Prof. Alex Theodoridis’ paper "Worth Weighting? How to Think About and Use Weights in Survey Experiments" has just been published in Political Analysis. Congratulations to Prof Theodoridis and his co-authors!
The paper studies the use of sampling weights to adjust online survey experiments or representativeness, which is increasingly common in Political Science research. The authors formally develop a number of tools for analysing survey experiments, and use both simulations and real-world survey data to examine these tools. Surprisingly, the authors find that the use of weights is not particularly important for retrieving population-representative estimates of experimental effects, at least when the quality of the survey is reasonably high.
