The results of the study on weather forecasts undertaken by undergraduate researchers, Matthew Standage, Rachel Bartlett and Shymali Abraham have been published in the latest issue of Meteorological Applications. The research explores how members of the public understand and interpret weather forecasts (rain probability and intensity) for their decision-making.
This project was a collaboration between the Reading’s Departments of Meteorology and Psycholology and the Centre for Information Design Research, funded by the University’s Undergraduate Research Opportunities Programme.
In their survey of 274 members of the public, the research team asked about the sources people used to get weather forecasts, and why they preferred the sources they used. They also asked the participants to make decisions based on forecasts for rainfall that included different probabilities (e.g. 30% chance of rain) and different intensities (light, moderate, heavy rain). The forecasts were presented both in graphic and tabular formats.
Among its findings, the study showed that although people have some understanding of probabilistic forecasts, they can find them difficult to interpret accurately. The results suggest that age and education affect the understanding of forecasts, and it might be that exposure to the ideas behind probability and uncertainty (during school education, for example) eases the understanding of probabilistic forecasts.
The study found that traditional channels for communicating forecasts, such as newspapers, radio and television, were preferred by participants over 40 years old, whereas younger participants preferred narrow-cast channels such as websites and mobile phone applications. These provide information that is more specific and immediately updated. However the younger group, still used television forecasts so it seems website and mobile phone applications add to the amount of weather information they receive.
The study takes a look at the challenges of explaining probabilistic information, the limitations of visualisations of uncertainty. It appears that although people might not fully understand the information in probabilistic forecasts, they are able to extract key information from them, for example to avoid planning a barbecue at a time when there is a high probability of rain.
The paper is available here.