One year and 21 fascinating data conversations

A little over a year ago we were still in pandemic mode. Still in (relative) lockdown, and being very cautious about going out and seeing people. One morning, while making coffee, I said to my husband that I was thinking about starting a podcast. I’d already put my book, Raising Heretics, out as a podcast, and it was much easier than I expected.

Also, my day job running The Australian Data Science Education Institute is all about helping kids become critical thinkers and data literate problem solvers, but it had long bothered me that I wasn’t doing anything to help adults become more data literate. And I know a remarkable range of folks doing incredibly interesting things with data.

My husband, astute as he is, said “You just want an excuse to talk to people!” and… well… He wasn’t wrong! I love talking to people. An excuse to talk to people that might just educate and entertain other people at the same time sounded like an excellent deal to me.

Covid being top of my mind at the time, I reached out to Professor Adrian Esterman and Juliette O’Brien, both of whom were instrumental in keeping the public informed of covid related data, such as case numbers, hospitalisations, and deaths, as well as educating people about what was going on. I didn’t know either of them, though I had seen a lot of their work and heard them speak on different occasions. To my grateful astonishment, they both said yes. In fact Adrian offered the following Monday – just a few days away! – as a possible recording date, and suddenly I had to make this podcast real!

My dear friend Jed Wesley Smith wrote and recorded a theme tune for me, and I scrambled to put together some questions, thinking that they would surely have to change over time. The questions were:

  • Who are you and what do you do?
  • What did you have to learn to do your work? What was missing from your formal education?
  • What do you wish everyone knew about data?
  • What are the worst data mistakes you’ve seen?
  • Have you ever seen data deliberately misused? How can we spot things like that?
  • What’s the first question you ask when you look at graphs in the media?

I’ve interviewed an astonishing range of people, including an epidemiologist, a data journalist, an astrophysicist, an author and academic in cybernetics, a government statistician, an open data advocate, designer and storyteller, a climate science activist and communicator, an expert on fairness and ethics in data and machine learning, an economist, an ecologist, and so many more. 21 extraordinary experts, so far.

In some cases people reached out to me and offered to share their expertise, like the Australia Institute’s Dr Richard Denniss, who was keen to help people understand the use and misuse of economic data and pulled no punches as he did so. In other cases I hunted people down – some I knew, some I didn’t – and cheekily asked them to share their time and expertise. In all cases I was blown away by the enthusiasm and generosity of these extraordinary experts. I’ve had very few people say no.

I was surprised, too, that the questions didn’t get boring over time and need changing, though I did add one: “What excites you about data?” which is probably my favourite question now. Every guest has a unique take on the questions, a different and thought provoking way of looking at data, and the conversation tends to roam free, so while the questions provide a kind of framework, there’s no telling where we’ll go along the way.

Some episodes have made me really angry, like Polly Hemming talking about carbon credits, and all the ways climate data is used to mislead and obfuscate, or Richard Denniss and Margaret Hellard talking about why we’re not collecting covid data anymore, and what the consequences of that might be. Some have been confronting, like Jarrod Hughes talking about Indigenous data sovereignty. All of them have given me a lot to think about.

Some themes keep coming up. I deliberately ask everyone what’s missing from their formal education, because very few people have been taught to work with real data. Even now, I hear tales of Masters of Data Science courses that work with textbook datasets, where every curve is perfect, and there are no errors or messiness in the data. This is one of the issues I tackle in my day job. There’s no point in working with perfect data, because it doesn’t prepare you for real problems, where the data is always flawed, and complicated, and understanding the domain is key to figuring out the right analysis. It’s been both gratifying and disturbing to hear my guests reinforce this theme over and over again.

It’s fascinating, too, that many people don’t set out to work with data. Instead, as their career progresses, they realise that data is a powerful tool that can significantly advance their work. As The Guardian’s Greg Jericho put it, “Data is great for calling bullshit, but it’s also good for saying ‘yep, what we think is happening is actually happening and here’s when and how and why.'” They have picked their data skills up along the way, as a means to an end, rather than setting out to become data scientists.

I love this, because it fits perfectly with the way Data Science should be taught – as a way of solving real problems, of understanding the world. A tool kids (and adults!) can use to change the world. Not a set of cookie cutter processes that we apply to random lists of numbers, in a meaningless display of academic conformity.

My guests have been united in their desire for us all to cultivate greater rational scepticism when we’re confronted with graphs and statistics. Ask where the data came from, what the sample size was, what the flaws are. Ask why the y axis doesn’t start at zero, and why this particular graph type was chosen. Ask what story it tells, and investigate whether the story is justified.

Above all, don’t be afraid of data. It can be an extraordinarily powerful tool in any field, and you don’t need wildly technical skills in order to make sense of a spreadsheet. Imagine if we all had the power to ask for evidence, and to critically interrogate that evidence. That’s the world I’m working towards, and I’m so grateful to all of my Make Me Data Literate guests who are helping me along the way!

Check out the podcast here on the website, or wherever you get your podcasts. And if you’ve got a wild and wonderful data story to share, let me know. Perhaps we can work together to Make us all Data Literate!

*You can support this work at givenow.com.au/adsei/

Leave a Reply