ACM Interactions

Authors:
Nikiforos Karamanis

---

 

I read the cover story of the July—August 2018 issue of Interactions on data visualizations by Danielle Albers Szafir with great interest, particularly since I recently gave an introductory webinar on this topic for the European Bioinformatics Institute (EMBL-EBI) (see [1:[R1]).

The article focuses on "understanding what people see when they look at a visualization" to design visualizations "that support more accurate data analysis and avoid unnecessary biases." This is valuable, particularly within the context of a how-to article that needs to be brief and practically applicable.

However, I think the cover story would have been even stronger if it:

• cited some additional seminal background work in relation to what people see; and
• mentioned, even in passing, the importance of studying what people do, which has been fairly firmly established within the UX and HCI communities.

In particular, I think it would have been useful to mention the ranking of visual channels ([2,3] and [1:R2]) especially given that position is considered an even better way to encode quantitative data than sequential or divergent color maps. In the recent review by O'Donoghue et al. [1:R3], Figure 2 provides an excellent visual overview of the ranking combined with succinct practical advice on the use of color maps.

Additionally, given that the article is about graphical integrity, I was expecting it to refer to Edward Tufte [4], to whom this principle is attributed. Mentioning Tamara Munzner's textbook [5] would have been useful too for those who are new to the field but want to study interactive visualizations in more depth.

My own audience is early-career life scientists, so I based my webinar on the Points of View columns on data visualization in Nature Methods [1:R4], which is a familiar and inspirational journal for them. Given that some of the examples in the cover story came from biology, citing this resource may also have been helpful for that readership.

In my role as a UX practitioner, I rely on particular methods such as interviews and contextual observations to understand the needs of life scientists, to capture these needs (typically as user personas and task models), and to formulate the question that we are trying to answer with a visualization (e.g., as a problem statement or a job to be done). In other words, I focus as much on what people do as on what people see by applying methods that are fairly established within the HCI and UX communities.

When I tell the story of how we designed a visualization or a whole Web application for a particular service in EMBL-EBI [1:R5], these methods stand center stage. Although the importance of qualitative fieldwork and analysis has been highlighted, for example, in the design study methodology framework of Sedlmair et al. [6], my impression is that these popular UX methods are still not routinely embedded in the everyday process of data visualization researchers and practitioners. The emphasis on what people see in the cover story reinforced this impression.

At EMBL-EBI we bring together experts from industry and academia to address current challenges in data visualization faced by our industry partners [1:R6] in an attempt to bridge the gap between data visualization researchers, the HCI community, UX practitioners, and domain experts (especially from the pharmaceutical and agro-food industry). We are also contributing to the UX for Life Sciences (UXLS) initiative (https://uxls.org/) to enable organizations that develop scientific software to adopt UX principles and methods.

I hope this blog post will help all of us who are part of these diverse and active communities focus on what people do in addition to what people see.

In closing, I'd like to thank Danielle Albers Szafir for writing the cover story and the editors of Interactions for publishing it. I welcome your feedback on these thoughts.

  References

1. Additional references for this article (R1, R2, etc.) available at http://bit.ly/karamanis-blogIX

2. Cleveland W.S. and McGill, R. Graphical perception and graphical methods for analyzing scientific data. Science 229, 4716 (Aug. 1985), 828–833.

3. Mackinlay, J. Automating the design of graphical presentations of relational information. ACM Trans. Graph. 5, 2 (Apr. 1986), 110–141.

4. Tufte, E. The Visual Display of Quantitative Information. Graphics Press. Cheshire, CT, 2001.

5. Munzner, T. Visualization Analysis and Design. CRC Press, 2014.

6. Sedlmair, M., Meyer, M., and Munzner, T. Design study methodology: Reflections from the trenches and the stacks. IEEE Transactions on Visualization and Computer Graphics 18, 12 (Dec. 2012), 2431–2440.

  Author

Nikiforos Karamanis is a senior user experience designer at EMBL-EBI. He enjoys spending time with life scientists, developers, and other stakeholders to help them work together to achieve their goals using lean user experience methods. [email protected]

 

Copyright held by author

The Digital Library is published by the Association for Computing Machinery. Copyright © 2019 ACM, Inc.