“The Chartjunk Debate” investigates, via a research study, the actual impact of chartjunk-y versus plain graphs to assess if chartjunk entirely deserves the bad rap set forth by Tufte who coined the term. Participants were shown one style or the other and asked about their comprehension and short- and long-term recall of the data. The findings showed that chartjunk did not actually have a negative effect on viewer’s comprehension and recall, and actually improved recall in the long term. The argument made here is that embellishment can in fact be useful if employed effectively, as it serves as a mnemonic device.

However, the study has a number of shortcomings in its limited and oversimplified scope, so the findings are not catch-all by any means. The embellished charts used in the study were particularly simple, contained small datasets, and were designed by talented graphic artists. Chartjunk that is just visually ugly (as is often the case) indeed can detriment the data, but if designed well, does have the power to support the message and impact of the data. On the other hand, unembellished charts that are just plain ugly and designed poorly do not fare any better either.

In conclusion, it seems that it really is not all about the embellishments - it all comes down to how the information is represented at its core, and then embellishments on top of that may improve or detract from our understanding.