You are not currently logged in.
Access your personal account or get JSTOR access through your library or other institution:
Logit-Based Interval Estimation for Binomial Data Using the Jeffreys Prior
Donald B. Rubin and Nathaniel Schenker
Vol. 17 (1987), pp. 131-144
Published by: American Sociological Association
Stable URL: http://www.jstor.org/stable/271031
Page Count: 14
Preview not available
Interval estimates for a binomial probability based on normal approximations on the logit scale are evaluated in terms of confidence and Bayesian coverage properties. These estimates, derived as Bayesian posterior intervals under the Jeffreys prior distribution, correspond to large-sample maximum likelihood procedures with a half success and a half failure appended to the data. The intervals work well and are suggestive of prior distributions for more general logistic regression problems.
Sociological Methodology © 1987 American Sociological Association