An alternative specification of the linear probability model

Missiakoulis, Spyros (1982) An alternative specification of the linear probability model. Statistician, 31 (4). pp. 339-343. ISSN 0039-0526. (The full text of this publication is not currently available from this repository. You may be able to access a copy if URLs are provided)

The full text of this publication is not currently available from this repository. You may be able to access a copy if URLs are provided. (Contact us about this Publication)
Official URL
http://www.jstor.org.chain.kent.ac.uk/stable/29876...

Abstract

Linear models have been proved to be inappropriate for the analysis of a dichotomous variable. There are three main problems associated with the estimation of linear probability model: heteroscedasticity, non-normal errors, and predictions outside the unit interval. Thus alternative non-linear probabilistic models have been proposed in order to incorporate these difficulties. The most commonly used are the probit and logit models.

Item Type: Article
Divisions: Faculties > Social Sciences > School of Social Policy Sociology and Social Research > Personal Social Services Research Unit
Depositing User: R. Bass
Date Deposited: 20 May 2011 14:46
Last Modified: 15 Apr 2014 13:32
Resource URI: https://kar.kent.ac.uk/id/eprint/26937 (The current URI for this page, for reference purposes)
  • Depositors only (login required):