Bookmark and Share

Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: st: Comparing coefficients across sub-samples

From   Lisa Marie Yarnell <>
To   "" <>
Subject   Re: st: Comparing coefficients across sub-samples
Date   Wed, 1 Aug 2012 01:24:08 -0700 (PDT)

Here are the full references, of the two that I had mentioned, that name effect sizes for standardized regression coefficients (standardized betas):
Cohen, J. (1988). Statistical power analysis for the behavioral sciences (2nd ed.). Hillsdale, NJ: Erlbaum.
Keith, T.Z. (2006). Multiple regression and beyond. Boston: Allyn and Bacon. 
James, I don't know whether Stata can do the z-score calculation for comparing two unstandardized betas for you, but I've never had trouble doing simple calculations like this by hand.
Also, I am not sure whether all of the subscripts and superscripts came out alright for the z-score formula below, but in the denominator, the standard errors for b1 and b2 are squared.  This should also be in the Keith (2006) book, both in the text and probably on the back cover as well, where he has a lot of equations that are useful in the context of regression.
            (b1 – b2)                       b1 and b2 are the unstandardized regression weights that you want
z = --------------------                                    to test the difference between
      √(seb12 + seb22)                   seb1 and seb2are the standard errors of these unstandardized
      ↑                                                    regression weights, found next to the weights themselves

----- Original Message -----
From: Clive Nicholas <>
Sent: Tuesday, July 31, 2012 9:43 PM
Subject: Re: st: Comparing coefficients across sub-samples

Lisa Marie Yarnell replied to James Fitzgerald:


> Here are rules that I have:
> Standardized regression coefficients:
> * Keith’s (2006) rules for effects on school learning: .05 = too small to be considered meaningful, .above .05 = small but meaningful effect, .10 = moderate effect, .25 = large effect.
> * Cohen’s (1988) rules of thumb: .10 = small, .30 = medium, >  (or equal to) .50 = large

Full references please, as per Statalist FAQ. I'm certainly not about
to bust my unmentionables finding out what they are.

Clive Nicholas

[Please DO NOT mail me personally here, but at
<>. Please respond to contributions I make in
a list thread here. Thanks!]

"My colleagues in the social sciences talk a great deal about
methodology. I prefer to call it style." -- Freeman J. Dyson

*  For searches and help try:

*   For searches and help try:

© Copyright 1996–2018 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   Site index