A new study suggests that the format in which charts are presented could mislead people into being overly optimistic or pessimistic about the trends the charts show.
Academics from the City, University of London and University College London found that people making predictions about how a trend would develop over time make poorer judgments when the trend is plotted as a bar chart than when it is the exact same data is represented as a line chart or as a chart consisting only of a series of data points.
The study included four online experiments with a total of over four thousand participants. In the first two experiments, participants were each given a single chart, either a bar chart, line chart, or scatter chart, filled with 50 data points representing a fictitious company’s weekly sales. Respondents had to click on the graph to see how many sales they thought the company would make over the next eight weeks. They were encouraged to provide accurate answers.
In the first experiment, the number of sales on the chart increased week-to-week, and participants generally predicted that sales would continue to increase; In the second experiment, the trends in the chart declined, making participants more pessimistic about future sales.
Still, many different types of trend participants thought that sales would be lower if the data was presented as bar charts rather than line charts or scatter charts.
The researchers wondered if the reason for this was that in bar charts, the area inside the bar is usually heavily shaded and therefore draws attention visually, lowering participants’ estimates compared to the other types of charts where it no shading there to attract the eye’s attention. However, in a third experiment, they found the same lower predictions for bars even when the bars remained unshaded.
In a fourth experiment, they tested a version of a bar chart in which the bars emanated from the top rather than the bottom of the chart. While subtle trends in the data suggest this might reverse the bias, the results were inconclusive.
Stian Reimers, Professor of Psychology and Behavioral Sciences at the School of Health & Psychological Sciences, City, University of London, who led the research, said: “In recent years it seems like we’ve spent a lot of time looking at time Series: whether it’s the number of COVID cases, electricity prices or inflation rates to find out what’s next. Our research shows that our predictions of what’s going to happen next aren’t just influenced by the trends we are in. This obviously has implications for all of us as we try to make decisions about whether it’s likely to be safe, to visit vulnerable relatives, or whether we can afford to visit them on a mortgage.”
These biases can affect not only the decisions of individuals, but also the many companies that perform analytics such as “demand forecasting,” which uses historical data to estimate future customer demand for a product or service and to predict; especially when these judgments are made without the help of people who are “eyeing” charts directly and assessing how they think a trend will develop.
However, Professor Reimers believes these biases may have merits: ‘It’s potentially useful because this type of format effect could help counteract some of the other mistakes people make when projecting trends into the future. Many of the other biases people display when they try to extrapolate trends are baked into the way we see the world and are difficult to change. The format we use for our charts is something we have complete control over, so it may be possible to use certain formats to undo people’s built-in biases and help people make more accurate judgments.
“Although we had many participants, this is only a small group of studies. It will be interesting to see how well these results generalize across different formats and expertise, and exciting to try to find the evolving ways of presenting data over time in ways that help people best grasp of the state of the world and most accurate in predicting what is likely to happen next.”
The research is published in Open Access International Journal of Forecasts.
Stian Reimers et al., Bars, Lines, and Dots: The Impact of Chart Format on Evaluative Forecasts, International Journal of Forecasts (2023). DOI: 10.1016/j.ijforecast.2022.11.003
Provided by City University London
Citation: New study suggests that when predicting trends, reading a bar chart versus a line chart skews our judgment (2023 January 26), retrieved January 26, 2023 from https://phys.org/news/2023-01 -trends-bar-line-graph-biases.html
This document is protected by copyright. Except for fair trade for the purpose of private study or research, no part may be reproduced without written permission. The content is for informational purposes only.