On methodological problems with measuring knowledge: Are women truly less financially literate than men?

Abstract When examining the role of knowledge in the decision-making process, researchers often measure it through multiple-choice questionnaires, interpreting the number of correct answers as an indicator of knowledge. In this study, we focus on the recognised methodological biases of this approach, recommending alternative ways of modelling the phenomenon of knowledge and its manifestation. We explore several ways in which the knowledge can be structured and interpreted and apply the models to the data from a group of Polish respondents surveyed on financial literacy. These models put into question some of the stylised facts on human knowledge. Particularly, we demonstrate that an alleged gender gap (lower financial knowledge of women compared to men) might be just an artefact, resulting from the oversimplified interpretation of people’s responses in the questionnaires. Finally, our models are not restricted to any specific area of knowledge and can be applied to existing micro-datasets, removing the need for re-surveying. This presents a unique opportunity to re-examine other established stylised facts.