Subscribe To Our Newsletter

Get tips and tools to tell your data story better.

No, thanks

 In Case Studies, Current Events, Data Analysis Concepts Simplified, Data Resources for Nonprofits, How To

Connecting the dots between funding and data equity isn’t always easy — or popular. But it’s important to understand the relationship between who is paying for a study and that study’s results. There is no way of getting around this:

Funding affects equity in data science.

Sometimes the impact is unintentional; sometimes it is less so. Let’s look at some examples.


What You See (and What You Don’t)

Goldman Sachs offers assistance to tens of thousands of college students. (Other financial corporations do the same — we’re just using research on GS as an example.) They run a program that assists some of the country’s neediest students, providing help with tuition, books, and transportation. They even provide emergency financial aid to students in crisis.

Research shows that this program makes it possible for these students to graduate. In some cases, it makes the difference between dropping out or going forward. They go on to get higher-paying and more fulfilling jobs — more than half of Goldman Sachs’ recipients show increased revenues.

All of that is pretty great. And the numbers definitely show that what Goldman Sachs is doing makes a difference.

But what are we not seeing? There is, for example, no research on how many students withdrew from their studies (or forwent college altogether) because of the 2007-08 economic crisis. A crisis brought about — at least in part — by Goldman Sachs.

Why don’t we have numbers on that?


Goldman Sachs funded at studies on how their assistance programs helped students graduate. They did not fund any research into how the economic crisis affected students. Obviously, lots of things that we do have a variety of impacts, and no one is going to measure all of them. It’s not specifically underhanded of GS to fund one study and not the other. But it does demonstrate how funding and data equity are inextricably linked.


What’s Being Hidden

This summer, the New York Times published a story about a much more worrying case of funding and data equity. A National Institutes of Health study on the possible benefits of moderate drinking was terminated after it came to light that researchers were working in close partnership with… you guessed it, significant players in the alcohol industry.

The study was intended to test the hypothesis that one drink a day is better for one’s heart than none, among other benefits of moderate drinking. But its design was such that it would not pick up harms, such as an increase in cancers or heart failure associated with alcohol, the investigation found.

Five beer and liquor companies were ready to fund the 10-year randomized controlled trial to the tune of $100 million. Researchers regularly communicated with representatives from the industry, fully aware that, with the right results, drink makers could use the study for marketing.

In this case, the link between funding and data equity — and the possibility for its abuse — is obvious.


Transparency is Key

Sometimes funding is deliberately used to manipulate research results. Sometimes, funding is only used to study one angle of an issue. Regardless of why funding and data equity are linked, it’s important to understand that they are.

So how can we address this issue?


Know Where Data is From

Knowing where our data comes from is critical. That, obviously, includes not only who collected and analyzed the data, but who paid for those processes. A data biography is a great way to introduce a level of transparency into your work. Make sure you understand not only how, but why data was collected. When we know the reason behind the data collection, it will be much easier to identify any motives that might have skewed the numbers.


Understand What Data Isn’t There

Recognizing the importance of missing datasets is also important. What numbers don’t we see? Why were they omitted? Would their presence change the results? Sometimes the information that isn’t there can say as much — or more — as the data that is.


Check Your Assumptions

It’s easy to assume we understand a situation when we have a cursory knowledge of what’s happening. But sometimes, there’s a big difference between what we know and what we think we know. Examining a situation from multiple angles can help us identify different perspectives — because frankly, objective data analysis isn’t really a thing.


Consult With an Expert

Not sure if there’s hidden bias in your data? Need help uncovering links between funding and data equity? At Datassist, we work in partnership with journalists, nonprofits, and social sector organizations to help tell accurate, educational, compelling data stories. We’d be happy to help your team. Get in touch with us now.


Recommended Posts

Start typing and press Enter to search

Inclusivity in design is about more than just not being overtly racist.We All Count - a project for equity in data