Going back to an earlier session on Epistemologies of DH, I wanted to reflect on a class discussion I found particularly interesting. One of the questions asked for the reading, “Why Data Science Needs Feminism” by Klein/D’Ignazio was “In what ways has data/technology been a tool of oppression but also a site of resistance?” My group discussed examples from readings outside the course, real life, and from the discussion text.
From the perspective of data being a tool of oppression, we discussed an example provided by Ruha Benjamin’s Race After Technology. She shares in her book how an algorithm created to track the recidivism rate of prisoners in Florida inaccurately predicted that black prisoners would re-offend at higher rates. And that’s just one example of the antiblack climate being perpetuated in justice systems. There’s also predictive policing software that leads to the over-policing of black neighborhoods because the data being fed into the software reflects surveillance priorities. I remember Benjamin stating that crime prediction algorithms were more like crime production algorithms.
We also shared ideas about future tools of oppression. For example, the OMNY system because of the type of data being collected and tracked. Why would OMNY need to collect our mailing address? Are we being profiled based on where we live? Another potential system discussed was the policing of sex trafficking with big data. The article brought up questions about a possible distinction between voluntary sex work and human trafficking. According to the article, we must be wary of labeling certain categories of sex workers as ‘trafficking victims’ when drawing only on open-source data.
As for how data could be a site of resistance, we drew an example from the discussion text. Darden, a black female data analyst at the NASA Langley Research Center realized her scientific accomplishments weren’t as acknowledged as the work done by her male counterparts. After consulting with Gloria Champine – a white woman working in Langley’s Equal Opportunity Office who had been compiling statistics about gender and rank- Darden’s realization was confirmed. The data revealed that the promotion rates of men and women with similar academic credentials were very disproportionate. She then takes the numerical data and creates a bar chart which she displays to her supervisor. He was “shocked at the disparity” and the chart proved to be the perfect solution to communicate a systemic problem. I think presenting the data in an easily digestible format worked in Darden’s favor. A bar chart is easy to read, and the simple design communicates the point without it being overwhelming.
Overall, I found our discussion very informative, and it was nice how we tied in readings and experiences from outside the course to answer the questions asked.
Article references:
