=ighASZf3eyA
Best, Stefan
Dear Martin,
You have to use regression analysis. Your IV will be the two continous factors and its interaction and your DV your continous response variable. You can use SPSS or Psych R Package.
Best
Regression analysis can be involved in such investigations. Then, further deep analysis can be involved just to approve your analysis.
Good luck