Scrambled, Over Easy, or Sunny Side Up?
How would you like your data?
Much less trivial than how we like our eggs is the question of how we like our data. While many of us would be quick to refuse the option of raw eggs, we’re far more willing to accept what we believe is raw data. After all, we may think, who can argue with statistics and cold, hard numbers? And how could straightforward, objective data be harmful?
However, some argue that this doesn’t exist— that saying “raw data” is equivalent to saying “jumbo shrimp”: it simply doesn’t make sense. So if we accept this claim and believe that numbers don’t equal neutrality, then we are forced to choose how we want our data “cooked” (D’Ignazio and Klein 159).
Despite our best efforts, information curated by subjective humans will never be completely unbiased. The key, then, is not to obtain neutral data, but to acknowledge its context, the lenses it has passed through, and the interpretations that allow it to be meaningful beyond numbers on a page.
“Interpretation is at the center of data analysis.” (Boyd and Crawford 668)
Context is everything from how the data is collected to the methods used to display its conclusions— anything that could introduce subjectivity. In an extreme example, if a study is done that concludes sixty-year-old males are more likely to contract diabetes, but the only people surveyed are sixty-year-old males with diabetes, what happens if we ignore that and blindly accept the conclusion?
Amassing a comprehensive frame of reference like this can be exhausting. Often it requires wading through scientific language that can be hard to understand or is not immediately accessible, especially if we see the data in a meme or social media post with no source information.
But to be responsible consumers, we need context. It limits our inferences and informs our interpretations, and without allowing it to influence the conclusions we draw, consuming “cooked” data is as dangerous as consuming it raw.
