Survey Experiments in Brazilian Public Administration: Concepts, Methods, and Examples

Abstract

Survey experiments have become a central methodological tool in public administration research because they enable the identification of causal relationships in controlled settings. Despite their growing use in the international literature, their application in Brazil remains limited, particularly in studies involving civil servants. This article provides a practical guide to implementing survey experiments in Brazilian public administration, discussing their advantages, challenges, and methodological best practices. It presents three types of experiments widely used in the field: list experiments, which reduce social desirability bias when measuring sensitive attitudes and behaviors; vignette experiments, which examine how controlled variations in hypothetical scenarios influence perceptions and judgments; and fact-checking experiments, which assess belief updating following exposure to accurate information. For each experimental design, the article provides empirical examples from Brazil and discusses its underlying causal logic. It also offers a step-by-step guide to formulating hypotheses, designing the questionnaire, collecting data, and statistically analyzing the results. The article contributes to the dissemination of experimental methods in the field and to strengthening evidence-based research in Brazilian public administration.