Stats & Regression
I'm stuck on a problem that requires me to find the model of best fit for a set of data.
I'm stuck for 3 reasons : a) I'm not sure whether the aim is to maximize the R^2 value or the F-score and b) I'm unsure how to treat a predictor with a coefficient of -1.352 and a p-value of 0.663 (is it significant or not?) c) I don't know how to assess the significance of interaction terms if one predictor is significant and the other isn't in the regression model before adding an interaction term.
I would really appreciate some help with this. Thanks in advance
I'm stuck for 3 reasons : a) I'm not sure whether the aim is to maximize the R^2 value or the F-score and b) I'm unsure how to treat a predictor with a coefficient of -1.352 and a p-value of 0.663 (is it significant or not?) c) I don't know how to assess the significance of interaction terms if one predictor is significant and the other isn't in the regression model before adding an interaction term.
I would really appreciate some help with this. Thanks in advance