topic Adjusted R square in Statistical Procedures

Adjusted R square

pawandh — Sun, 14 Feb 2016 06:02:02 GMT

Adj R2=1- (1-r2)(N-1)/N-p-1

where N=total sample size

p=no of predictors

r-coeff of determination

If i get a negative adj r2 then what can i conclude from that??

Re: Adjusted R square

Norman21 — Sun, 14 Feb 2016 10:04:03 GMT

The predicted R-squared helps you find out how well the model fits the original data. Generally, if this is low (approaching zero, or negative), then your model is not very good.

"Adjusted" R-squared lets you compare regression models with different numbers of predictors.

So, to answer your question, your model is not very good and you should try another.

Norman.

Re: Adjusted R square

Rick_SAS — Sun, 14 Feb 2016 20:26:24 GMT

It is possible. It usually means that you have many explanatory effects compared to the number of observations. I can't remember the rule-of-thumb right now, but some experts recommend a minimum of 10-20 observations per regressor. If you have only 100 observations and you try to use 75-100 effects, your model will be bad and you might get a negative adjusted R-squared value.

Limit yourself to main effects instead of many interactions, or choose fewer explanatory variables.