Description:
[video 6 begin]
Spotting bias
Sometimes data may be generated, selected or analyzed in such a way that it favors one group, outcome, or conclusion. This is called data bias. Bias is another way to describe prejudice, partisanship, or favoritism. The Australian Association of Mathematics Teachers (AAMT) has identified several circumstances that can result in biased data: “survey questions that are constructed with a particular slant, choosing a known group with a particular background to respond to surveys, reporting data in misleading categorical groupings, non-random selections when sampling, and systematic measurement errors.” Bias can be unintentional or deliberate. All authors have subjectivity (a background that influences their opinions and ideas), but if an author is not careful, he or she may introduce steps in data gathering and analysis that produce an incomplete picture of the facts. When evaluating a visualization, consider who gathered the data being represented, who created the visualization, what the author’s message is.
Going back to the drone visualization from earlier, reflecting on the story being told there. What could be missing from the data communicated to you, the viewer? [image of point in visualization that mentions Obama]. What do you think the author’s purpose is in highlighting the point in time when President Obama was elected to office, and then a rapid display of drone strikes? Is there a larger narrative here?
Were there more drone strikes because Obama was in office? Or, can there be an unknown third factor that was responsible. Were there other historical factors? Or, did drone technology become more effective or less expensive? Are the authors of this visualization presenting the data in a way that inevitably leads you to a specific conclusion.
[video 6 end]