Designing Passwords for Web Survey Access: The Effects of Password Length and Complexity on Survey and Panel Recruitment

Abstract Online probability panels that recruit participants via postal invitation letters use passwords to manage access to the survey. While previous research has examined primarily whether providing a password affects response rates, less attention has been given to the impact of password strength, defined by length and complexity, on response propensities. Password length refers to the number of characters in a password, while complexity refers to the set of characters (e.g., lowercase letters, digits). This study evaluates the influence of password length and complexity on various participation levels (i.e., survey access, response rates, and panel registration) as well as the propensity to consent to data linkage and item response rates for income questions. We conducted an experiment in the first wave of a German online probability survey and manipulated password length and complexity. Additionally, we included a group using the default length and complexity settings (eight uppercase letters) of the survey hosting service. The participants were randomly assigned to one of these groups. The findings indicate that longer and more complex passwords increase both participation rates and the propensity to consent to data linkage between survey and administrative data.