14 votes

Why does the normality assumption not affect Linear Regression in large samples?

The assumption of the Normality of the error term in a regression that applies Least-Squares estimation methods, is used to make statistical inferences about the coefficients after estimation, it is ...
Alecos Papadopoulos's user avatar
4 votes

non-significant p value in a multivariable cox regression following exhaustive model selection

There are lots of problems with "exhaustive model selection" and with stepwise. These have been covered here many times. I am not hugely familiar with glmulti (it appears to do some kind of ...
Peter Flom's user avatar
  • 117k
3 votes

Why does the normality assumption not affect Linear Regression in large samples?

And is that for both prediction and inference (i.e getting BLUE estimators)? No, not for both The estimator for the coefficients will approach a normal distribution, independent of the distribution ...
Sextus Empiricus's user avatar
2 votes

Can I predict values from a multiple regression model with fewer predictors than there are in the model?

This problem was studied here with several solutions offered. One of the solutions is based on fast backwards (approximate) step down variable selection to fit a temporary submodel using only the ...
Frank Harrell's user avatar
1 vote

models MIMIC, predicted probabilities?

This is a slightly weird thing to want to do, because you're interested in the predicted level of a categorical variable, given the level of a factor. But the level of the factor is determined by the ...
Jeremy Miles's user avatar
  • 17.5k
1 vote

What metric should I use for a Regression model with a gamma distributed target?

Use the (negative) likelihood of the distribution as your loss function. You can also turn it into a pseudo r2 for easier interpretability ( and negative likelihood is a 1:1 relation to pseudo r2). I ...
Georg M. Goerg's user avatar
1 vote

R=X*Y is the relationship. Is predicting R and X and obtain Y same as predicting X and Y to obtain R?

The equation states equality, not causality The equals sign gives you equality, nothing more, nothing less. You have three numbers that are in some relation to one another that's defined by the ...
Scriddie's user avatar
  • 2,084

Only top scored, non community-wiki answers of a minimum length are eligible