Tag: DataScience

One of the things ADSEI does in its lesson plans is ask the question: What is wrong with this data?

This is a really crucial question, because there is no such thing as a perfect dataset. All data has issues. Often its not the data you want, it’s simply the data you were able to get. For example:

• whale observations tell you how many whales were seen, when what you really want to know is how many whales were there. Some whales might have breached but not been observed (shades of Schrödinger’s Whale), or swum by without breaching, or even been spotted twice but accidentally counted as two whales when it was really just the one.
• speed cameras tell you the instantaneous speed of the car when what police really want to know is: has that car exceeded the speed limit at any time on this trip?
• counting the litter found in the schoolyard tells you how much litter you found, not how much litter was dropped – some of it may have blown away or hidden under things. It also only tells you how much litter was there that day. What if a year level was out on excursion, or it was a wet day timetable…

And even when the data is actually what you want, there may be data that’s missing or flawed for various reasons. For example:

• Facial recognition systems that were trained on images of faces that were almost exclusively white and male.
• Phone polls that can’t include people with unlisted numbers.
• Internet polls that can’t include people without internet access.
• Surveys where people don’t or can’t tell the truth – for example about healthy eating, or sexuality, or where people don’t actually know the truth, for example about why they did things, or things they don’t remember (like what did you have for breakfast yesterday? Or how often do you eat broccoli?).
• Skipped data where someone forgot to record a daily observation or the system went down and didn’t record any values.

Consider the reporting around the Corona virus. We have a reported death rate of around 2% which is highly speculative, because we have no idea how many mild cases of corona virus there are out there that are not being identified or reported. Some sources report the numbers and stress the uncertainty, while others report them as solid facts.

This is a kind of scepticism and critical thinking that we don’t often leave room for – in education, business, or journalism. Often we are in such a rush to get the “right” answer that we don’t have time to pause and evaluate the data we’re working with, to consider the flaws and uncertainty that are built in to any dataset, and any analysis.

If we can teach our students, from pre-school onwards, to question their data, to ask “how many ways is this data flawed” rather than assuming the data is perfect, then perhaps we can build a world which centers critical thinking and evaluates evidence.

This is why using real datasets rather than nice clean sets of fake numbers is crucially important to teaching data science. Because real world datasets are never nice, clean, and straightforward. There is no need for scepticism and critical thinking in textbook examples. But kids who have used real data in their learning are equipped to tackle real world problems.

Can you share some examples of flawed data? What consequences have you seen from people assuming data is perfect?

I want to start by asking you a question: What gets you out of bed in the morning? What really motivates you?

For me it’s the chance to make a difference in the world. It’s wanting to leave the world a better place than I found it.

And that’s something that STEM skills are perfect for. They are for problem solving, for designing better ways to do things. For bringing clean water, clean power, increased food production, solutions to climate change, safer transport, personalised medicine, and a whole host of innovations to the world.

But when I first started teaching in a high school – a science school, no less – we were teaching “STEM” as “fun stuff”. Drawing pretty pictures. Making robots follow a line. Playing with toys.

How many of us are motivated, I mean really motivated, by toys? Some of us are, especially technical people! But those are generally the people we’ve already GOT in tech! I’m much more interested in the people we haven’t got yet.

All too often we ask those kids who are not already into tech to get out of bed for the chance to have fun. And fun is great – I like to have fun, we all do! And not all of my fun is finding an interesting new dataset and analysing the hell out of it, I promise. I do have other ways of having fun besides writing an interesting new Python script. Really I do. But fun doesn’t get me out of bed in the morning. Fun is a hobby. A diversion. A toy. That’s not what we need kids to understand about STEM.

We are handing our kids a world in desperate need of creative solutions. Of innovation and entrepreneurship. Of change.

And we’re telling them that STEM is fun! It’s for designing 3D jewellery. It’s sparkly. It’s pink. It’s useless.

We are doing kids a huge disservice. They’re kids, therefore fun is the way to reach them, right? It’s like saying we want more women in tech so we’re going to paint some things pink and offer some courses in the chemistry of makeup (a real suggestion that was made at an actual school). It’s like saying “women do hardware too, let’s sell them some pink hammers.” (and that’s also a real example)

When we were teaching computing using “fun toys” the overwhelming feedback I was getting – from science students – was “Why are you making me do this? It’s not relevant to me. I don’t want to do it.”

Can you guess what happened when we made the year 10 computer science course a data science course instead of a “fun toys” course? We were teaching the same basic coding skills. We still had them learning about selection, iteration, variables, and functions. But now we were using real datasets and finding real questions to answer, real problems to solve. Do you know what happened?

Suddenly they could see the point. They found it useful. They found themselves using the skills in other subjects, especially in project work. And the numbers who went on to the year 11 elective computer science subject increased by around 30%, with double the number of girls.

And none of it was pink!

That first data science course I had a student who was super interested in politics, and there was a federal election, so we used data from the Australian Electoral Commission. Turns out you can download csv files containing every single vote from any Australian election.

We used the senate votes for Victoria for the 2016 Federal election. Over 3 million lines of csv, they contained polling booth, electorate, and a 151 position comma separated string containing the contents of every box on each ballot paper.

3 million lines of csv won’t even open in excel, so the kids had to program just to open the file. They learned about using a small section of the file in order to test their code, so that it didn’t take ages to run. They learned about what questions a dataset could answer.

They found their own questions – from which party’s voters were more likely to follow the how to vote cards, to where Pauline Hanson voters came from. They asked questions about their own electorate or polling booth and how they compared to the whole state. About female representation and share of the below-the-line vote. About preference flows and about how polling compares to actual results. Every student asked a different question, which meant that every student had to write different code to find the answer (goodbye plagiarism!).

