Filter by author: Melanie Keller

Completing a questionnaire about science that comes from science likely requires at least some trust in science. Why else would one take the time to do so? Thus, the question of whom we survey is not trivial in science communication research. Therefore, this review examines how science communication researchers approach sampling in 404 quantitative survey studies published in the Journal of Science Communication, Public Understanding of Science, and Science Communication. Our findings show that they heavily rely on non-probability samples, often drawn from access panels. Sampling procedures are often vaguely described. As a consequence, we propose five steps to improve sampling in science communication research: report sampling procedures, reserve non-probability samples for experimental or exploratory work, apply probability sampling for generalizations, foster innovation, and design meaningful quotas beyond demographics.