Solutions to Turnout Over-Reporting: What Is Out There, What Works, and Can We Do Better?

Valid measurement of voter turnout is crucial to electoral studies. One major problem in obtaining valid turnout measurements is over-reporting, i.e. survey respondents who did not vote report having voted. Aiming to identify effective solutions to turnout over-reporting, this doctoral thesis consists of four separate but interrelated papers , plus introductory and concluding chapters. The introductory chapter reviews the causes and consequences of turnout over-reporting, providing the basis for an in-depth research into solutions. Each of the papers then addresses a question about solutions. Paper 1 critically re-examines an influential study of turnout over-reporting. The examination results highlight the need for better solutions to over-reporting. Addressing the question of "What is out there?", Paper 2 conducts a meta-analysis of studies that have experimented on innovative solutions to turnout over-reporting. Addressing the question of "What works?", Paper 3 experimentally compares two promising solutions – item-count and pipeline techniques – and finds that the former is, overall, better than the latter for preventing turnout over-reporting. Addressing the question of "Can we do better?", Paper 4 improves the design and analysis of the item-count technique, making it an even better solution to turnout over-reporting. From the results of these research papers, the concluding chapter considers the implications for developing effective solutions to turnout over-reporting, and laying the foundations for future advances in the measurement of turnout. Furthermore, the concluding chapter also discusses how the results of this doctoral research can contribute beyond election studies, towards scientific studies on a wide range of topics on which people often misreport.