Problemas de cobertura en la encuesta telefónica

Abstract

This paper aims to analyse one of the main problems of surveys that are conducted over
the telephone, specifically in relation to the process of appraising and learning about those
individuals who cannot be interviewed over the telephone because they do not have one: this is
defined as coverage error. Despite the «massive» introduction of the telephone into the majority of
homes in the developed countries, not every household possesses this means of communication, and
there are groups which still lack this piece of equipment. The big problem produced by lack of
coverage is the loss of representation which occurs when a particular type of person has to be
bypassed, and so this decrease in representativeness becomes a loss of inference ability of the
sample, thereby producing a warp in the inference process. Quantification of these groups and
knowledge of their characteristics are the main aims of this paper, insofar as they allow us to
learn about the magnitude of coverage error in the telephone surveys carried out in Spain.