Last week I attended the 7th edition of the Barcelona Summer School in Survey Methodology, organized by the Research and Expertise Centre for Survey Methodology (RECSM) of the Pompeu Fabra University (UPF).
During the week several programs covering diverse topics on survey methodology were proposed. I attended two modules: one on Questionnaire design, measurement quality and comparability and the other on Survey Experiments.
The first one was focused on the design of a questionnaire, including the definition of the concepts of interest. Especially focusing on the "3-step procedure" proposed by Saris and Gallhofer (2007) which starts from the concepts, assertions can be made, and then requests for an answer. Other topics covered decisions about the scale format, the use of a “don't know” answer, the grouping, and ordering of the items, the mode of data collection, etc. We also learned about Survey Quality Predictor (SQP) software, a tool useful to decide what the best formulation for a question is, i.e. the formulation which will give the highest quality data. The instructor was Melanie Revilla, a researcher at the Research and Expertise Centre for Survey Methodology (RECSM) and an adjunct professor at Universitat Pompeu Fabra (UPF, Barcelona, Spain).
The second module examined survey experimental methods for causal inference, including theory testing, treatment design, estimation of heterogeneous treatment effects, convenience sampling, and generalizability. The course also concluded with an applied session in which students design survey experiments using Qualtrics. Luckily I am already familiar with this survey platform. The instructor was D.J. Flynn is Assistant Professor of Political Science in the School of Global and Public Affairs at IE University in Madrid.