A Note on the Measurement of Digital Preference in Age Recordings

Abstract

1. The field of biased and unbiased errors in recorded ages is vast, and there are many opportunities for research in it. This note is restricted to a consideration of a small part of the field. First, only errors from digital preference are considered, i.e. errors to move the age to one with a preferred end digit such as 0 or 5 or in general any even number. Secondly, concentrations at isolated ages which are a genuine feature of the population are not considered. Thirdly, the use of supplementary data (such as birth registrations) or supplementary calculations (such as graduation) is ignored.