Hi,
Am I correct in assuming that if a variable was found to be good predictor in univariate analysis, it should also be found to be a good predictor in multivariate analysis as multivariate takes into account multiple variables including interaction terms.
I am going through a research...