STEM, Part culture war, part cargo cult: My latest Fin column

The Future of STEM (Science, Technology, Engineering and Math)Here’s yesterday’s op ed for the Fin published as Technology education is about more than funding:

STEM is all the rage in education – that’s Science, Technology, Engineering and Maths. Part culture war against Australian mediocrity, part cargo cult, a principal goal is more money – for universities and school education. It’s hard not to agree. Lateral Economics’ HALE index of wellbeing values Australian human capital at $18 trillion – over three times all other physical and natural capital combined. Growing it puts every other means of enriching our future in the shade.

Still, I smell a rat.

We’ve nearly doubled educational spending per student in the last few decades. That’s funded popular measures with little impact – smaller class sizes – and politico/educational fads some of which have proven disastrous – like whole language learning. If STEM is simply cranked up and bolted onto the existing system, expect business-as-usual, expensive-as-usual disappointment.

Traditional STEM teaching often turns kids off. If they were ever invited into the debate, they’d see the STEM agenda as baby-boomer finger wagging telling them to eat their greens. They’d ask what exciting jobs will exist for STEM graduates. They’d ask what STEM skills will be in demand in ten or twenty years. And we don’t know.

The vast riches of Silicon Valley use STEM skills sure enough. However not only has old-fashioned entrepreneurialism been the dominant input, but the main STEM contribution has been computer coding. While that’s taught in universities, the valley is full of practitioners who’ve mostly taught themselves with the help of free internet resources and their own workplaces. Silicon Valley has embraced data science but no thanks to university courses. Even today, while they crank out stats grads, our universities barely teach data science.

The innovation we desperately need to get STEM right is institutional. In 2010 I proposed a different approach lunching with the secretary of a state Education Department followed by discussions with his senior managers. I suggested we tap into free resources all around. The net is brimming with free resources. Want to learn how to build a website or learn JavaScript? Head to CodeAcademy.com.

Second, while teacher training, support and specialisation in STEM should be better resourced, on its own that would achieve very little. The last thing we should do is insist on widespread STEM in-service training for existing teachers – for instance in teaching computer skills – who’d simply go through the motions.

Meanwhile an immediate human resource is students. We should find those already doing it for themselves and empower them to enthuse and teach their peers – not to mention reverse mentoring their teachers. And if we’re to do that, we must make room for it in kids’ timetables and in the recognition they receive – their marks.

However that requires some real transformation of existing routines and priorities. And incumbent organisations find that almost impossible. Much better to seek funding for some new, bolt-on initiative. As we’ve loaded the curriculum with recent enthusiasms and political correctness, what priorities have we jettisoned? Stats was more useful than trigonometry even when I was a kid: Much more so now. But the relative weightings in the maths curriculum haven’t changed in 30 years. I learned more about computer coding in school in the the early 1970s than my kids have done in today’s schools.

What I’m proposing can’t simply be ‘rolled out’. Just as a manufacturer wouldn’t release a new product without extensive design, prototyping and testing, that’s what should happen here. We should draw out in-system entrepreneurs, fund experiments and pilots, fixing or jettisoning the failures, identifying, tweaking and growing successes and rewarding those behind them.

After a long lunch with the Education Department Secretary whom I knew from another life in which I extemporised on these issues, he invited me to a showcase of students’ achievements in IT projects. There I met Ben, a year 8 student. He’d built an iPhone app to hone his brother’s mental arithmetic.

“How do you find maths” I asked.

“Boring! We keep doing the same stuff.

“How’d you like to teach other students to write iPhone apps?

“Awesome!

“Wait right there.

I fetched the Departmental Secretary. Here was an opportunity to get going with what I’d proposed. Excited, he summoned his Innovation Chief saying “I want to start on this tomorrow!”

The next year I asked Ben how things had gone. I still have his reply: ”Nothing really went anywhere with my school, didn’t really surprise me”.

The STEM agenda could handsomely enrich our future, but only if it’s part of wider transformation which, though it would cost nothing, offers a richer prize than any amount of new STEM funding.

Postscript: the interview of the blogpost of the column on this page with the mp3 here.

This entry was posted in Cultural Critique, Economics and public policy, Education, Innovation, Science. Bookmark the permalink.

