Is anyone aware of published guidelines for interpreting the magnitude of effect from standardised regression coefficient (beta)? I usually rely on correlation benchmarks for interpreting coefficients as small, medium, or large but I know this is a very rough approximation.
References to authoritative guidelines would be greatly appreciated.
Hello Michael,
I think the best answer to your question is, it depends on the variables under investigation. A difference that experts might consider noteworthy in one application might be wildly different from the corresponding standard for salient differences in another application. That said, I'm not familiar with any absolute guidelines for interpreting the relative size of a standardized regression coefficient.
What you can do with standardized regression coefficient estimates is determine the relative amount of emphasis given to individual variables (of those evaluated) in that regression model, based on your data set. If multicollinearity is present, then the coefficient values may be distorted such that you might inaccurately conclude that some variables are unimportant. For this reason, examining the zero-order correlations of predictors with the DV simultaneously is recommended.
Good luck with your work!
- Badges
- Science topic
- Similar topics
- Mathematics
- Statistics
- Regression