Snippet: How data visualisation can mislead

**2 min read**

We live in an era of ‘big data’ and ‘big information’. We work in an increasingly quantified world devoid of the nuance that qualitative information provides. We also live in an era of ‘scanning’. We don’t want to read or know detail. ‘Short reports’, soundbites, video clips and tweet-driven insights are the norm.

Through technology the world has accelerated to the extent that we cannot keep up with it, except when things are simple enough to absorb with the greatest ease.

Enter ‘data visualisation’ (dataviz) – something everyone loves. 

Accessible and useful data can contribute massively to the greater good. But if not detailed, nuanced or accurate enough in collection or interpretation, it can also be a toxic recipe that feeds disaster.

If we care about ‘credible evidence’, we have to engage with the values and ideologies reflected in data collection criteria, their interpretation and visualisation. We accept visual presentations such as charts to be true because they appear to present a scientific, expert, neutral point of view.

I was reminded of this when I recently again read the work of Joanna Boehnert of Ecolab. She has presented and written very convincingly (see also here) about the politicisation of data visualisation.

She beautifully demolishes the widely quoted data of Max Roser: “With this highly popular work, ideological agendas are served discretely and buttressed with what appears to be hard facts …. These data visualisations obscure complex problems and in some instances are severely misleading.”

DataWith excellent examples, Joanna demonstrates how his data visualisations
• conceal complicated realities;
• reflect particular values and ideological
assumptions;
• promote unstated political agendas and
parochial interests; and
• fail to capture power relations, attitudes
and behaviours that cannot be reduced
to a number.

Data visualisation can be a powerful tool to create simplistic, misleading narratives – flattening out complexity; defining things in ways that change the nature of the data and hence the message it projects; setting boundaries too narrow to be meaningful; and distorting through decontextualisation.

Do we look critically enough, with nuanced interpretations, at the data that inform our findings?

Let us take care to be vigilant about both the light and the dark sides of the increasingly superficial, over-simplified data and information worlds on which our work increasingly depends.

Share this article

Zenda Ofir

Zenda Ofir is an independent South African evaluator. Based near Geneva, she works across Africa and Asia. A former AfrEA President, IOCE Vice-President and AEA Board member, she is at present IDEAS Vice-President, Lead Steward of the SDG Transformations Forum Evaluation for Transformation Working Group, Honorary Professor at Stellenbosch University, and Richard von Weizsäcker Fellow at the Robert Bosch Academy in Berlin, Germany.

4 Comments

  1. I agree with you, data visualized poorly does not inform, it obscures what is actually happening. It also divides, since not everyone can interpret charts, so the story is not told.

  2. Very apt! Data visualizations can help understand our data, basic concepts, clarify the story for others and emphasize our message. However, poor data visualization can do the opposite!

Leave a Reply

Your email address will not be published. Required fields are marked *