Monday, November 28, 2016

UWA Marks Policy

Now universities are from time to time accused of "grade inflation".  That is, they give higher marks than they used to.  Of course in the long term this can't really be true, because by now all students would have to be getting 100% all the time.  But I'm making light of a real problem.  How do we ensure that a mark of 80% is an impressive mark?

UWA, bless its cotton socks, has an idea to fix this.  At some point in the near future no unit will be allowed to give more than 50% of their students a mark of 70% or more.

Superficially, this is attractive.  But as is so often the case, the devil is in the detail.  The real world doesn't cooperate.  I wish it would, but sadly, no.  Not only won't the real world cooperate, but it fails to comply in multiple ways.

Firstly, lets look at a case where the rule will work.  You have a course that students enrol in and when they come out they are a professional widger, and can go out and get employment with that qualification.  Every year 300 new widging students start, and every year about 250 of them graduate.  You lose some because they change their minds, or they can't hack the work.  But most battle bravely through.  So you essentially have the same cohort of students throughout, and it doesn't really make sense if they get really high marks for "Widging in the field", and really low marks for "Rural widging".  If that happens, you'd think that the lecturer of "Widging in the field" maybe wasn't very confident in their lecturing, and so set a much too easy exam in the hope of hiding their incompetence (only to discover what the education people have told us a lot, the quality of the lecturer doesn't matter as much as you think it does).

Now lets look at a case where it won't work.  Take my favourite group of students, physics.  Take the 3rd years.  Look at their marks in PHYS3011.  You'll find more than 50% of them getting marks over 70%.  Should they?  Well, one thing you can do is look at how they went back in PHYS1001, a big first year physics unit.  They were just a small part of that unit.  And that unit had well under 50% of its students scoring over 70%.  But how did our PHYS3011 students go in PHYS1001?  Well over 50% of them got scores over 70%.  Unlike the widgers who kept much the same cohort throughout, the majority of students who did PHYS1001 don't do PHYS1002, and only around a third of those who do PHYS1002 end up doing PHYS3011.  Because physics is hard, and weaker students drop out, leaving just the best at the end.

And its not just PHYS1001 where our PHYS3011 cohort did very well.  Look at just about any non-physics unit they do, and you'll find that they break the "less than 50% over 70%" assessment rule.  So our PHYS3011 students are breaking the assessment rule all over the place, but the only time it actually shows up is in PHYS3011, where they are all together and undiluted by other students.

So that is one way the assessment rule breaks down.  Here's another.  Consider Advanced Music Ensemble, a 3rd year music unit (sorry name may not be quite right).  These students also get too many marks over 70%.   But if you look at their performance in non-music units they take, you won't see exceptional performance.  They probably won't break the assessment rules in anything other than units involving musical performance.  These units are very different to others at UWA.  Almost all units involve learning a lot of material and then demonstrating your understanding of and ability to apply that material.  But virtually no units are concerned with excellence of performance and technique.  No one watches the maths student produce a proof and marvels at their virtuosity.  Maybe they lose a few marks for an ugly proof, but as long as its a valid proof, they are fine.

No, with Advanced Music Ensemble, there is nothing for it but to leave the marking to the experts who know what they are looking for, and recognise it when they see it.

But lets make a point about this.  If the School of Music, or the School of Physics send out students with marks of 80 and 90%, and those students audition for orchestras, or enrol in PhD's at other universities - and if they embarrass themselves, then that would tarnish UWA's reputation.  And we don't want that.

So you see that there are competing forces at work.  Yes, the School of Music might want to have 200 students doing 3rd year Advanced Music Ensemble, and it can probably attract that many by marking really easily.  But they have the countervailing force of their reputation.  And these force balance out and ensure that the marks given out actually mean something.

But there are units at UWA which have very little incentive to keep marks down.  Broadening units that are not part of a major.  And if you look at these units their marks have gone up in the years since they were introduced.  Its hardly surprising.  You get paid for the number of students you teach.  The students want a broadening unit which requires little work and is likely to give you a high mark.  Students and academics interests align, and you get easy units with high marks.  I'm being a tiny bit cynical here.  Some of these units are really good.  And some of the students who take them do so because they are genuinely enchanted by the subject.  But none of that stops grade inflation.  And for these units, the assessment policy is likely a good thing.  I still think more effort is required by the university to ensure that these units run well, but fixing the assessment is a good move.

