The SWE colloquium on Thursday 8 October (12:00-13:00) will be presented by Björn Meder, University of Potsdam. The talk will be totally virtual, streamed live through Zoom.

Understanding of and reasoning with verbal uncertainty terms

Dealing with different degrees of frequency and probability is important not only in science, but also in our everyday lives. Relevant information, however, does not always come in the form of numerical estimates or direct experiences, but is instead obtained through qualitative, rather vague verbal terms (e.g., “the virus often causes coughing” or “it’s likely to rain”). Understanding how people make sense of such verbal expressions and how they represent and utilize this kind of information is therefore critical to understand cognition and behavior in many real-world situations. I will focus on three key issues. First, when does a shared understanding of verbal uncertainty terms emerge in development? Second, how can we formally represent the vagueness of verbal uncertainty terms and build computational models of reasoning with such information? Third, how good are people when making probabilistic inferences based on verbal uncertainty terms, compared to reasoning with numerical information and relative to normative benchmarks? I conclude by discussing ideas and pathways for investigating judgment and decision making with verbal information within a computational modeling framework.