Go to Statistics Portal

INTERVAL ESTIMATION

Statistics Directorate    
French Equivalent: Estimation par intervalle

Definition:
The estimation of a population parameter by specifying a range of values bounded by an upper and a lower limit, within which the true value is asserted to lie, as distinct from point estimation which assigns a single value to the true value of the parameter. The unknown value of the population parameter is presumed to lie within the specified interval either on a stated proportion of occasions, under conditions of repeated sampling, or in some other probabilistic sense.

Context:
The first of these two approaches is that of confidence intervals due to Neyman (1937), which regards the value of the population parameter as fixed and the limits to the intervals as random variables. A second approach is that of fiducial limits due to R.A. Fisher (1930) where the population parameter is regarded as having a “fiducial probability” distribution which determines the limits.

Source Publication:
A Dictionary of Statistical Terms, 5th edition, prepared for the International Statistical Institute by F.H.C. Marriott. Published for the International Statistical Institute by Longman Scientific and Technical.

Cross References:
Estimation

Statistical Theme: Methodological information (metadata)

Created on Tuesday, May 21, 2002

Last updated on Wednesday, January 04, 2006