You are not currently logged in.
Access JSTOR through your library or other institution:
The Analysis of Relationships Involving Dichotomous Dependent Variables
Paul D. Cleary and Ronald Angel
Journal of Health and Social Behavior
Vol. 25, No. 3 (Sep., 1984), pp. 334-348
Published by: American Sociological Association
Stable URL: http://www.jstor.org/stable/2136429
Page Count: 15
You can always find the topics here!Topics: Logistics, Discriminant analysis, Mathematical dependent variables, Maximum likelihood estimation, Discriminants, Coefficients, Logistic regression, Least squares, Physicians, Mathematical independent variables
Were these topics helpful?See somethings inaccurate? Let us know!
Select the topics that are inaccurate.
Preview not available
Several methods of modeling dichotomous dependent variables are reviewed from both theoretical and practical points of view. We conclude that logistic or probit models are the more theoretically appropriate, but that there are often insignificant practical differences between the results of logistic regression, discriminant analysis, and linear probability functions. Discriminant analysis and linear regression are recommended for exploratory analysis if local software limitations make them more accessible or less expensive than logistic regression programs; logistic analysis is recommended for more exact model fitting and hypothesis testing.
Journal of Health and Social Behavior © 1984 American Sociological Association