33 Responses to STEM, Part culture war, part cargo cult: My latest Fin column

  1. derrida derider says:

    But the problem with telling kids to eat their greens is not only that they won’t – its that eating your greens genuinely IS good for you.

    And so is formal, rigorous training in maths and computer science; that kid would write far better iPhone apps if he had such training. Lord knows there is more than enough inventive but simply awful code emanating from Silicon Valley – Vista and similar disasters did not come out of nowhere.

    Now a talented teacher using a well thought through curriculum can make sure “rigorous” is not an antonym of “boring”, but such talent and thinking is always going to be scarce.

    • Chris Lloyd says:

      “Even today, while they crank out stats grads, our universities barely teach data science.” Advertising alert: Melbourne Business School have just launched a Masters of Business Analytics (which is data science with a sexier title). These students will have no problem getting good jobs. They are almost employed already and it is only first term.

    • Chris Lloyd says:

      Sorry this was not supposed to be a reply to you DD.

  2. derrida derider says:

    Whoops – for “antonym” read “synonym”

    • Nicholas Gruen says:

      Thanks DD,

      I agree. I plead only the demands of compression into 720 words.

      No question that teacher quality is crucial. But it maddens me that the institutions can’t be focused on the thing itself, which is the quality of the learning for students, not teacher qualifications or any other thing. Those things can be powerful means to the ultimate end, but always figure in these kinds of debates as ends in themselves.

      However there is a deeper issue in my mind but not taken up in the column at all. STEM disciplines are taught as domain subjects in school. They’re understood as a body of technical capability. So maths consists of a whole bunch of things we do to numbers in primary school and as you go up through high school it becomes a wider, more formal set of capabilities – trig, calculous, algebra etc.

      This is also true at uni and in other disciplines. So law comprises a few general subjects – an introductory method course and an elective reflective method course – philosophy of law. The rest is just a bunch of subjects that you pull down off the shelf. Contracts, Trusts, Companies, Family, Crim and on it goes. I think that’s a pretty dismal way to conceptualise both knowledge and education.

      In case this sounds as if I want students to study lots of philosophy and methodology, I don’t. I want them to do what one does in history at uni – which is to do the practice in a critical, reflective and above all useful way. And I’d like to see the various disciplines related to one another.

      There’s almost none of that – it’s expected to come by osmosis or from having a bolt-on subject – philosophy of law or economics or mathematics or education or whatever.

      Economics education is a kind of caricature in which students make a mad dash firstly to pick up the neoclassical superstructure – budget lines, indifference curves – the kind of stuff that had Keynes marvelling at its simultaneous cleverness and uselessness of the whole construct “Simple things are made to appear very difficult and complicated, and the emptiest platitudes paraded as generalisations of vast import”. Undergrad students often – I’d say usually – remain in a kind of unarticulated and resigned shock that they never really engage with ‘the economy’ or economic life, but instead are inducted into seeing it all through neoclassical constructs like budget lines and indifference curves. It’s not as if such constructs mightn’t be useful in some circumstances (not that many mind you, but some). But they’re taught as the building blocks of the discipline (when they’re the building blocks of a particular family of techniques and way of thinking about the economy).

      Then there’s the dash for technique as students learn one model and one technique after another. What is not taught, or what is assumed will be picked up by osmosis – assuming it’s regarded as a problem at all – is the task of using the techniques in a ways that are well judged and helpful in the circumstances.

  3. I learned more about computer coding in school in the the early 1970s than my kids have done in today’s schools.

    Really? Given that PC’s weren’t even around in the 70’s, I thought coding would not have appeared in schools until the late 80’s.

    • Andrae says:

      Not all computers are PC’s. These days most computers are phones—back in the 70’s, most computers were batch oriented mainframe and time-sharing mini-computers.

      As a kid growing up in the 80’s I was a direct beneficiary of the enthusiasm and exposure kids in the 70’s were given to punch card, teletype, and later dumb-terminal based computers in their schools.

