False Alarms and Missed Events: The Impact and Origins of Perceived Inaccuracy in Tornado Warning Systems

Recent research suggests that the real and perceived accuracy of tornado warnings impacts individual trust in forecasters and ultimately the decisions that individuals make when confronted with a warning. This paper adds to this research by asking and answering an important set of interrelated questions about public perceptions: 1) how accurate are individuals in their recollection of the number of tornado warnings that were issued for their area in the recent past? Do perceptions match reality? Or, do individuals routinely over- or underestimate the number of tornado warnings that the National Weather Service (NWS) has issued for their area in the recent past? 2) Are individual recollections of past warnings systematically biased? In other words, are some people predisposed to over- or underestimate the number of warnings that have been issued for their area in the recent past? 3) What impact do these biases have on individual trust in weather forecasters?

To answer these questions, we systematically match and then compare NWS warning data with “social” data collected in September 2012 via a census-balanced Internet survey of 4,006 respondents that reside in tornado-prone regions of the United States.