Development and Application of Statistical Methods in the International Tobacco Control Four-Country Survey

• Activity Number: 271
• Type: Roundtables
• Date/Time: Tuesday, August 6, 2013 : 7:00 AM to 8:15 AM
• Sponsor: Section on Statistics in Epidemiology
• Author(s): Mary E Thompson*+
• Companies: University of Waterloo, Canada
• Keywords: survey design ; prospective study ; cross-country survey ; conceptual models ; tobacco control ; policy evaluation

Abstract

The International Tobacco Control (ITC) Four Country Survey, led by Geoffrey Fong of Psychology, University of Waterloo, is a prospective study of more than 8,000 longitudinal respondents, with nearly annual waves that began in 2002. Developed to evaluate tobacco control measures mandated by the WHO Framework Convention on Tobacco Control, the survey is based on theory-driven conceptual models and has longitudinal and quasi-experimental designs as key statistical features. The replenishment of the sample with a new cohort at each wave allows the study of time-in-sample effects in models of change. The project requires innovative statistical methods in the analysis of complex survey data to account not only for sampling errors, but also for attrition, measurement errors, and a transition to mixed mode collection. Important policy implications, such as the importance of attention to the design of graphic health warning labels, have emerged from this study. The roundtable discussion will provide an opportunity to talk with Mary Thompson, survey statistician of the ITC Project, about the challenges and opportunities for statisticians in the field of health policy evaluation surveys.