.png)
A pedagogical approach to teaching rigor in data visualization faces a challenge in that teaching the precise 'rules' of displaying data visually is highly dependent on one's research area and intended narrative. Rules fail as soon as there is a disconnect between what was didactically learned, and the subject of a study or a data type. Traditional instruction often begins with technical guidelines—use this chart type for these kinds of data, avoid pie charts, minimize visual clutter—treating visualization as the endpoint rather than part of a continuous process of scientific inquiry. But therein lies the rub: rigor in data visualization begins long before choosing a figure type or a display choice.
The decision to display data and how to use that data to develop your scientific story is the very first rigor-related choice that you make. Which experiments to showcase, which variables to highlight, which comparisons to draw—these foundational decisions shape the narrative before a single axis is plotted. Therefore, teaching rigor in data visualization broadly cannot be about prescribing specific plots for every data analysis case, but rather about raising awareness of where rigor needs to be attended to, and how choices about what to display hold tremendous power for clarity, transparency, and impact in scientific discovery.
To address this pedagogical challenge of developing rigor awareness, rather than merely teaching rules, we developed a curriculum that emphasizes critical decision points across the entire data visualization process, while providing practical skills development opportunities. In the C4R unit titled Effective Data Visualization for Research Communication, our team designed eight lessons aimed at graduate learners to increase awareness about how rigor comes into play in data visualizations, to help learners understand the roles of data visualization in story reporting and scientific discovery, to equip learners to identify key considerations when designing and analyzing data visualization, and to empower them to feel confident applying rigorous practices in data visualization creation.
The critical theme that weaves the lessons together is the concept that there is not a one-size-fits-all solution when researchers are working with data visualizations, but rather that there must be an awareness of how rigor comes into play. The first lesson begins with defining roles of data visualizations in research, emphasizing the importance of upholding rigorous practices to make the plots, graphs and figures reliable sources of information. Learners engage in trials of plot interpretation, emphasizing how dynamic elements of each figure emphasize rigor principles themselves. Based on descriptive scenario activities, Lesson 2 asks learners to derive critical decision points in the visualization process, including pre-visualization decisions (which experiments to include), data preparation decisions (transformations, exclusions, aggregations), representation decisions (figure type, scales, and color choices), contextual decisions (comparisons shown, limitations), and sets the stage for critical thinking about approaches in the following lessons. Lessons 3 and 4 ask learners to focus on features of plot types and how they aid or hinder interpretation of data, while providing foundational information about general concepts of data, analysis and figure type matching. Learners engage with contrasting examples showing how the same data can tell different stories. Practices for variability reporting in visualizations is discussed in Lesson 5, while Lesson 6 provides practical guidelines for displaying multidimensional data. Finally, Lesson 7 educates about plot design choices to promote transparency in the data display and accuracy in interpretation, while Lesson 8 serves as a knowledge check exercise that provides learners with clickable dynamic choices to build a publication quality figure.
The collective goal of these lessons is to provide an awareness of what types of choices are implicated in rigor in data visualizations while providing just-in-time, focused skills practice. Will the unit answer every point about rigor in data visualization? Of course not. Rather, we propose that the initial step to creating more rigor is to be aware of critical choices and slowly, we can “make science better every day.”
Joining Sara on the OSU CoLAB team are Martin Haesemeyer, PhD, Cole Vonder Haar, PhD, Andrew Fischer, PhD.
