Access

You are not currently logged in.

Access JSTOR through your library or other institution:

login

Log in through your institution.

An Inconsistent Maximum Likelihood Estimate

Thomas S. Ferguson
Journal of the American Statistical Association
Vol. 77, No. 380 (Dec., 1982), pp. 831-834
DOI: 10.2307/2287314
Stable URL: http://www.jstor.org/stable/2287314
Page Count: 4
  • Download ($14.00)
  • Cite this Item
An Inconsistent Maximum Likelihood Estimate
Preview not available

Abstract

An example is given of a family of distributions on [-1, 1] with a continuous one-dimensional parameterization that joins the triangular distribution (when θ = 0) to the uniform (when θ = 1), for which the maximum likelihood estimates exist and converge strongly to θ = 1 as the sample size tends to infinity, whatever be the true value of the parameter. A modification that satisfies Cramer's conditions is also given.

Page Thumbnails

  • Thumbnail: Page 
831
    831
  • Thumbnail: Page 
832
    832
  • Thumbnail: Page 
833
    833
  • Thumbnail: Page 
834
    834