      A classic essay on the topic was written by the fiction author Neal Stephenson in 1999 (http://www.cryptonomicon.com/beginning.html)

      • Well, Andrae, as a high school student in the Queensland State system in the mid 70’s, I had no exposure at all to coding, and I followed mainly a maths/science stream. And really – don’t kids in high school now who want to go down an IT route get some exposure to it?

        I’m not meaning to be rude here – but it just seemed to me that line sounded a tad exaggerated. Or, at least, it didn’t sound like it reflects the general situation (regardless of Nicholas’ own experience) very accurately.

        • Nicholas Gruen says:

          Steve,

          I think your right that my experience was a little unusual. It was a private school. But there would be plenty of ways to give our state school kids access to coding instruction.

          You can say ‘we don’t have the money’ but we do if we make this about the kids and the education and not about business as usual which is institutions and the grown ups’ interests first.

        • Patrick says:

          You can’t say “we don’t have the money” because all you need to do is give them an hour a week (or two!) with computers (which nearly every school has now) and codeacademy or khan academy or code.org or any of the numerous sites that have free coding lessons.

          You could do more, but that would be a good start.

      • Nicholas Gruen says:

        Steve, the little computer coding I did at Haileybury College in third and/or forth form (year 9 and/or 10) was minitran, some cut down variant of fortran. We didn’t do that much of it, but it was done on punch-cards which were packaged off to Monash Uni’s computer and came back as printouts the next week. Almost invariably one had made some small mistake and the program wouldn’t run.

        It was pretty trivial. Neither of my kids have learned any coding whatsoever. I know more html than they do. It’s important not to misunderstand what I’m suggesting here. I’m not embracing coding as the cargo cult of the future – there’s so much of it going on you really have to learn it all and it will make you rich or well informed in the internet age. I’m OK with the idea that it’s a specialised skill that many people won’t learn.

        The thing that makes me think it’s an excellent thing to be at least an important elective, though I expect I could easily be convinced that some coding familiarity is as important as basic maths is that coding is a discipline by which one takes ideas out of a human head and conveys them in a logical form to a machine. That requires the kinds of skills that maths teaches. I’ve observed that those with strong domain knowledge of coding use it as a rich source of metaphors about meaning and about life. It’s a world of thought of great simplicity and power.

    • David Walker says:

      In the mid- to late-1970s you trekked off to some institute in a distant suburb and used a mainframe. (I also had a mate whose dad was an astronomy lecturer, so we got a bit of time on an Adelaide Uni mainframe …).

      Like Nick – and contrary to the idea which says kids these days are so computer-literate – my kids mostly know just enough about computers to get done what they need. I did a deal with the 15-year-old in a recent holiday where he sat down and learnt some Python off a Youtube course, just to get a feel for how it all worked. Python is a data analysis language too, so we figured it was a good chance to be useful somewhere, sometime. Highly recommended.

      Now to teach him about relational databases …

  4. R. N. England says:

    Bureaucrats can’t do without statistics. Engineers can’t do without trigonometry.

    • Nicholas Gruen says:

      Thanks R.N. Both bureaucrats and engineers can learn, and will learn all sorts of domain skills. I think some familiarity with stats is a general literacy. The papers are full of implicit statistical arguments – so a little more familiarity with statistical reasoning seems to me to have general usefulness. Likewise in one’s life – making decisions about what to do – particularly financial decisions – some basic familiarity with statistical concepts seems useful to me.

  5. conrad says:

    “and politico/educational fads some of which have proven disastrous – like whole language learning”

    You are confusing the use of money with something where all the free gains are. Things like language learning and so on are basically implemented by each Department of Education in the States and then schools are told to teach like that. For things like language learning which take more or less no resources to teach apart from the teacher, teaching them well should of course cost you nothing more than teaching them poorly. So the problem is really who is dreaming up the fads and forcing people to teach them. Thus when you are thinking of things that are “bolted on” this is really not the fault of schools, but those who are telling them what to do, but these are people least likely to listen — why endanger their own jobs by getting new ideas in?

