Comparing Regression Coefficients Between Same-sample Nested Models Using Logit and Probit

Sociological Methodology - Tập 42 Số 1 - Trang 286-313 - 2012
Kristian Bernt Karlson1, Anders Holm1, Richard Breen2
1Aarhus University
2Yale University

Tóm tắt

Logit and probit models are widely used in empirical sociological research. However, the common practice of comparing the coefficients of a given variable across differently specified models fitted to the same sample does not warrant the same interpretation in logits and probits as in linear regression. Unlike linear models, the change in the coefficient of the variable of interest cannot be straightforwardly attributed to the inclusion of confounding variables. The reason for this is that the variance of the underlying latent variable is not identified and will differ between models. We refer to this as the problem of rescaling. We propose a solution that allows researchers to assess the influence of confounding relative to the influence of rescaling, and we develop a test to assess the statistical significance of confounding. A further problem in making comparisons is that, in most cases, the error distribution, and not just its variance, will differ across models. Monte Carlo analyses indicate that other methods that have been proposed for dealing with the rescaling problem can lead to mistaken inferences if the error distributions are very different. In contrast, in all scenarios studied, our approach performs as least as well as, and in some cases better than, others when faced with differences in the error distributions. We present an example of our method using data from the National Education Longitudinal Study.

Từ khóa


Tài liệu tham khảo

10.1002/0471249688

10.1177/0049124199028002003

Amemiya Takeshi, 1975, Annals of Economic and Social Measurement, 4, 363

Bartus Tamás, 2005, margeff, 5, 309

Blalock Hubert M., 1979, Social Statistics, 2

10.1086/230638

10.1017/CBO9780511615412

10.1111/j.1468-0084.2007.00445.x

Curtin Thomas R., 2002, User’s Manual. National Education Longitudinal Study of 1988: Base-Year to Fourth Follow-up Data File User’s Manual (NCES 2002–323)

Goldberger Arthur S., 1991, A Course in Econometrics

Hoetker Glenn. 2004. “Confounded Coefficients: Accurately Comparing Logit and Probit Coefficients across Groups.” Working paper 03-0100. College of Business Working Papers, University of Illinois at Urbana-Champaign.

10.1002/smj.582

Kendall Patricia, 1950, Continuities in Social Research, 133

Lazarsfeld Paul L., 1955, The Language of Social Research, 115

Lazarsfeld Paul L., 1958, Daedalus, 87, 99

Long J. Scott, 1997, Regression Models for Categorical and Limited Dependent Variables

Long J. Scott, 2005, Regression Models for Categorical Dependent Variables Using Stata, 2

10.1017/CBO9780511810176

10.1080/0022250X.1975.9989847

10.1093/esr/jcp006

Powers Daniel A., 2000, Statistical Methods for Categorical Data Analysis

Simon Herbert A., 1954, Journal of the American Statistical Association, 49, 467

10.2307/270723

10.1177/0049124187016001006

10.1016/0165-1765(88)90029-8

10.2307/1912526

10.1177/0049124109335735

10.2307/2095465

Wooldridge Jeffrey M., 2002, Econometric Analysis of Cross Section and Panel Data

10.2307/1928444

Thông tin
Thông tin xuất bản

Sociological Methodology

Tập 42 Số 1

286-313

Thông tin tác giả