When Words Matter: Exploring Public Responses to Different Risk Words in the SPC Convective Outlook

As one of the most visible severe weather forecasts presented on both social media and television coverage in advance of storms, the Storm Prediction Center’s (SPC) Convective Outlook is often the subject of intense debate on how to properly present convective hazard risks to users in government agencies, private businesses, and the general public. While a useful forecast tool, the Outlook was not originally designed to play such a key role in hazard communication, and studies have shown that the words used in the current scale do not consistently or intuitively communicate thunderstorm threats to message recipients. Some alternative scale labels have been proposed by researchers, ranging from a set of Words of Estimative Probability (WEPs) to alternative words associated with risk, as well as numbered levels similar to those used to scale tornadoes and hurricanes. Although these alternative scales have been tested in surveys of the public before, none have yet tested multiple labeling variations for the SPC Outlook against each other in a controlled scenario with geospatial and visual elements of the product held constant.

In this study, we asked a representative sample of the US public to report their perceived concern, likelihood of response, and perceived effectiveness of their response when shown a verbal forecast using one of five levels of risk across four versions of the Convective Outlook scale:

• The Current scale (“Marginal”, “Slight”, “Enhanced”, “Moderate”, and “High”),
• A scale based on recent Spanish translation efforts (“Minimal”, “Low”, “Moderate”, “High”, “Extreme”),
• A scale based on the Likert scale (“Very Low”, “Low”, “Medium”, “High”, “Very High”),
• A scale that uses numerical levels (“Level 1 of 5”, “Level 2 of 5”, etc).

We also prompted survey participants with a forecast map that displayed their location in a Convective Outlook forecast and asked for their concern, response likelihood, and response effectiveness once again. Participants were finally asked to align their randomly assigned risk word to a series of definitions currently used to describe the Convective Outlook scale, as well as whether those shown one of the alternative scale designs preferred it to the current scale in the outlook. Our results will add to the conversation surrounding the potential revision of the SPC convective outlook scale language and highlight the positive and negative effects on public perception that these different proposed risk communication scales have.