Is “chart junk” useful? An extended examination of visual embellishment

Abstract

Although many well-cited theories or guidelines for visualization design advocate 'minimalism', designers tend to include a wide variety of visual embellishments in their charts. Researchers have examined the effects of visual embellishment on comprehension and memorability of charts under specific conditions, such as charts with a small number of data points that were viewed with no time limit (Bateman et al., 2010). This paper extends previous studies and investigates the effects of visual embellishment given different time limits for viewing these charts. Similar to the Bateman et al. (2010) study, we compared embellished charts (selected from the work of Nigel Holmes) and plain, grayscale charts, but we limited our selection to those that consisted of larger data sets (10 or more data points). Results showed that the presence of a time limit affected comprehension and short-term recall performance, while the type of chart significantly affected short-term recall. In addition, the type of chart affected the time needed to review the chart while answering the questions. Participants found Holmes charts more attractive and memorable.

Memorability of Enhanced Informational Graphics

“…The assessment of "concept grasping" by Borgo et al (2012) was similar to our assessment of theme recall, and the findings align as previously discussed. Recall of chart value details is difficult, and our study did not find differences due to While some of these studies (Bateman et al, 2010;Borkin et al, 2013;Li and Moacdieh, 2014) look at illustrative graphics and elaborate infographic embellishments more similar to Holmes (1984), others have investigated more minor manipulations and stylistic enhancements (Borgo et al, 2012;Haroz et al, 2015). Our research tends toward the side of more limited embellishments in an attempt to preserve the underlying visual encodings of data, though our custom created embellishments may provide a somewhat greater degree of enhancement while maintaining high experimental control for the presented study.…”

Map‐based Visualizations Increase Recall Accuracy of Data

“…In contrast, we test for unprompted recall of the data in the visualizations. Bateman's study has been somewhat controversial, and Li et al [LM14] recently reported a replication, limiting their selection to those charts that consisted of data sets with 10 or more observations. They found that the presence of a time limit affected comprehension and short‐term recall performance, while the type of chart significantly affected short‐term recall.…”

“…For the long term recall condition, phase three occurred four days after phase two. We chose the length of our long‐term recall condition as in previous long‐term evaluations, e.g., Li et al [LM14].…”

Crowdsourcing for Information Visualization: Promises and Pitfalls

“…Bateman et al [15] conducted a study to test the comprehension and recall of charts using an embellished version and a plain version. Bateman's study has been somewhat controversial, and Li et al [62] recently reported a replication, limiting their selection to those charts that consisted of datasets with 10 or more observations. They found that the presence of a time limit affected comprehension and short-term recall performance, while the type of chart significantly affected short-term recall.…”