If a researcher ran a regression at time 1 using 3 independent variables to predict Y (i.e., weight), is it possible to predict a change in weight (either gain or loss) at time 2?


TS Contributor
It's certainly possible to predict it, but the validity and accuracy of the prediction is a different question.

It depends on how "good" the predictive model fits the data, and how well the theory or hypothesis is framed so that predictions outside of the model bounds (extrapolations) will be reasonably accurate.