CMS.631 | Spring 2017 | Undergraduate
Data Storytelling Studio: Climate Change
Instructor Insights

Misconceptions about Data Storytelling

In this section, Rahul Bhargava describes two common misconceptions about data storytelling students tend to bring to the course. The first is that data are objective and can point to truths unencumbered by social biases. The other is that one has to be a computer science expert to analyze data. He shares perspectives he offers students that counter these prevailing ideas.

The Truth Is, Facts Don’t Speak for Themselves

Many students come to the course believing that data are objective. This misconception is perpetuated by what Dana Boyd, founder of the Data & Society Research Institute, and Kate Crawford, Principal Researcher at Microsoft Research, describe as the branding of Big Data as a way to find patterns of truth. The idea is that if you have enough data, the noise in that data set won’t matter, and the truth will rise to the surface. It’s a perspective students have absorbed. If you put “Big Data” in any course title, you’ll get a waiting list of people who want to take it because it’s so hyped right now. 

In Data Storytelling Studio, I offer students a different perspective: Data isn’t truth, it’s a mirror of our socially constructed world. Most of the data sets we work with are, in some way or another, capturing human activity or the results of human activity in the physical world. So, of course, algorithms are biased based on our beliefs about culture, race, gender, class, etc. For example, it’s not unusual for a data set to leave out the perspectives of homeless people because this demographic tends to be marginalized in society. I’m explicit about sharing with students these kinds of biases in data. It doesn’t mean you can’t use data—it just means you need to recognize and account for the ways in which the facts don’t speak for themselves, but rather, reflect how we interact with each other at a larger social level.

In fact, the idea that facts could ever speak for themselves is a total misunderstanding of data. Everything from data collection (decisions about “who counts”) to presentation (choices about what kind of chart to use, where the vertical axis starts, what colors to use, etc.) comprise rhetorical decisions that change how someone understands what you’ve done. The minute you make the smallest decision about how to gather or present information, you’ve already turned data into speech. It’s not objective truth; it’s rhetoric.

You Don’t Need to be a Whiz Bang Computer Scientist to Work with Data

"Computers are helpful for data analysis, and yes, we’re going to use computer-based tools in the course, but your brain is also helpful for data analysis, and your brain can do something a computer can’t do: identify a story."
— Rahul Bhargava

The second misconception students tend to bring to the course is that you need to be a whiz bang computer scientist to work with data. I try to dismantle this idea by telling students that data analysis is just counting in special ways. I’m good at counting. My seven-year-old daughter is good at counting. This is something regular people can do. You don’t need to be an expert in the statistical language R to analyze data. My message is that, yes, computers are helpful for data analysis, and yes, we’re going to use computer-based tools in the course, but your brain is also helpful for data analysis, and your brain can do something a computer can’t do: identify a story.

In fact, many of the digital tools we use in the course are not built in a way to help you figure out which visual media will best tell your story. This is because you don’t tell the computer what your goal is. You can’t tell it that your audience already agrees with you or that your audience is predisposed to disagree with you. You can’t tell it that people are coming to your display after lunch or that you’re trying to convince people at a music festival to recycle so you need a display that will get their attention. In other words, a computer can’t help you figure out which technique might be appropriate for your audience. If you rely solely on a computer, you’ll show up at a music festival with a conventional bar chart on a poster. No one will stop to talk with you.  But if you show up with a giant bar chart made out of recycled trash, then maybe people would stop, because they’d be like, “Hey, dude, what’s up with your giant pile of trash?” And then you could talk with them about an important environmental issue. The tools are helpful, but data analysis is not all about computer science. It’s about using what you know about the world to make smart decisions about how you engage people with your data.

Course Info
As Taught In
Spring 2017
Learning Resource Types
notes Lecture Notes
co_present Instructor Insights