Lastly, will the assessment policy have any bad effects?  Well, yes.  If you want a mark of over 70% (a grade of "D" or "HD"), then you are not attempting to exceed a standard that could be defined for the unit before you start it, you are competing against your fellow students.  If you help a weaker student improve their mark, your mark could go down as a consequence.  This is ludicrous.  We could even reach the ridiculous situation of encouraging weak students to continue because we know that more weak students will allow us to reward our more able students with higher marks.  Now of course this is not the sort of thing we'd do - but in running a unit we are faced with a lot of seemingly minor decisions, and if we have in the back of our mind that more weak students helps us reward better students...

In summary, this policy looks just like the UWA restructure.  The university has identified a problem, and produced a simplistic solution that ignores the unavoidable complexities.  And in doing so they will create a new set of problems, which at least in the short term will be worse than those they sought to fix.  Like the UWA bosses, I wish for a simpler world.  Unlike them I don't pretend to be living in one.


Friday, November 18, 2016

What makes good teaching?

A fellow worker in the School of Physics at The University of Western Australia (UWA) got a teaching award, as he does most every year, because he’s a fantastic teacher.  So the Dean of our new faculty (Engineering et al) asked him to come in for a chat.  And he asked him what made for good teaching.  Now I can teach, but I’m not that great at it.  But I can tell you what makes good teaching.  

The key things are values and incentives.  If you have those right, the rest is just details.  What follows is based solely on my experience in Physics. 

The main value is that we love the kids we teach. They are not just students, they are partners in the teaching and learning process.  Many of us are parents.  We know how we want our own kids to be treated, and we should not do any less with other people’s children.  We love them because they are interested in physics.  We owe it to them to provide a physics education that will allow them to join the worldwide community of physicists.  We owe it to them to fail them if they don’t master a unit.  Allowing them to waste their time would be cruel.  We owe it to them to catch cheats.  Allowing cheating devalues their degrees.

While we can all individually have such values, there is great benefit to the Head of School having these values.  Many decisions in physics teaching are easy to make.  We know the Heads values, and we apply them.  So less work for him, and more responsibility for us.  Its a win - win situation.  I worry a little about the university as a whole.  Is aiming to be in the top 50 universities in the world a value driven goal?  How does that goal translate into the decisions that everyone who works at the university makes on a day to day basis?  

But good values alone are not enough to ensure good teaching.  We are regularly faced with decisions on how things are done.  And being human, we sometimes go for the easy option even if its not “the right thing”.  However if your choices have consequences, you might think harder.  In Physics, we get paid for each student, and we like having lots of students.  If we teach badly, we will lose students, and money.  There is our incentive.  Many of our students are very talented, and they could major in maths, engineering, computer science etc.  To keep as many as we can makes us teach well.  It makes us choose our teachers well.  What happens if the incentives aren’t there, or the incentives are perverse?  In Physics, we are battling with one such situation.  

By the time our students are in the 3rd year of their degree, they are unlikely to change majors.  So one incentive to be excellent teachers is gone.  Worse still, by trying to “be nice” to the students in the earlier years, we’ve left a lot of content for third year, and we’ve left some of our not so excellent teachers for third year.  So traditionally 2nd semester in 3rd year has been a nightmare for the students.  This was recognised a couple of years ago (and may have led to us having fewer Honours students than we expected), and we’ve put a gun teacher in charge of 3rd year.  Its a lot better this year, but there is still work to do - its no easy thing to turn an oil tanker round.  So if you want excellent teaching, and your students are “captive”, then you will have your work cut out for you.  Its pretty easy to indulge in fuzzy thinking when there is no cold hard loss of money attached to bad decisions.

Another example of perverse incentives relates to specialist broadening units.  These are units that are not part of any major, and it is entirely up to the student which of these they take.  Now from the student’s point of view, they want an easy unit for which they’ll get a high score.  This frees them up to focus on their major.  From a School’s point of view, they want lots of students so they get lots of money.  So to attract large numbers of students, the school has an incentive to make the unit easy and to mark easy.  And if you look at these units in the time since they were introduced, you will see grade inflation.  That is, the Schools have responded rationally to the incentive.  Why do we not see the same in other units?  Most other units are part of a major.  The school running them wants to be sure that students who pass that unit are capable of proceeding to higher units.  The school wants to turn out quality graduates, as this reflects on the school.  So they have an incentive for a mark of 80 to actually mean that the student is pretty good.  But this incentive is not present for broadening units.