And then the important part happened: they had to visualise their results. To create an image, more interesting than an ordinary graph, that conveyed their results in a convincing, valid, and compelling way.

They learned about channels of information, about the human visual system and attention. About colour blindness and the problems with the rainbow scale. They learned which types of graph are appropriate for different types of data, and how to customise their graphs so that they don’t mislead their audience.

As well as learning to analyse and visualise data themselves, they also learned to be critical data thinkers, reviewing graphs and statistics they are presented with using critical questions like “How was that data collected? What was the sample size? And where is the zero on that scale?”

We have a tendency to bend at the knees when presented with statistics and graphs. It seems to automagically make information more credible. But they are very easy to manipulate. So it’s crucial, in this era of fake news and anti-science, that our kids learn to be critical thinkers.

Another reason we need our kids to learn data science skills is the increasing dominance of Big Data and Machine Learning in every aspect of our lives. They are determining our healthcare and our access to home loans. They’re directing our traffic and influencing our consumption and behaviour – even our votes! They’re controlling our justice systems and our borders. But how many of you really feel like you have a good understanding of how the algorithms that do these things actually work? How many of you are confident in the fairness, impartiality, and accuracy of these systems?

And this is a highly educated audience. Think about that for a moment. These systems are running our lives and we have no say in how they operate. We don’t even understand them.

So it’s crucial that we educate upcoming generations to have informed, intelligent conversations about these systems. So that we can have that long delayed community conversation around the way we manage our data – and the way it manages us.

And to do that, we need to engage kids with data in the classroom. To show them its relevance, and to build their Data Science and technological skills.

The problem with finding cool datasets and building them into interesting lessons is that it’s hugely time consuming and highly skilled work. When I used the electoral data it took me hours to make sense of the dataset. I couldn’t even find anyone in the electoral commission who could explain it to me, so I had to derive it from first principles. The only reason I had the capacity to do that is that I was part time, so I used my own time, unpaid, to find the dataset, make sense of it, and build a project around it. Most teachers simply don’t have the time to do that – or, to be honest, the skills.

It’s also important to acknowledge that student motivation is not the only issue we face in teaching tech in schools. The problems are many. Tech has an image problem almost as bad as teaching does! So kids don’t see themselves as the type of people who go into tech (and this affects boys as well as girls).

We attract the kinds of people into tech that we already have – generally people with a very narrow personality and background distribution. This conference is obviously full of the exceptions to that rule. 🙂 But it’s a real problem if you want innovative solutions that meet the needs of everyone, not just the tech nerds of the world.

We lack skilled teachers, in part because the correlation between that classic tech personality type and the kind of person who loves to teach seems to be, frankly, quite low, but also because if you have tech skills you can EASILY earn a LOT more and work a LOT less hard by NOT going in to teaching. But we also have a large cohort of teachers who are flat out terrified of technology. So if we force those teachers to teach our shiny new Digital Technologies curriculum, they can’t help but convey that fear to their students.

That’s why I founded the Australian Data Science Education Institute (which, by the way, is a registered charity). To find and make sense of the datasets, to build cool projects around them that are aligned with the curriculum, and to train teachers in the skills they need to incorporate data science into their teaching. We start from where teachers are and build their skills gradually, in the context of their own disciplines.

We don’t expect them all to program on day one. We start with spreadsheet skills and projects that both teachers and students find relevant and interesting.

Using Data Science teaches kids why STEM matters, and gives them the opportunity to use STEM skills to change the world. So we use this template for finding, analysing, and solving problems in the local community.

• Find a problem
• Measure it
• Analyse the measurements
• Communicate the results
• Propose a solution
• Implement the solution
• MEASURE IT AGAIN

And that’s the crucial part that we need to make the default position anywhere where we try new things: That we measure & analyse them to see if they work. Because in governments, in schools, in businesses: too often we see new programs implemented as a matter of ideology, and the only “assessment” that happens is for the champion of the program to say “It was awesome!”

And when you say “How do you know?” Everyone goes suspiciously quiet and changes the subject.

Incidentally, that’s why ADSEI collects feedback data on all of its courses, and why we’re also building a feedback mechanism for our online resources.

We also have a template for exploring global issues:

• Find a dataset
• Explore & Understand it – and this means understanding the domain, a fact we tend to lose sight of.
• Find a question it can answer
• Analyse it to find the answer
• Communicate your results

ADSEI’s ultimate goal, of course, is to put itself out of business. To build Data Science into the way teachers are trained to teach. To build a community of Data Scientists and teachers who can support each other by sharing resources, project ideas, and cool datasets.

I think my job is safe for the moment!

For now we have grants from the Victorian Department of Education and Training, Google, and the Great Barrier Reef Foundation. We’ve developed teaching resources for Monash University, CSIRO, and the Digital Technologies Hub. We have delivered workshops and talks at conferences and schools, and we are working with the wonderful people at Pawsey Supercomputing Centre and the West Australian Marine Science Institute.

And ADSEI has only been in existence for 18 months.

Over the next few months we’ll be running workshops in Perth, Melbourne, and Alice Springs.

Next year in October we’ll also be running the Inaugural International Conference on Education and Outreach in Data Science and High Performance Computing, with the support of the awesome Australasian eResearch Organisation – Sponsors welcome!

So if any of this sounds like a mission you can get behind, join the slack channel, check out the website, send me an email (linda@adsei.org) or tweet at me wildly. Because Data Literacy and Data Science skills are something all kids need to experience, before they decide that Data Science is too hard, too boring, or not relevant to them!

If Data Science is going to drive us to the future, I want to put all of our kids in the driver’s seat!