[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: Logistic Regression
In addition, I think logistic regression has become popular due to the easy
translation of parameter estimates into relative odds ratios (a somewhat
intuitive comparison of one group to another), especially for binary
outcomes (such as abused / not abused). While there are many other ways to
look at categorical data, few have this wonderful (and sometimes misuesed)
feature.
I would pose that the best tool for an analysis depends entirely on the
question being asked. Logistic is very nice if you have cross-sectional
data and want to look at the relative odds of one group experiencing an
outcome when compared to another.
Aron
At 04:02 PM 6/27/2001 -0400, you wrote:
>“Unlike discriminant function analysis, logistic regression has no
>assumptions about the distributions of predictor variables; in
>logistic regression, the predictors do not have to be normally
>distributed, linearly related, or of equal variance within each
>group” (Tabachnick and Fidell, 1996, p. 575)
>
>
>From: Kelli K Connell <kconnell@uta.edu>
>To: Child Maltreatment Researchers
><CHILD-MALTREATMENT-RESEARCH-L@cornell.edu>
>Subject: Logistic Regression
>Date sent: Wed, 27 Jun 2001 11:14:56 -0500
>Send reply to: CHILD-MALTREATMENT-RESEARCH-L@cornell.edu
>
>> I am doing a study looking at the differences between substantiated and
>> unsusbtantiated cases of neglect. It seems that most maltreatment
>> researchers uses logistic regression for the analysis. Why is the trend to
>> use logistic instead of discriminant analysis?
>>
>> Thank you.
>>
>> Kelli
>> Kelli Connell, LMSW
>> Project Coordinator
>> Center for Child Welfare
>> UT-Arlington School of Social Work
>> Phone: 817-272-2527
>> Email: kconnell@uta.edu
>>
>
>
>
>_________________________________________________________
>Richard E. Heyman, Ph.D.
>Research Associate Professor
>Department of Psychology
>State University of New York at Stony Brook
>Stony Brook, NY 11794-2500
>(631) 632-7857
>FAX: (631) 632-7876
>Homepage http://www.psy.sunysb.edu/marital
>RMICS Coding http://www.psy.sunysb.edu/marital/coding.htm
>_________________________________________________________
Aron Shlonsky, MSW, MPH
Research Associate
UC Berkeley
School of Social Welfare
Center for Social Services Research
16 Haviland Hall, #7400
Berkeley, CA 94720-7400
(510) 642-8139
shlonsky@uclink4.berkeley.edu