    On this note, as I’ve found out recently, there are places like SA that are truly appalling. Their *good* schools (e.g., Unley High where Julia Gillard went, and even expensive private schools like Scotch which top the TER scores in Vic) are getting results only a bit above average for Aus (which is terrible given that Aus isn’t doing too well at things like maths anymore — just imagine what the poor schools are like). This suggests that the problem here is the top echelon of education bureaucrats running the show, who should be sacked for such an abysmal outcome affecting so many students, or if you are nicer, told to fix things in, say, 5 years or get the chop (maybe I’m just griping here). Perhaps you might have some better ideas about how to fix the root of this problem, which might then at least allow newer ideas to enter into the debate.

    • john Walker says:

      Conrad
      Re ” there are places like SA that are truly appalling”
      Is this from published information?

      • conrad says:

        These are the NAPLAN high school results. I personally don’t think the NAPLAN has super validity (the primary school one is especially bad). Alternatively, when, say, your 20th best school, including elite private schools taking 25-30K a year is on the 72nd percentile for Aus, you have very serious problems. Like-for-like comparisons are good too. Scotch in Adelaide is on the 68th percentile (!).

        http://secondary-school-ranking-adelaide-south-australia.street-directory.com.au/

        • john Walker says:

          Thanks,
          The ACT figures also look a bit ‘odd’, care to comment?

        • conrad says:

          Not sure what is odd about them — remember that the ACT only has a population of about 400K (1/4 the amount in SA), and I suspect it might have a different age distribution too.

          As a random comparison, by just multiplying the numbers by 4 to equate Canberra and SA (obviously there are lots of other factors), if you took the 10th best school in the ACT, you get to the 68 percentile, whereas the 40th in SA gets you 61 (+10%). Similarly, if you took the 20th school in the ACT it gets you to the 54 percentile whereas the 80th in SA gets you to 51. So the top end is more toppy in Canberra, whereas the middle end gets closer. SA gets worse at the bottom end (not in these figures — I can’t find a good online source right now), but there are serious demographic differences in groups there compared to Canberra which makes comparisons hard (i.e., lots of poor groups).

          Digging around more, here is a an article good for satire: http://www.adelaidenow.com.au/news/south-australia/naplan-results-show-south-australia-is-the-worst-performing-state-in-the-country/story-fni6uo1m-1227027376576

          SA Primary Principals’ Association president Pam Kent said teachers would be keen to see the end of the testing. “Teachers would be happy if there was less pressure and focus on NAPLAN and if that means no NAPLAN testing, I think they would prefer that,’’ she said.

          Personally, I’m not a fan of constant testing, but you would probably want to say this before the results of these things come out, and not after.

        • john Walker says:

          Thanks
          “SA gets worse at the bottom end (not in these figures ..”
          They do not publish all the results?
          BTW does SA proportionally have more remote indigenous students?

        • conrad says:

          They do publish all of them — I just don’t have them conveniently hanging around and can’t find any place with them. SA certainly has remote indigenous students, and many really are “remote” in SA, although they don’t have high numbers overall (http://www.abs.gov.au/ausstats/abs@.nsf/mf/4705.0). They also have rural populations in the middle of nowhere which you don’t get as much in places like Vic.

  6. Mayan says:

    Any reasonably bright person can code an app, but it takes something more to design and/or select an algorithm that can deal with serious data processing.

    I did an engineering degree, largely for the maths (if I had my time over, I’d have just done a maths degree because it floats my boat) and really haven’t found it overly useful because, no matter how much Australian employers, both private and public, say they want STEM talent, they don’t use it. Generally, Australian managers don’t want to deal with anything technical. They don’t get it and want nothing to do with it.

    As for maths in schools, it really is taught badly. It’s a wonder I like it. At university, we had texts that deal with subject matter as discrete blocks. However, there is a mid-level university maths text by Gilbert Strang which weaves together most of the second year applied maths syllabus into a coherent whole. He shows how it fits together and how it relates to the real world. If only such an approach were used in school syllabi,

    • MarkH says:

      I agree with Mayan’s comments.

      Nick, the basic gist of your argument seems to be about how we can improve the teaching of STEM skills to Australian students.

      But there is a more fundamental issue: there is little point in teaching STEM skills to Australian students because these skills are not valued very highly in Australia.

      To prepare younger people for a successful professional life in Australia – rather than teach STEM skills – schools would be better off teaching students how to become “managers” and “leaders”

      I am not sure why but Australia seems to have quite a (misplaced) obsession about the importance of “managers” and “leaders”.

