What is social data literacy?

Every dataset is the product of choices. Someone decided what to measure. Someone decided how to categorise the people in it. Someone decided which questions were worth asking and which weren't. Someone decided how to present the findings.

Most of the time, those choices go unnoticed. When you've always measured something a particular way, it just feels like how it's done. That choice becomes invisible.

Social data literacy is the ability to see those choices. It enables you to recognise how social context shapes the way data is created, interpreted, and communicated, and to factor that into your work.

Where the choices hide

The choices show up at every stage of the data process: creation, interpretation, and communication. See The myth of neutral data

• Data Creation: someone decides what gets measured and what doesn't, which communities are counted, and which categories are used to group people.

  
  

• Data Interpretation: someone decides what the data means. The same numbers can support very different conclusions depending on the questions an analyst brings to them.

  
  

• Data Communication: someone decides what to include and what to leave out. These choices influence what an audience understands and what actions they take.

At every stage, you can make these choices actively or by default. The defaults aren't necessarily wrong—sometimes the standard approach is the right one. The problem is when you don't know you're making a choice, so any assumptions associated with it go unexamined.

What it looks like in practice

Social data literacy requires pausing at each stage of the data process to ask whose perspective is shaping what you're seeing.

A survey that groups everyone over 65 years together makes invisible the different experiences of someone who is 66 and someone who is 95.

When one team consistently scores lower than others in an engagement survey, the most familiar interpretation is that the team is doing something wrong. This is an example of deficit thinking (blaming people for their struggles instead of looking wider at what governs their environment). This low scoring team has been restructured three times in the last two years. Considering this, changes the question from what is wrong with this team to what has been done to them.

And here's a data communication example: a report shows that exits from emergency housing increased after a policy change. While this accurately describes what happened, it also leaves out where people went after they exited. Adding rough sleeping data from the same period (which also increased) changes what the audience understands and what they can act on.

You can make these choices actively, or you can rely on defaults. Social data literacy is knowing the difference.

Why social data literacy matters now

Data has traditionally been created, interpreted, and communicated by people. People made choices at every stage, consciously or not. As AI takes over more of these stages, the choices become harder to see, which makes the ability to notice them more important, not less.

This is the focus of my current research and writing. If you'd like to follow along as these ideas develop, the best place to do that is my newsletter.

Kat Greenbrook is a data storytelling consultant, author, and workshop facilitator based in Wellington, New Zealand. She is the founder of Rogue Penguin and the author of The Data Storyteller's Handbook.