Testing for nonresponse bias in population surveys

Population surveys are an important tool for collecting data in the social sciences, but low response rates risk compromising the representativeness of the results. Response rates have also typically declined over time, underscoring the potential for systematic bias due to nonresponse. In this study, we will conduct a field experiment with a sample size of n=15,000 to estimate nonresponse bias across a range of variables, with a focus on trust and prosociality. We will attach registry data to obtain complete data on some outcome variables.