After reading Drabinski’s ideas and listening to Kate Crawford I started wondering about how bias manifest itself in data visualization and what are the solutions being presented in order to tackle this problem. Media organizations, businesses and large public institutions resort to data visualizations in order to provide people with access to data, but it is important to remember and understand that data visualization remains a storytelling practice influenced by the perspective of the people who create and design them.
A simple presentation of a graph can drastically change and affect our perception of important public issues depending on how it is designed. An example can be the analysis of monthly change in jobs created by the NYT on 2012 where the authors give us the opportunity to scan the graph with Democrat and Republican lenses.
There are other cases where there is risk of unintentional bias. The usage of certain colors and shapes can alter the ways in which we visualize and understand certain social groups. Take as an example this map of concentrated poverty in Minnesota where the people living in these areas considered that they were being portrayed as an infestation.
The readings for this week reminded me of the importance of leaving room for discovery even when we may have a preconceived notion of how we wish to visualize the data and, above all, the need to develop techniques that promote transparency in our work as designers.
Thanks, Alonso! I’m really glad to see that you were able to connect this week’s reading to your own work in Data Visualization. I wonder if any of our authors proposed any solutions or strategies that you might translate to your practice — e.g., Perec’s experimentation, Foucault’s “epistemic rifts,” Drabinski’s queering pedagogy, Nowviskie’s call for systems that encompass other (Afrofuturist, indigenous, etc) ontologies, etc.