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

From |
John Antonakis <john.antonakis@unil.ch> |

To |
statalist@hsphsun2.harvard.edu |

Subject |
Re: AW: st: AW: beta coefficients for interaction terms |

Date |
Sun, 21 Jun 2009 14:55:00 +0200 |

Hi: It depends on what you want. Here is the original regression: . reg mpg head length ia2,

------------------------------------------------------------------------------

-------------+----------------------------------------------------------------

------------------------------------------------------------------------------ From the -spost- suite of commands, we get: . listcoef, std regress (N=74): Unstandardized and Standardized Estimates Observed SD: 5.7855032 SD of Error: 3.5456553 -------------------------------------------------------------------------------

-------------+-----------------------------------------------------------------

-------------------------------------------------------------------------------

When you run . reg mpg head length ia2, beta

------------------------------------------------------------------------------

-------------+----------------------------------------------------------------

------------------------------------------------------------------------------

. reg smpg shead slength sia2

------------------------------------------------------------------------------

-------------+----------------------------------------------------------------

------------------------------------------------------------------------------

Best, J. ____________________________________________________ Prof. John Antonakis Associate Dean Faculty of Business and Economics University of Lausanne Internef #618 CH-1015 Lausanne-Dorigny Switzerland Tel ++41 (0)21 692-3438 Fax ++41 (0)21 692-3305 Faculty page: http://www.hec.unil.ch/people/jantonakis&cl=en Personal page: http://www.hec.unil.ch/jantonakis ____________________________________________________ On 21.06.2009 13:59, lschoele@rumms.uni-mannheim.de wrote:

