Data Visualization
Leave a comment

Lies, Damned Lies, and Statistics: Exploring the Relationship Between Aesthetics and Interpretation

By Mac Hill

The Oxford Dictionaries declared “post-truth” as its 2016 word of the year. The year saw the rise of “alternative facts” and politicians declaring traditional and respected news sources “fake news” when they did not agree with a story. The truthfulness of data and data visualization has also been called into question. According to the Guardian US’s data editor, Mona Chalabi, 4 out of 10 Americans distrust the government’s economic statistics (2017). While this statistic is frightening, it makes sense. Data can be skewed and selectively used as evidence to strengthen un-truths and falsehoods. The same is true of visualizations. The visual decisions designers make can be highly subjective and change the viewer’s interpretations. Designers have to understand how aesthetic choices can skew a viewer’s perception in order to produce clear and objective visualizations.

So how can designers explore this relationship between aesthetics and interpretation? Try lying. By exploring how a visualization can lie to a viewer, designers gain a greater knowledge of their own power over a viewer. The project, aptly named “Lies, Damn Lies, and Statistics” after the Benjamin Disraeli/Mark Twain quote (“There are three kinds of lies: lies, damned lies, and statistics.”), prompted students to combine two unrelated data sets in a series of data visualizations that suggest a correlation, or even causation. These visual explorations employed both traditional statistical methods, such as line graphs, histograms, bar charts, and scatterplots, and more experimental methods like 3D modeling and animation to present the data in a skewed way. The final visualizations demonstrate challenges to visual literacy and techniques that change a viewer’s interpretation, specifically through color, material, shape, and metaphor.

Color and Material:

Linking color to data is somewhat difficult. Traditionally, sequential color schemes that move from light to dark are used for quantitative data, with low values in lighter tints and high values in darker tints (Kelleher and Wagener 2011). Adding more colors to the scheme or varying the direction of the scale can throw a reader off, pushing the acceptance of the established conclusion provided by titles.

by Bree McMahon

The same can be said of material in 3D visualizations. Using reflective materials distorts the sizes of objects and changes the viewer’s interpretations.

by Mac Hill



Viewers often have trouble accurately estimating differences in 2D areas, and when those areas are presented in radials or circles, it makes it even more difficult for the viewer to understand what’s being compared. While radials and pie charts might be visually pleasing, they are difficult for a viewer to interpret and lend themselves to false conclusions (Quach 2016).

by Clément Bordas

by Mac Hill



Traditional data visualization methods argue that scales should be aligned and similar to aid in interpretation. For this project, designers skewed and cut off scales as a means of lying to the reader (Kelleher and Wagener 2011).

by Amber Ingram

by Grace Anne Foca


Chart Type:

Not every visualization fits every data set. As readers, we’re trained to interpret different visualizations in different ways, so when a designer chooses a chart that would traditionally be inconsistent with a data set, it challenges the reader and skews the final interpretation. Clever designers can use this to their advantage, letting the chart type lie to the viewer (Quach 2016).

by Dajana Nedic


Information Overload:

Data visualizations can be a challenge for a viewer’s short-term memory, the home of sense-making. Despite our short-term memory’s capacity for high-speed information processing, it’s limited in its capacity. (Few) Visualizations that present numerous data sets in one visualization can challenge a reader’s recall, making them a challenge to read and encouraging the viewer to give up and accept the proposed conclusion rather than explore the data.

by Rachael Paine

For more details on this project please visit

Mac Hill is a Master of Graphic Design Candidate at North Carolina State University, and the Senior Editor of And So. Coming from an undergraduate degree in English Literature, she is interested how designers communicate about their work and aesthetic decisions. 


“3 ways to spot a bad statistic.” Chalabi, M. (Director). ().[Video/DVD] Retrieved from

Few, S. “Data presentation: Tapping the power of visual perception.” Retrieved from

Harris, R. L. (1999). Information graphics: A comprehensive illustrated reference. Oxford University Press.

Kelleher, C., & Wagener, T. (2011). “Ten guidelines for effective data visualization in scientific publications.” Environmental Modelling & Software, 26(6), 822-827. doi:10.1016/j.envsoft.2010.12.006

“Most of trump’s charts skew the data. and not always in his favor.” Washington Post. Retrieved from

Quach, A. (2016, -03-12T00:20:48.476Z). “Why pie charts often suck.” Retrieved from

Tavernise, S. “As fake news spreads lies, more readers shrug at the truth.” New York Times. Retrieved from

“Word of the year 2016 is… | oxford dictionaries.” Oxford Dictionaries. Retrieved from

Leave a Reply