Poor education field. Why is it we always seem to be the last to know? As a career educator, I get excited hearing about new ways of thinking, knowing, or doing. Often I’m disappointed to find out that in fact, that what is new to us has been used in business or other fields for years. Such as it is with data visualization. Journalists and evaluators (among many others) have ridden the dataviz bandwagon for years now.
I’m quite happy to report, though, that we’re coming around! My June 2013 issue of American Educator, a journal on education research published by the American Federation of Teachers (AFT) opens with none other than a story on what is affectionately known as “chart junk!”* The article references a study published in the May 2013 issue of the Journal of Educational Psychology in which elementary teachers were shown two bar charts – one with colorful pictures running up each bar, and one with solid color bars – and asked which they thought would be more effective in teaching their students to read charts. All teachers chose the chart with pictures, some even claiming they would not even use the one with solid color bars. Additional experiments with students, however, showed that when they are taught to read bar charts with pictures of countable objects in the bars (that match the y-axis), they will continue to rely on that method, even when the countable objects do not match the y-axis, and even when they are later taught how to properly read a chart using bars and axes.
In other words, the presence of colorful pictures (read: chart junk) got in the way of the children’s understanding the actual mathematical relationships involved in interpreting charts – that is, the relationships between the axes and the bars.
According to the study authors, “Although such added pictorial information may be visually appealing, such information may hinder, rather than facilitate, learning and/or transfer.” This runs contrary to what many of us were taught in our teacher prep programs. Make it colorful, we were told. Make it interesting. Make it fun. After all, we want students to be engaged, right?
Not so fast, teachers. “Extraneous information included in the learning material can capture the learner’s attention and divert it from these relations during learning. Not only can extraneous information hinder learning, it can also hinder subsequent performance.” Ooops! Perhaps engagement (at least visually speaking) is not the goal after all.
“…it is important that the designers of instructional material, including textbooks and lesson plans, not simply rely on intuition as to what features may seem desirable or visually pleasing. They should recognize a priori the potential pitfalls of including such extraneous information in learning material intended for children whose ability to inhibit extraneous information is still developing.” Turns out it’s more important that children learn the mathematical principles at issue here. Turns out we teachers need to engage children in the learning without the chart junk!
The study provides pretty compelling evidence that the dataviz folks know what they’re talking about! If adults’ ability to separate the chart wheat from the chaff is hindered by chart junk, why would this not be so for children? I may never look at an elementary math textbook the same now.
While still a novice, I’m fascinated with dataviz, learning as fast as I can, and have written a bit about it here, here, here, and here.
*Edward Tufte, the great statistician and dataviz guru coined the term “chart junk” in his 1983 book The Visual Display of Quantitative Information.
References
Kaminski, Jennifer A. & Sloutsky, Vladimir M. (2013). Extraneous perceptual information interferes with children’s acquisition of mathematical knowledge. Journal of Educational Psychology, 105(2), p. 351–363. DOI: 10.1037/a0031040.
Keep it simple to avoid data distractions. (2013). American Educator, 37(2). p. 2.
I think you (and the researchers) are jumping to conclusions. The visual refinement, or whatever you’d call it, of a chart should not be random decoration, it has to carry some meaning. Like, using an oil barrel as a unit for representing xx thousand barrels of oil might facilitate comparison as well as actual values.
In 2010, a group of Canadian researchers “conducted an experiment that compared embellished charts with plain ones, and measured both interpretation accuracy and long-term recall”. They found that “people‘s accuracy in describing the embellished charts was no worse than for plain charts, and that their recall after a two-to-three-week gap was significantly better” (Useful junk?: the effects of visual embellishment on comprehension and memorability of charts / Bateman et al).
Of course, “embellishment” should only be used for simple charts and not-too-complex data. And creating bar charts with units that do not match the y-axis, well, that would be plain stupid.
I appreciate your comments. I think one of the researchers’ main points was that if children are first taught to count the pictures vs. accurately reading the y-axis scale, they have a tendency to continue to use that strategy, even when it is not appropriate (i.e when counting the pictures would result in an inaccurate interpretation of the graph). It may be “plain stupid” to create a graph such as this, and I would hope a teacher would not use such a graph in a classroom, but they do exist. Here is one example where the number of flags is not indicative of the population of G8 countries as the size of the flags differs: http://reference.wolfram.com/mathematica/example/CreatePictorialBarCharts.html
I get your point. That chart sucks!
This is awesome, Sheila! These are the very studies we need more of, not just in education. Thanks for sharing!