Can you say "predict" in a regression write-up if your data is cross-sectional? Or should you stick to language such as X is "associated" or "related to" with Y.

Does predict always imply change over time?

  • 4
    $\begingroup$ There are few rules in predictive modeling and regression. So, no, prediction is not time dependent and, yes, you can use it to describe cross-sectional data. $\endgroup$
    – user78229
    Jul 24, 2023 at 11:07
  • 2
    $\begingroup$ You can always predict what potentially could have been observed. After all, the whole point of a probability model is to describe such potentially observable data. Sure, once it has been observed that's all there is, but other data could have occurred. If not, don't use a probability model. $\endgroup$ Jul 24, 2023 at 15:23


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Browse other questions tagged or ask your own question.