Annals of Statistics

Defining the Curvature of a Statistical Problem (with Applications to Second Order Efficiency)

Bradley Efron

Full-text: Open access


Statisticians know that one-parameter exponential families have very nice properties for estimation, testing, and other inference problems. Fundamentally this is because they can be considered to be "straight lines" through the space of all possible probability distributions on the sample space. We consider arbitrary one-parameter families $\mathscr{F}$ and try to quantify how nearly "exponential" they are. A quantity called "the statistical curvature of $\mathscr{F}$" is introduced. Statistical curvature is identically zero for exponential families, positive for nonexponential families. Our purpose is to show that families with small curvature enjoy the good properties of exponential families. Large curvature indicates a breakdown of these properties. Statistical curvature turns out to be closely related to Fisher and Rao's theory of second order efficiency.

Article information

Ann. Statist., Volume 3, Number 6 (1975), 1189-1242.

First available in Project Euclid: 12 April 2007

Permanent link to this document

Digital Object Identifier

Mathematical Reviews number (MathSciNet)

Zentralblatt MATH identifier


Primary: 62B10: Information-theoretic topics [See also 94A17]
Secondary: 62F20

Curvature exponential families Cramer-Rao lower bound locally most powerful tests Fisher information second order efficiency deficiency maximum likelihood estimation


Efron, Bradley. Defining the Curvature of a Statistical Problem (with Applications to Second Order Efficiency). Ann. Statist. 3 (1975), no. 6, 1189--1242. doi:10.1214/aos/1176343282.

Export citation