User states, “If I want to do 1-to-1 regression for 20 statements, it means that I will have to run it 20 times. In SPSS, I can write an excel macro to write the SPSS syntax to run it 20 times so I'm wondering if I can do something similar in Q. One reason why I do not like multiple regression is because of multicollinearity and results can be unstable wave on wave. Shapley value can deal with that problem but from my experience, the coefficients of different independent variables in the model become very similar (especially when there is high correlation) since the coefficients are calculated based on averaging the coefficients from so many models. Hence why I would like to stick to 1-to-1 regression. I can easily do it in excel with the slope() function. Only downside is that I can't use it easily if I need the data weighted.”