Reviewing the theory and practice of political surveys. What some masters teached us

Abstract

In recent years published polls made mistakes when “forecasting” results of elections and referéndums. It is not only a Spanish problem. This has weakened confidence of public opinion and customers on this research technique. The logic of any applied science or technique is the accumulation of knowledge, however, this field seems to be an exception because slips are repeated in the United States, Great Britain, France or Spain. The origin of this problem is the lack of a consistent theory about voting memory and cognitive processes of respondents in their answers to surveys. The theory seems replaced by a priori and mutual monitoring between those responsible for the surveys that are published. Political polls influence public opinion. The distortions in the estimates that were published between 2014 and 2016 affected party strategies in the election campaign and the outcome of the elections. There is a risk that the published surveys may have become an ingredient of the show with which media seeks to attract or influence audiences. A final conclusion is that, in this regard, the ban on publishing polls should be maintained in the week prior to the elections.