Data ethics and literacy
📥 Click here to download this document and any associated data and images
This section will include 1) key questions and concepts in data literacy, 2) ethical data practices, and 3) centering equity in data.
Data literacy
Data literacy begins with applying a critical lens and asking the right questions about data. Before using a dataset, you should always ask yourself: (1) who collected the data, (2) why did they collect the data, (3) what does the data say, (4) what does it leave out, and (5) how can it be used to support your goals?
Next, it is important to consider the limitations of the scale, or geography, used for analysis. When analyzing communities, census tracts are often considered the gold standard. However, they come with sampling limitations, especially at smaller scales. Not everyone responds to the census, so a lot of data is imputed. Aggregating data can also lead to misleading generalizations about neighbourhoods that don’t ring true from one block to another. It’s important to consider context and local definitions of the neighbourhood, since that can drastically change the story your data tells.
Finally, effective and trustworthy communication is the final step in building data literacy. Visualizations should be clear, credible, and honest about the limitations of the data. Avoid distracting visuals, “chartjunk”, misleading precision, and vague labels. Instead, aim for clarity and purpose in your design. How you frame your data also matters. Shifting from deficit-based narratives to asset-based ones can powerfully reframe discussions and impact how your analysis is received.
Data ethics
Using data ethically starts with recognizing the people behind the numbers. Behind every dataset are many people: data generators (that’s often you and me), data collectors, and data utilizers. Each group brings different intentions and potential biases, which can shape how data is framed and interpreted. Each time you encounter a new dataset, take a moment to reflect on who collected the data, who is represented in the data, who benefits, and who might be harmed.
Other best practices for the ethical use of data include protecting privacy, preventing reidentification, ensuring transparency, and making methods replicable. Following clear codes of conduct helps reduce the risk of unintended harm. Frameworks like the Locus Charter offer specific guidance for spatial data, emphasizing the importance of minimizing intrusion and safeguarding vulnerable populations. Another great resource to explore is these Ten Simple Rules for Responsible Big Data Research.
Centering equity in data
Centering equity in data means challenging deficit-based narratives and telling stories that reflect systemic injustices and community strengths. For example, mapping redlining alongside current patterns of displacement highlights how historical discrimination continues to shape neighborhoods. Reframing data around what was lost - like generational wealth due to exclusion - can deepen understanding and fuel advocacy. However, be aware of the way data is organized. For example, dimensions of group and individual identity can’t be examined in isolation from each other. Understanding intersectionality reveals forces of oppression and privilege and how they work differently across income, race, gender, ability, sexuality, and immigrant status.
Additional readings
The videos and content above represent a brief introduction to the topics of data ethics and centering equity in data. To learn more or dive in deeper, Wwe encourage you to check out the following additional readings:
Schwabish, J., & Feng, A. (2020). Applying Racial Equity Awareness in Data Visualization.
Below are the citations and resources mentioned in the video:
D’ignazio, C. & Klen, L. F. (2023). Data feminism. MIT Press.
Ethical Geo (2021). Locus Charter. Retrieved from:
Tufte, E. R. (1983). The Visual Display of Quantitative Information. Graphics Press.