Go to Statistics Portal


Statistics Directorate    
French Equivalent: Biais inhérent

A rather loosely defined expression which in general means, or ought to mean, a bias which is due to the nature of the situation and cannot, for example, be removed by increasing the sample size or choosing a different type of estimator.

An example of inherent bias is the systematic error of an observer or an instrument; a further example, in the interrogation of human population, is the distortion of truth by the respondent for reasons of prestige, vanity or sympathy with the investigator.

It is possible also to speak of the inherent bias of a method of estimation, although in this context the word “inherent” appears redundant. For example, in the theory of index numbers it may be shown that the standard formulae of Laspeyres and Paasche possess an inherent bias due to the methods of weighting and averaging the items.

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.

Statistical Theme: Methodological information (metadata)

Created on Sunday, May 19, 2002

Last updated on Tuesday, April 29, 2003