Data Equity and the Impact of Our Choices
6 Steps to Help Organizations Build Equitable Data Practices
This article shares highlights from a recent learning session hosted by Certified B Corporation ResultsLab and led by the Founder of We All Count, Heather Krause. The session focused on the foundations of data equity: What data equity means, why “data is objective” is a myth, and how you can begin to build a more equitable data practice.
At ResultsLab, we believe when making decisions we should always include the perspectives of the individuals and communities who will be affected by the outcomes of those decisions. Elevating the voices and lived experiences of individuals affected by inequities will strengthen communities and create a more just future for all.
We are a Certified B Corporation that works with other social good organizations to build capacity for this informed decision-making and are inclusive in our approach. We are on our own journey in diversity, equity, inclusion, and justice efforts as an organization and are committed to examining how our own social identities shape our worldviews and actions. We are continuously reflecting on and adjusting our beliefs and practices to best serve our communities. As part of this journey, we invited Heather Krause, Founder of We All Count, to share her expertise on increasing data equity in data science.
What Is Data Equity?
“How to use quantitative data so that it aligns with the intended experience of the people you care about.”
Let’s break that down a bit:
- Quantitative data is the numerical data, the numbers, the things we can count. This is most commonly where the false belief that data is objective comes in, and why this definition of data equity focuses on quantitative data.
- Intended experience, in this case, is the purpose of your research. What’s the question you’re curious about that you’re trying to answer with data?
- People you care about are the lived experiences, communities, and world views that you want to center. Whose perspective do you need to keep coming back to? There is no right answer here, and it will vary from project to project and from organization to organization.
Why the Belief ‘Data Is Objective’ Is a Myth
Now that we have an understanding of what data equity is, we need to understand that ideas like “data is objective,” “numbers don’t lie,” or “tools used for math are value neutral’” are all a myth.
We want to think data is objective, but we need to realize that we are making hundreds of subjective choices throughout a data project, and figuring out “what to choose is the heart of data equity.”
Understanding Types of Choices in a Data Project
Choices we don’t know we’re making. For example, when we look at the average classroom size, we can get two different answers that are both correct, depending on the perspective from which we choose to count our numbers. If we choose to center the perspective on the teacher, we get one answer; if we choose to center the perspective on the students, we get a different answer. Both are mathematically correct, but we made a subjective choice on whose perspective to center — and likely a choice we didn’t know we were making. Learn more in “Not Your Average Average.”
Choices we know we’re making but don’t realize affect equity. For example, when we choose to suppress data that is “not statistically significant.” In this scenario, we know that we are choosing to categorize a small sample size as “not statistically significant,” but we might not be aware of the equity issues this can create. Read more about the equity issues when using “not statistically significant.”
Choices we know we’re making that have equity implications, but we don’t know how to make the choice in the way that supports our equity goals. For example, when we design a research question we know we are making choices to craft this question, and we know this will guide the following activities in the project, so there are equity implications that could follow. However, there are slight variations in the way a question is asked, which can be problematic. How we choose to frame a question can put the onus to change on the wrong subject, creating a misalignment with our equity goals. The way we frame questions can shift and influence ultimately who or what is expected to change. For example, do we expect an individual to change or the system/environment that surrounds the individual? Read more on “Framing Research Questions that Reflect Who is Expected to Change.”
6 Steps You Can Take Right Now
Whether you work for a grassroots organization in a rural community or you’re the director of business intelligence for a corporation, you can apply these six steps and related tips and resources to your work.
1. Recognize that we are making subjective, human choices in our data work.
Tips and resources:
- Recognize how we make subjective choices in what we count. Revisit the average classroom size example.
- Recognize how we make subjective choices in the data we analyze. Revisit the equity issues when using “not statistically significant” example.
- Recognize how we make subjective choices in how we design our data projects. Revisit the framing research questions example.
2. Identify as many choice points in the data process as possible.
Tips and resources:
- Familiarize yourself with the Data Equity Framework — this provides a way to walk through the common steps of a data lifecycle, and learn how to notice the choices that are most common in each of those steps.
3. Try to make choices around data that reflect the equity we want to see.
Tips and resources:
- Continuously come back to whose world views or lived experiences you aiming to center. Are we amplifying the wrong narratives about the people and communities we care about?
- Go beyond the assumption that you are following best practices and test aspects of your data project with the individuals and communities your project is centering.
4. Expand the group of people who get to make meaningful choices about data.
Tips and resources:
- Inclusive Data Workshop: Learn and practice simple techniques for updating your organization’s data practices to better integrate and represent the community you serve.
5. Talk about our data choices and stand by them.
Tips and resources:
- Find communities where you can grapple with the real data challenges we all face every day, where you can share your thinking and test your ideas.
- ResultsLab Impact Collective CoLab: A community of practice that includes monthly connects with like-minded professionals to collaborate with and problem-solve to accelerate your work.
- Talking Data Equity: Informal discussion and Q&A series around applying data equity in the real world hosted by We All Count.
6. Be ready to try to make even better choices next time.
Tips and resources:
- There will always be room for improvement when it comes to equity. It’s an ongoing journey.
Recommended Resources
- Data Equity Framework by We All Count. This tool provides a way to walk through the common steps of a data lifecycle and learn how to notice the choices that are most common in each of those steps.
- Inclusive Data Workshop by ResultsLab. Learn and practice simple techniques for updating your organization’s data practices to be more integrated with and representative of the community you serve.
This article was originally published by ResultsLab.com. B The Change gathers and shares the voices from within the movement of people using business as a force for good and the community of Certified B Corporations. The opinions expressed do not necessarily reflect those of the nonprofit B Lab.
