I am wondering if there is a table or some reference that explains the strength of the beta coefficient.
$\begingroup$
$\endgroup$
3
-
$\begingroup$ possible duplicate of Meaning of p-values in regression $\endgroup$– AndyCommented May 4, 2014 at 12:54
-
$\begingroup$ That question was about p-values, this one is about beta. I think that makes it different. $\endgroup$– Peter FlomCommented May 4, 2014 at 14:15
-
1$\begingroup$ Define what you mean by the "beta coefficient". $\endgroup$– Scortchi ♦Commented May 4, 2014 at 16:07
Add a comment
|
2 Answers
$\begingroup$
$\endgroup$
It depends of the phenomenon that you are studying. If you can tell me more about your regression, I could try to interpret it.
By the way, you should do a t-test, to check if your beta is statistically different from zero. You could check this link to understand what I'm talking about: http://en.wikipedia.org/wiki/Student%27s_t-test.
$\begingroup$
$\endgroup$
2
Any book on regression will explain the $\beta$ coefficient.
$\beta$ refers to the standardized coefficient. It is a measure of how much the dependent variable changes when the independent variable changes by 1 standard deviation. (But see comment below about econometrics).
Is it strong? That is context-dependent. If you tell us the context, someone here may be able to answer.
answered May 4, 2014 at 14:18
-
1$\begingroup$ In econometrics $\beta$ refers to a coefficient, standardized or not. $\endgroup$– AksakalCommented May 4, 2014 at 14:20
-
$\begingroup$ OK, good to know. In social sciences they use $\beta$ for standardized and B for not standardized. $\endgroup$ Commented May 4, 2014 at 14:23