If Data Could Talk: The Ethics of Visualizing Data on Race

Author: Andy Cotgreave

Publisher: Tableau

Publication Year: 2020

Summary: The following podcast episode/video features the host and speakers discussing how our data is not race-neutral, and that we as a society are starting to wake up and be more aware of the inequities that hide within the data. Simply put, data is not just data. Every single data point, attribute, and calculation is a story. We can think that taking the mean of a continuous variable is a story. However, the story told is not objective. We as data scientists need to consider what point of view that means is coming from. Additionally, when we display a mean, we must understand our audience and ask if they know what that one number means. Are you able to see the people in the data? Do you know the risks of presenting this information to your audience? Can you potentially marry that number with others to bring a more diverse perspective? The podcast also introduces the idea of pipelines. There is an inequity in career and talent pipelines where having a lack of diversity in our teams leads to a lack of diversity in thought. There is also an inequality in the data and work pipeline. When conducting our work, its culture is quick and fast and it seems like there is no time to pause to consider ethical considerations. However, we need to be able to have that pause with diversity. While it may delay the product release, the discussion and transparency can find potential risks and biases in the product.