Edward Tufte, expert in information design, and author of Beautiful Evidence, Visual Explanations, Envisioning Information, and The Visual Display of Quantitative Information describes a technique called "Small Multiples." Let's use this technique to inspect data generated with a cytokine bead array run on a flow cytometer.
Complex life-science experiments can yield challenging data sets. In one case, researchers at David Wagner’s laboratory at the Webb-Waring Center, University of Colorado School of Medicine, in Denver, Colorado used a Bender MedSystems human cytokine bead array to study differences in cytokine production in 19 donors (7 healthy controls, 12 with Type 1 Diabetes (T1D)). Four different experimental conditions—unstimulated, stimulated with anti-CD3, stimulated with anti-CD40, and stimulated with the combination of anti-CD3 and anti-CD40—were considered.
We can use the following row of scatterplots to inspect the summarized data representing unstimulated production for all 10 cytokines.
![](/images/stories/tufte/scatteruntreated75.png)
Now, let's include more data in our analysis.
![](/images/stories/tufte/scatteranti-cd375.png)
![](/images/stories/tufte/scatteranti-cd4075.png)
![](/images/stories/tufte/scatteranti-cd3andanti-cd4075.png)
Including results for all 4 stimulation environments yields the following scatterplot array, which is what Tufte would call a "small multiples" presentation.
According to Tufte, small multiples: • reveal repetition and change, pattern and surprise—the defining elements in the idea of information; • directly depict comparisons, the essence of statistical thinking; • create visual lists of objects and activities, nouns and verbs, helping viewers to analyze, compare, differentiate and decide; and • amplify, intensify, and reinforce the meaning of images.
In this case, the scatterplot array provides a concise mental model of experimental results, providing a basis for informed conversation among all members of the research team. |