COVID-19 is a new disease, so people only have partial knowledge about the associated health risks and may not be able to assign a well-defined probability to certain events, such as the probability they will get infected or transmit the virus to others. To deal with this, MiSoC Co-I Delavande and Director Del Bono use a novel survey design to elicit imprecise probabilities about COVID-19 related risks by allowing respondents to report a minimum and maximum subjective probability associated to different health events. As expected, the data reveal considerable imprecision about various health risks. For example, the mean minimum for the chance of contracting COVID-19 in the next 3 months (conditional on current behaviours) is 15%, while the mean maximum is 37%. Importantly, individuals who express more imprecise beliefs, or greater ambiguity, are found to be less likely to adhere to social distancing and hygiene practices. More details can be found here.