Part 2: my zine review of Opening Data 2 from 2017 by the Allied Media Project, the Detroit Digital Justice Coalition, the Detroit Community Technology Project, Our Data Bodies, and the Media Democracy Fund
The follow-up companion to Opening Data Volume 1 digs deep into community research. Whereas Opening Data 1 focused on facts, analysis, and theory from a data justice lens, Opening Data 2 gets its hands dirty with interviews, workshop agendas, and worksheets. In the democratic and open source spirit of the information being provided, the reader is asked to do further exploration of how data shows up in their own community.
The most compelling idea from this zine may seem rather obvious—that is, data can be both beneficial AND harmful. Meaning, the same set of data can be used to harm people or to help them. A data set isn't inherently good or inherently bad, it's all about how actual people analyze and use it. Focus groups in Detroit were asked to imagine and identify perceived benefits and perceived harms of data sets on the Detroit Open Data Portal.
I thought that I might practice this line of inquiry within Tableau's COVID-19 data hub as though I were Donald Trump, our nation's most harmful citizen. Trump is a notorious cherry-picker with pandemic data, using whatever statistics suit his talking point (generally, that everything is okay and that he is the best). For example, although he refuses to use rising cases as an indicator of a problem in the US, he uses that metric to cast Europe as the epicenter of the pandemic. So, as Trump, I took a look at a data set.
First of all, I don't think Trump would be able to make sense of this gorgeous visualization. It took me a bit of time to understand it as well. It would be very easy to have a look at this infographic and say, "Look how blue that map is on the left! We're ready! Open up!" According to the Opening Data rubric, this data might cause harm to the community. If cherry-picked, it might encourage the Trump administration to hasten the country's reopening since they are looking for evidence to prove a point. To be fair, they seem skilled at cherry-picking literally any kind of data they're presented, no matter how it is presented.
In terms of benefits, generally, tracking corona virus data is helpful for our communities. For example, this visualization gives in-depth, data-driven answers about reopening down to specific cities. Granularity in visualization and presenting the viewer with the ability to "drill down" has become more interesting to me the more I learn about data. In his book "Envisioning Information," Edward Tufte encourages very complicated visualizations—the more layers the better. However, can trend lines get lost and usurped in very complex data? Who is our audience when we create visualizations? Finally, is cherry-picking and harm inevitable in the wrong hands no matter how a chart is constructed?