When created with intention, data-based research can present a rich tapestry of evidence. Each data point a length of yarn, intertwining to present an engaging visual reflection of the research. But without enough high-quality yarn, enough colors and lengths, enough time spent weaving, or enough consideration of the final design from the start, the tapestry will have holes, tears, and lopsided edges. Worst of all, it will not accurately convey the intended design.

In the original Do No Harm Guide, we distilled interviews with nearly 20 data experts to demonstrate how data practitioners can avoid these kinds of pitfalls when collecting, analyzing, and presenting their data. The guide centered empathy as a key tool for ensuring equity in data work, asking researchers and analysts to consider how their data were collected, whose input they had sought out, and how they ultimately presented their work.

For this follow-up volume, Do No Harm Guide: Additional Perspectives on Data Equity, we handed the pen to data experts and practitioners whose voices have been traditionally underrepresented. As with the original guide, the lessons within this accompanying volume are not meant to be prescriptive—rather, they should inspire greater thoughtfulness and consideration in data work.

Key Findings

Through the five included essays, the authors explain how data practitioners have historically fallen short in data collection, application, and representation, leading to lasting harm and distrust of researchers and fewer pathways for advancement in data fields. From these essays, researchers and data analysts should consider the following:

• Incorporate community-based participatory research. In many cases, researchers and data analysts approach their work from a position of cold detachment, treating communities and people transactionally. Through the principles of community-based participatory research methods, researchers can more equitably work with communities and establish positive, collaborative relationships. These principles include creating long-term relationships, focusing on assets instead of deficits, creating transparent summaries of findings, and much more.
• Seek advice and input from the communities. Countless documented cases exist of researchers inflicting harm on the communities they mean to study. As a result, many of these communities, which are already marginalized, have come to distrust researchers. To better work with these communities, researchers and data analysts should create a two-way street of collaboration and partnership. Rather than imposing a set of research questions and methods on a community, researchers can openly seek input and advice from the community.
• Create a pipeline for more diverse data practitioners. Data and data projects can only be made more equitable if work teams are diverse, reflect a variety of experiences, and are trained to foreground empathy and equity in their work. Data practitioners can also learn to apply their analysis through an equitable and inclusive lens. Using and refining existing data to create more informed treatment or policy options can lead to more equitable outcomes.

By considering these lessons when collecting, analyzing, and communicating data, researchers and analysts can create work that is not only more equitable and inclusive, but also more likely to be embraced by the people who are best positioned to use it in their lives and in their communities.