Linear versus logistic regression when the dependent variable is a dichotomy Authors
First Online: 16 February 2007 DOI:
Cite this article as: Hellevik, O. Qual Quant (2009) 43: 59. doi:10.1007/s11135-007-9077-3
The article argues against the popular belief that linear regression should not be used when the dependent variable is a dichotomy. The relevance of the statistical arguments against linear analyses, that the tests of significance are inappropriate and that one risk getting meaningless results, are disputed. Violating the homoscedasticity assumption seems to be of little practical importance, as an empirical comparison of results shows nearly identical outcomes for the two kinds of significance tests. When linear analysis of dichotomous dependent variables is seen as acceptable, there in many situations exist compelling arguments of a substantive nature for preferring this approach to logistic regression. Of special importance is the intuitive meaningfulness of the linear measures as differences in probabilities, and their applicability in causal (path) analysis, in contrast to the logistic measures.
Boyle R.P. (1966). Causal theory and statistical measures of effect: a convergence.
Am. Sociol. Rev.
Davis J.A. and Schooler S.R. (1974). Nonparametric path analysis—the multivariate structure of dichotomous data when using the odds ratio or Yule’s Q.
Soc. Sci. Res.
Fox J. (1997). Applied Regression Analysis, Linear Models and Related Methods. Sage Publications, Thousand Oaks, CA
Greene W.H. (1993). Econometric Analysis. Macmillan Publishing Company, New York
Heath A., Jowell R. and Curtice J. (1987). Trendless fluctuation: a reply to Crewe.
Heath A., Mills C., Roberts J.: Towards meritocracy? Recent evidence on an old problem. In: Crouch C., Heath A. (eds) Social Research and Social Reform: Essays in Honour of A.H. Halsey. Clarendon Press, Oxford (1992)
Hellevik O. (1983). Decomposing proportions and differences in proportions: approaches to contingency table analysis.
Qual. Quant. 40: 79–111
Hellevik O.(1984) Introduction to Causal Analysis. Exploring Survey Data by Crosstabulation. London: George Allen & Unwin; (1988. Oslo: Norwegian University Press).
Hellevik O. (1996). Fagkritikk av oppdragsforskning.
Sosiologisk tidsskrift 4: 219–228
Hellevik O. (1997). Class inequality and egalitarian reform.
Hellevik O. (2000). A less biased allocation mechanism.
Hellevik O. (2002). Inequality versus association in educational attainment research:. comment on Kivinen, Ahola and Hedman.
Hosmer D.W. and Lemeshow S. (1989). Applied Logistic Regression. John Wiley & Sons, New York
Kanagy C.L., Humphrey C.R. and Firebaugh G. (1994). Surging environmentalism: changing public opinion or changing public?
Soc. Sci. Quart. 75: 804–819
Kivinen O., Ahola S. and Hedman J. (2001). Expanding education and improving odds. Participation in higher education in Finland in the 1980s and 1990s.
Lægreid P. and Olsen J.P. (1978). Byråkrati og beslutninger: En studie av norske departement. Universitetsforlaget, Oslo
Reynolds H.T. (1977). The Analysis of Cross-Classifications. Free Press, New York
Rothman K.J. (1986). Modern Epidemiology. Little, Brown and Company, Boston
Rothman, K.J., Greenland, S. (eds.): Modern Epidemiology, 2nd edn. Lippincott-Raven Publishers, Philadelphia (1998)
Rubin D.B. (1997). Estimating causal effects from large data sets using propensity scores.
Ann. Internal Med. 127: 757–763
Skog O-J. (1998). Å forklare sosiale fenomener. En regresjonsbasert tilnærming. Ad Notam Gyldendal, Oslo
Veierød M.B., Weiderpass E., Thörn M., Hansson J., Lund E., Armstrong B. and Adami H-O. (2003). A prospective study of pigmentation, sun exposure, and risk of cutaneous malignant melanoma in women.
J. Nat. Cancer Inst. 95: 1530–1538
© Springer Science + Business Media B.V. 2007