Measuring scientific interest: the effect of knowledge questions on interest ratings
Research into the publics' interest in, knowledge of and attitudes towards science has captured the attention of politicians and educators. Sample survey research has been employed to assess the diverse aspects of the public understanding of science. However, surveys are subject to various biases that may affect the findings, calling into question both the reliability and the validity of the measures concerned. In this study we look at one common bias—that of context effects. Context effects occur when a question influences responses to later questions. The effects of answering one of four different sets of science questions (physical or life science, and easy or difficult questions) on what people report as their interest in science and what they think science is, were investigated using a split ballot format ( n = 2099). Two approaches from social psychology, framing and consistency, are used to predict the effects of these knowledge questions on subsequent responses. Context effects were found and were more in line with the framing explanation. The results signal the need for caution in interpreting findings from surveys of the public understanding of science.