Truncating bar graphs persistently misleads viewers.

Abstract

Data visualizations and graphs are increasingly common in both scientific and mass media settings. While graphs are useful tools for communicating patterns in data, they also have the potential to mislead viewers. In five studies, we provide empirical evidence that y-axis truncation leads viewers to perceive illustrated differences as larger (i.e., a truncation effect). This effect persisted after viewers were taught about the effects of y-axis truncation and was robust across participants, with 83.5% of participants across all 5 studies showing a truncation effect. We also found that individual differences in graph literacy failed to predict the size of individuals' truncation effects. PhD students in both quantitative fields and the humanities were susceptible to the truncation effect, but quantitative PhD students were slightly more resistant when no warning about truncated axes was provided. We discuss the implications of these results for the underlying mechanisms and make practical recommendations for training critical consumers and creators of graphs.

The Science of Visual Data Communication: What Works

“…The prevalence of this deceptive effect has led to quantitative prescriptions for how to set y-axis boundaries to produce accurate measures of statistical effect sizes by typical viewers (Witt, 2019;B. W. Yang et al, 2021).…”

Graphs do not lead people to infer causation from correlation.

“…Things like highlighting the number of people affected (i.e., foreground) in an icon array increases risk aversion in comparison to showing icon arrays with people affected and people unaffected (i.e., foreground + background) depending on the risk probability (Okan et al, 2020). We know that altering features of graphs such as truncating the y-axis can lead to exaggerated effect sizes (Correll et al, 2020; Yang et al, 2021). And that in the judgmental forecasting literature giving participants a graph improves the forecasting of linear trends (Harvey & Bolger, 1996), but harms the forecasting of exponential trends, potentially from providing a false sense of confidence (Fansher et al, in press).…”

Summary Plots With Adjusted Error Bars: The superb Framework With an Implementation in R

“…In the third panel of Figure 1, a raincloud layout is used that shows the raw data with jittered dots and their distribution with a half-violin plot (Allen et al, 2019;Lane, 2019;Marmolejo-Ramos & Matsunaga, 2009;Rousselet et al, 2017;Weissgerber et al, 2015;Yang et al, 2021).…”