News users confront numbers on a daily basis. From stock prices to sports scores, government spending figures, or COVID-19 case counts, the media landscape is full of quantitative data. As part of our "Meaningful Math" research, we're uncovering ways media organizations can support better statistical and data reasoning through journalism. We've investigated how frequently numbers appear in news reporting, what can make numbers-heavy stories easier to understand, and the kinds of skills people need to understand quantitative content in news stories.

In this new paper published in Numeracy, we asked: Is the news people get related to their ability to understand and interpret numbers and statistics? In other words, is there a relationship between people's news consumption behaviors and their quantitative reasoning abilities?

Interestingly, it doesn't appear there is a strong relationship between news consumption habits and people's quantitative reasoning skills. An individual's ability to interpret numbers and statistics doesn't seem to be very strongly connected to the media outlets they turn to, or the topics and stories they pay attention to.

Methods

We asked 1,200 U.S. adults to take a 10-question multiple-choice quiz about quantitative reasoning. Each question consisted of a short text that dealt with some of the statistical concepts (natural variation, sampling, etc.) that one might encounter in a newspaper story. The scores on the quiz gave us a basic understanding of each reader's quantitative reasoning abilities. We also asked how often respondents looked at news across different topic areas and 11 different news outlets. We studied 527 stories from these 11 outlets to understand how quantitative each outlet's coverage was, using a coding scheme developed for our earlier study, "Surveying the Landscape of Numbers in U.S. News." By combining the survey data with our news sample, we were able to look for connections between people's news use and their quantitative reasoning.

Results

Our analysis of survey data showed that people's media consumption habits could be categorized into six distinct clusters of news use patterns. Our identification of these 'news repertoires' is consistent with previous studies that explored these questions using other methods (Edgerly, 2015; Dvir-Gvirsman, 2020). By grouping people into clusters based on their news use patterns, we could predict their performance on the quantitative reasoning quiz, but the relationship between news repertoire and people's scores was weak. There was no direct, straightforward relationship between high scores and topical interests.

It is important to point out that individuals' media behaviors vary along far more criteria than this study explored—including news medium, reasons for reading news, and sharing behaviors. We weren't able to measure things like how much effort respondents put into understanding stories with numbers, or the extent to which they interact with numbers in other contexts (like work or hobbies). Both of these phenomena warrant further study, and are important avenues to explore. Our data used self-reports, and didn't include detail on the particular stories any individual respondent read.

Let's Put it to Work!

For Journalists and News Outlets

• Be clear about what can and can't be inferred from quantitative data
  Be aware that the audience engaging with your work varies in how accurately they can draw conclusions from statistical information, like study or poll results. Make stories accessible by being clear about what can and can't be inferred.

Photo by Roman Kraft on Unsplash