      If we are going to talk about how best to teach STEM in Australia, I think we first need to raise the profile of STEM.

      • Nicholas Gruen says:

        Thanks Mark,

        I think you’re trying to get something off your chest. That’s fair enough but the column is about teaching STEM better, which seems to me worth doing whether Australia does more or less STEM in the future. I’m not much into grand objectives so just improving STEM is fine for me. It might be good if the society respected it more – and maybe teaching it better would be a start – but that’s not a precondition for improving the rather poor way we do something now.

  7. Ivo Vekemans says:

    I would love to see mathematics taught at schools. Sadly, what they teach is numeracy, a system designed systematically crush the love of structure inherent in the young.

    I could go on but this tells it better than I ever could:
    https://www.maa.org/external_archive/devlin/LockhartsLament.pdf

  8. conrad says:

    I did read that article — and whilst its all very well to imply that all children have a great innate interest in maths and that we’re killing it, the reality is:
    a) There is massive variation in both the ability and motivation of children in maths from the earliest age. Also, especially early on, lots of kids like quite repetitive stuff (e.g., counting backwards in steps, times-tables etc.), and these skills form the basis of later symbolic manipulation used in maths (a limited cognitive system humans start with that must be developed further). I agree though that more visual-spatial stuff like the examples should be taught (a separate approxamative system we start with and ignored a lot in current teaching).
    b) The author seems to think that if you think of interesting creative ways of teaching ideas, this will work. It won’t for all the lower end of the distribution. They will just sit there and be frustrated. If you’ve done higher level maths, then you will probably realize that at some stage, someone will talk about a topic for which you simply have no idea. This occurs at a different times for everyone. Sometimes quite early.
    c) Part of the reason what teachers teach is specified so well is that there is that there are logical paths of progression in terms of what children learn. I always use the example of graphics calculators. If these are introduced to early, children won’t develop visual-spatial skills. So teachers need to know more or less what to teach at certain ages. Also, do you want a teacher who got 58 on their TER doing whatever they want?
    d) Most parents want to know that their children have learned something, and I’m sorry to say that even for this reason alone, we would use testing.
    e) Thanks to the wonders of students-as-clients education, anything will be done to help the lower common denominator students. Since many kids (and adults) don’t like to think, you will get people teaching to this standard when they can, and if thinking causes problems, the teacher will have no alternative but to do this. In high schools, outcomes of this means “you got a good TER” and the teacher will remain on good terms with the parents. I can’t imagine, say, Carey, letting their lower students get poor marks even if it means repetitive boredom for others. In universities, lots of students don’t like to have open problems or think at all, and will complain on teaching evaluations if they are forced too “I got a poor mark because the assignment wasn’t well specified” (curiously, this doesn’t work in reverse). So guess what happens there.
    f) If you look at the nations where people are good at maths (and you look at, for example, whose citizens immigrate to work at NASA), they come from countries with lots of testing and where kids get lots of problems. Last time I checked, 25% of NASA’s employees were Asian, many coming from overseas or experiencing rub-off effects from their immigrant parents.

    So I think the complaints are simplistic, and better solutions cost money. For example, top echelon of kids should be given free run as the author talks about, but this would require special classes and other things that cost money most schools don’t have — most couldn’t even find the teachers if they wanted to, since they don’t even have qualified ones for the normal kids.

  9. paul frijters says:

    Nick, your romanticism does you credit. Whilst it is easy to say of such stuff that ‘it will never work’, I am all in favour of experiments with it on the off-chance that it might, even if only for a small group.

    On the wider point of greater appreciation for thinking in our society, I tend to point to incentives. With so much human capital walking in for free, and with a high reliance on things that cant run away (minerals), you should not be surprised when the incumbents invest a lot of time in the networks that help them appropriate this wealth, at the detriment of other investments. The massive levels of hidden unemployment in the education sector probably reflect those network investments.

  10. Nicholas Gruen says:

    Here’s a record of an email exchange I had with someone which furthers the discussion.

