Back to Programme

"I Was Never Good at Science": Determinants of Non-substantive Answers in Science Surveys

Ana Muñoz van den Eynde (CIEMAT)
Josep Lobera (Universidad Autónoma de Madrid)

Keywords: Survey research and questionnaire design

Abstract

Public opinion polls are grounded in two important assumptions. Firstly, it is supposed that every person have an opinion about everything and is ready and willing to provide it in the moment he or she is asked about. But there is growing evidence pointing out to the fact that people need to know about the issue in order to give an opinion, and –more importantly- they need to feel that they know about it. Secondly, it is supposed that human beings are rational and invest a lot of cognitive effort in giving an elaborated answer to every question they are asked about. However, empirical research points in another direction: people do not invest such a cognitive effort when answering to opinion polls. It is also known that this occurs more frequently among people less knowledgeable about the issue they are asked about. The consequence of both assumptions has a lot to do with the “nonattitudes” problem identified by Philip Converse. Nevertheless, this issue has been little studied –and almost neglected– in the field of the public understanding of science. Science as a formal object of attitudes and opinions is extraordinarily complex and, as a consequence, it is reasonable to anticipate the presence of significant cognitive and evaluative barriers to the development and/or the manifestation of attitudes about science.

In this contribution we test the hypothesis that the combination of the complexity of the scientific issue and some elements of the questionnaire design is a decisive factor to expel from the study those persons that are less knowledgeable and/or that feel less knowledgeable about science. To identify the factors that increase the vulnerability of people to be excluded from the survey, we analyze non-substantive responses to Public Attitudes to Science 2014 survey (PAS 2014), a study looking at attitudes to science, scientists and science policy among the UK public. We have selected this data set because PAS 2014 includes more than 50 statements, with different degrees of abstraction, measuring attitude to science with reasonable internal consistency. This allows us to engage with available literature showing that scientific knowledge develops an important role in non-substantive answers in this kind of surveys.
In order to improve the representativeness of surveys of public perception of science, efforts should be made to include citizens who experiment more difficulties having or expressing an opinion about science and, as a consequence, are prone to exclude themselves from science and technology surveys. The results of our analysis will help to improve questionnaire design in science surveys and, therefore, will enhance our understanding of the factors contributing to the public attitudes to science, scientists, and science policy.