Thank you very much.I was just a little confused, since in the example I was given the DVwas standardized.Example you should code . sysuse auto, clear . egen shead = std(headroom) . egen slength = std(length) . egen smpg = std(mpg) . gen ia2 = shead*slength . egen sia2 = std(ia2) . reg smpg shead slength sia2 Best, Lisa Zitat von John Antonakis <john.antonakis@unil.ch>:Hi: As far as I know, there is not reason to standardize the dependent variable (when standardizing the independent variable to obtain their standardized effects). Given the range of the DV will change substantively when standardizing it is evident that the intercept now is qualitatively different from before and might not be different from zero. I'd stick to having the DV in raw units. HTH, J. ____________________________________________________ Prof. John Antonakis Associate Dean Faculty of Business and Economics University of Lausanne Internef #618 CH-1015 Lausanne-Dorigny Switzerland Tel ++41 (0)21 692-3438 Fax ++41 (0)21 692-3305 Faculty page: http://www.hec.unil.ch/people/jantonakis&cl=en Personal page: http://www.hec.unil.ch/jantonakis ____________________________________________________ On 21.06.2009 12:54, lschoele@rumms.uni-mannheim.de wrote:Hibut it is important for my paper to show the value of theintercept, and I get different values if I use either thestandardized or the not standardized dependent variable. If I havethe null hypothesis that is is equal to zero, do I have to use thestandardized or not standardized dependent variable? And comingback to my original question: Why do I have to standardize thedependent variable?Best, Lisa Zitat von Alan Neustadtl <alan.neustadtl@gmail.com>:If it is not meaningful for your paper how is that complete. I am sure yo ucould think of other things to include for "completeness" but you left out. In general the null hypothesis for the intercept is that it is equal to zero. If that is unimportant it is not needed. Best, AlanOn Sat, Jun 20, 2009 at 10:26 AM, <lschoele@rumms.uni-mannheim.de>wrote:Hi Alan, I do not have a hypothesis for the intercept, but I want to show the significance in my paper due to completeness. Best, Lisa Zitat von Alan Neustadtl <alan.neustadtl@gmail.com>:In your case what is your null hypothesis regarding the intercept? Why do you want to test for significant difference of this estimate and some other value? In many cases the intercept is 1) beyond the range of the data (and therefore a poor estimate), and 2) theoretically uninteresting. Best, AlanOn Sat, Jun 20, 2009 at 8:51 AM, <lschoele@rumms.uni-mannheim.de>wrote:Hi Statalist, I have one more question regarding this theme. Why do I have to standardize the dependent variable as well? If I standardize it, the constant won't be significant anymore. Without standardizing the constant is highly significant in my case. Thank you Lisa Zitat von Ulrich Kohler <kohler@wzb.eu>:I think Lisa refer the section on standardized regressioncofficients ofthat book, particularly to the second item on pg. 201 (Englishedition).That item states that one should not use b*s_x/s_y to createstandardized regression coefficients in the presence ofinteractionterms. Based on Aiken/West 1991 (28-48) it is recommended that oneshould standardize all variables that are part of theinteraction inadvance. Hence, instead of coding . sysuse auto, clear . gen ia = head*length . reg mpg head length ia, beta you should code . sysuse auto, clear . egen shead = std(headroom) . egen slength = std(length) . egen smpg = std(mpg) . gen ia2 = shead*slength . egen sia2 = std(ia2) . reg smpg shead slength sia2The estimated coefficients of the constituent effects thenshow how muchstandard deviations the dependend variable change when theindependentvariable changes by one standard deviation and the othervariable of theinteraction term is at its mean.Standardized regression coefficients are often used to findout which ofthe independent variables have the "largest" effect. I mustadmit that Ioften fail to understand why students want to know that. Butleavingthat aside, if an effect is not constant over the range of anothervariable (i.e. in the presence of an interaction term) thequestion ofwhich independent variable have the largest effect seemspointless.Many regards Uli Am Donnerstag, den 16.04.2009, 14:47 +0200 schrieb Martin Weiss:<>Your -gen- statement computes the interaction, but Stata wouldtreatthisnew variable as a covariate in its own right, w/o anyconnection toothercovariates. A similar issue arises with quadratic terms of acovariate(http://www.stata.com/statalist/archive/2008-08/msg00307.html).The book you mentioned has a subsection on the topic on pages222-226,andthe English version seems to be a straightforward translationof it,AFAIK(http://www.stata-press.com/books/daus2.html, page 222). Icannot findthe stuff on the beta coefficient there, though. They do say that you shouldcheck for missings with -rowmiss- and that you shouldsubtract the meanfromthe variables before standardization. The latter is easilyaccomplishedvia ***** sysuse auto, clear *enter your vars to be standardized here local stdvars "price weight trunk turn" foreach var of local stdvars{ summ `var', mean gen std`var'=`var'-r(mean) } *****-egen, std()- would divide by the standard deviation inaddition to mycode... HTH Martin -----Ursprüngliche Nachricht----- Von: owner-statalist@hsphsun2.harvard.edu [mailto:owner-statalist@hsphsun2.harvard.edu] Im Auftrag von lschoele@rumms.uni-mannheim.de Gesendet: Donnerstag, 16. April 2009 14:07 An: statalist@hsphsun2.harvard.edu Betreff: Re: st: AW: beta coefficients for interaction terms How do I tell stata, that it is an inetraction term? Here is what I did: gen appearance_attention=apperance*attentionIs that telling stata, that the new variable is an interactionterm?I am referring to the book "Datenanalyse mit Stata" by Kohler,Kreuter"Note that you can effect the standardization yourself via -egen,std()-" What standardization do you mean? Thez-standardization or the"normal" standardization for the beta coefficients, that Ineed forthe interpretation? Best Lisa Zitat von Martin Weiss <martin.weiss1@gmx.de>:<>Well, did you tell Stata in any way that a specific variableis an"interaction term"? If not, Stata probably treats it as justanothercovariate in your regression. BTW, which book are you referring to?Note that you can effect the standardization yourself via -egen,std()- HTH Martin -----Ursprüngliche Nachricht----- Von: owner-statalist@hsphsun2.harvard.edu [mailto:owner-statalist@hsphsun2.harvard.edu] Im Auftrag von lschoele@rumms.uni-mannheim.de Gesendet: Donnerstag, 16. April 2009 12:35 An: statalist@hsphsun2.harvard.edu Betreff: st: beta coefficients for interaction terms Hi Statalist,I am working on a regression model with interactions betweensomevariables. I read in a book, that I can't use the "normal"standardized beta coefficients for the interaction terms.They saidthat the interpretation of the beta coefficients is not possible until you z-standardise the interaction variables before you do the regression. Does anyone know, if stata does the z-standardization for the interaction variables automatically, so I can use the normalstandardized beta coefficients (shown in the stata output)for theinterpretation? I am using the 9.1 version of stata. I hope someone can help me. Best Lisa * * For searches and help try: * http://www.stata.com/help.cgi?search * http://www.stata.com/support/statalist/faq * http://www.ats.ucla.edu/stat/stata/ * * For searches and help try: * http://www.stata.com/help.cgi?search * http://www.stata.com/support/statalist/faq * http://www.ats.ucla.edu/stat/stata/* * For searches and help try: * http://www.stata.com/help.cgi?search * http://www.stata.com/support/statalist/faq * http://www.ats.ucla.edu/stat/stata/ * * For searches and help try: * http://www.stata.com/help.cgi?search * http://www.stata.com/support/statalist/faq * http://www.ats.ucla.edu/stat/stata/-- kohler@wzb.eu 030 25491-361 * * For searches and help try: * http://www.stata.com/help.cgi?search * http://www.stata.com/support/statalist/faq * http://www.ats.ucla.edu/stat/stata/* * For searches and help try: * http://www.stata.com/help.cgi?search * http://www.stata.com/support/statalist/faq * http://www.ats.ucla.edu/stat/stata/* * For searches and help try: * http://www.stata.com/help.cgi?search * http://www.stata.com/support/statalist/faq * http://www.ats.ucla.edu/stat/stata/* * For searches and help try: * http://www.stata.com/help.cgi?search * http://www.stata.com/support/statalist/faq * http://www.ats.ucla.edu/stat/stata/

**References**:**Re: AW: st: AW: beta coefficients for interaction terms***From:*lschoele@rumms.uni-mannheim.de

**Re: AW: st: AW: beta coefficients for interaction terms***From:*Alan Neustadtl <alan.neustadtl@gmail.com>

**Re: AW: st: AW: beta coefficients for interaction terms***From:*lschoele@rumms.uni-mannheim.de

**Re: AW: st: AW: beta coefficients for interaction terms***From:*Alan Neustadtl <alan.neustadtl@gmail.com>

**Re: AW: st: AW: beta coefficients for interaction terms***From:*lschoele@rumms.uni-mannheim.de

**Re: AW: st: AW: beta coefficients for interaction terms***From:*John Antonakis <john.antonakis@unil.ch>

**Re: AW: st: AW: beta coefficients for interaction terms***From:*lschoele@rumms.uni-mannheim.de

- Prev by Date:
**Re: AW: st: AW: beta coefficients for interaction terms** - Next by Date:
**st: Reshape and Changing Data** - Previous by thread:
**Re: AW: st: AW: beta coefficients for interaction terms** - Next by thread:
**Re: AW: st: AW: beta coefficients for interaction terms** - Index(es):

© Copyright 1996–2016 StataCorp LP | Terms of use | Privacy | Contact us | What's new | Site index |