    I do have comments on your piece. I started to get caught up in the perceived shortfall in quality and quantity of STEM trained people when I was in the university sector. I was on an national enquiry which found the principal cause was lack of properly trained and motivated science teachers in schools so students did not get access to current knowledge and as the teachers often did not have science degrees themselves they lacked confidence and enthusiasm. Hence a vicious cycle with uninspiring teachers, fewer students study science, resulting in smaller pool of graduates for jobs, and teaching not competitive with other occupations. So I guess money could be useful here. The other problem was curricula, which has received recent attention though with what results I don’t know. Your examples illustrate the effects of precisely this.

    As far as demand or need for STEM qualified people goes, I think there is strong evidence we need quality and quantity for workforce and research requirements across a range of industries that Australia values and which contribute to the economy not only Information technology and computer science; what about biomedical science, neuroscience and pharmacology; biotechnology; agricultural sciences; resources and anergy; biodiversity, environment and water; and I am sure others.

    To which I replied:

    Thanks for your comments, I appreciate your taking the time.

    I agree with your points and 90% of my excuse is limitations of space. With less than 720 words one has to pick one’s targets. I realised when I filed it that read sceptically it suggested that coding and data science were the only STEM disciplines worth teaching, which is obviously not what I think. I like your list of skills we’re likely to need.

    If you also detected some disillusion with the system you’d also be correct. It pains me to see how much policy is driven by the basics of rent seeking. It’s not that I object to rent seeking. If a socially important activity won’t pay it’s way, then as I argued in this column, rent seeking is a dirty job but someone’s got to do it. The problem is that this gets in the way of systems transforming themselves. Whenever there’s a new need, the system will package it up as a new policy proposal and seek funding for it.

    Finally I can’t speak for all disciplines and my hunch is that things are a fair bit better in the STEM disciplines, but at least from the vantage point of my own understanding of my discipline, a vast amount of academic work is worse than useless – I’m using those words quite carefully, not ‘rhetorically’.

    When I think of the great economists of the ages – ending in Marshall and Keynes and their focus on specific economic problems in the world and the bizarre hall of mirrors contemporary academic economics has become, it’s pretty sad. As Paul Krugman laments, probably over half the high status academic macro-economists in the US operate within a framework which is essentially ridiculous and ridiculous for reasons that are easily explained.

    But to crown it all with a system that thinks of incentives for universities and their academics the way BF Skinner thought about lab rats in mazes, it’s pretty sad. It’s bad enough for the Americans to turn academia into a rat race for articles in prestigious journals, but for us to try to encourage Australian academic economists to do that is risible. It means Australian economists can’t be rewarded for studying Australian economic issues!! Quite a situation we’ve managed to get ourselves into. As Orwell once wrote – or wrote words to the effect of – Only people with years of training could do anything so stupid.

    Anyway, thanks for picking me up on an important point. To repeat, I agree with you!

  11. Nicholas Gruen says:

    Some corroboration of the idea that peer support can help with learning

    2. Peer Effects in Computer Assisted Learning: Evidence from a
    Randomized Experiment
    by Marcel Fafchamps, Di Mo – #23195 (DEV ED)

    Abstract:

    We conduct a large scale RCT to investigate peer effects in computer
    assisted learning (CAL). Identification of peer effects relies on
    three levels of randomization. It is already known that CAL improves
    math test scores in Chinese rural schools. We find that paired
    treatment improves the beneficial effects of treatment for poor
    performers when they are paired with high performers. We test
    whether CAL treatment reduces the dispersion in math scores relative
    to controls, and we find statistically significant evidence that it
    does. We also demonstrate that the beneficial effects of CAL could
    potentially be strengthened, both in terms of average effect and in
    terms of reduced dispersion, if weak students are systematically
    paired with strong students during treatment. To our knowledge, this
    is the first time that a school intervention has been identified in
    which peer effects unambiguously help weak students catch up with the
    rest of the class without imposing any learning cost on other
    students.

    http://papers.nber.org/papers/W23195?utm_campaign=ntw&utm_medium=email&utm_source=ntw

  12. Pingback: Memefication | Club Troppo

  13. Pingback: Overton Window – Overton Juggernaut: Part One | Club Troppo

Leave a Reply

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

Notify me of followup comments via e-mail. You can also subscribe without commenting.