Given that Physics has an incentive to deliver quality teaching, it is not surprising that we do it by allocating our teaching resources in a way we think is optimal.  Its like a hockey team.  You put your best players at inside left and inside right.  You probably put your worst players at full back (my old position).  So Physics knows there are some people who you shouldn’t unleash on students.  Or at the very least, you don’t let them loose on 1st and 2nd years.  We try not to put our weaker teachers in the early weeks of semester - we still want students there on the all important census date!


So that is values and incentives.  After that you still have to get the best out of your teaching staff.  If you overload them, then they will forget about values and incentives and just struggle to survive - and clearly the students will suffer.  So your values dictate that you can’t overload them.  You also have to get your support staff to reduce the load on your teaching staff.  High quality teachers should not be wasting their time on drudgery (unless they like doing it as a form of meditation).  In Physics we have shared out unit coordination duties in a way that works for staff and students.  Moving to a different way of doing this might reduce teaching quality.  I say this as someone with a vested interest - my unique role in Physics looks like it will vanish next year as the university restructures to a more "public service" model.

Friday, November 11, 2016

With a day to think about it.

Democracy sort of worked in the US.  The voters delivered a message to the establishment that they aren't happy with the status quo.  There are lots of horrible things about who they chose, but that was the choice they were faced with.

Trump will fail.  He's got some good ideas, but he's hamstrung by his low tax policy.  That means he's going to really disappoint a lot of people who didn't think things could get worse.

30 years ago, we were better off than we are now.  Not richer, but definitely better off.  Come up with a plan to fix that up, and I'll vote for you.  But not if you're a cunt like Trump.

Actually, scrap that.  I've got that plan.  So vote for the Newstart Party.  Follow me.  We may not fix it, but the smart dudes will be so scared of us that they will fix it.  All you need to do is follow me.

Wednesday, November 9, 2016

Fuck. He won.

I live in a world that I don't understand any more.  It is changing so rapidly.

Donald Trump just won the US presidency.

And he won by appealing to white people.  They are the biggest single group in the US, and Trump made it very clear he would represent their interests.  The blacks and latinos knew what this meant for them, and they voted against him, but there are not enough of them, and the white vote won.

Hillary made the stupid mistake of being inclusive - the sort of president who would govern for everyone.  I guess white people looked at that and couldn't see why they had to share.  Poor parenting.

In Australia, we had an election earlier this year.  There is an organisation called "GetUp!" who campaign on the progressive side of politics, my side of politics, and I gave them money, and I worked on a polling booth for them.  It kind of worked, as the conservatives won, but only just.  And now I realise, that if I want progressive politics, I have to pay.  And that is fine, I'll chip in to balance the millions donated by big business.

But the same is true of newspapers too.  If I want to read quality online, ultimately I'm going to have to subscribe to The Guardian, or The New York Times, or The Sydney Morning Herald, or The Melbourne Age, or New Matilda, or Crikey.  Because the Murdoch media is on the side of evil.  Because our ABC has been starved by government and cowed.

So now I'm living in a world I don't like.  A world where the relentless drive of the rich is wearing us down.  A world that is heading back in time to the 1930's, but without the shining light of an FDR.  We are heading back to the era of robber barons (or oligarchs), to an era of patronage, to an era whose values led to the Russian revolution.  To an era of jingoistic nationalism.  To an era of increasing inequality.

And the American people, they know good, but they said, "Stuff that", and voted for Trump.


Sunday, October 30, 2016

Science and the future

On todays ride, Mark and Paul were talking material science.  The discovery of some new alloys that were amazing.  And I had a sudden revelation about the future.  You see, I'd recently been thinking of a somewhat dystopian future, where we only bother training a million or so people to be the scientists, engineers, and designers for the whole world.  Because the team at Apple that comes up with the new iPhone, with all its design and technological magic, isn't huge, but the product and its clones sell worldwide.  No one who isn't an absolutely amazing designer/engineer/programmer/scientist will work on it.

And what has happened with the phone will spread like wildfire.  How long befor whole houses are designed by the worlds best?  You'll just tell them the block of land you've got, and they will know the local council planning rules, the local building regulations, and they'll know the houses you've browsed online, and they'll know your budget, and they'll come up with designs.  Largely automated designs.  And if you like one, their automated factory will produce your house, and it will be shipped to you and assembled - far more cheaply than a conventional builder or architect could do it.

And so I thought that there was no longer any need for your common or garden engineer/material scientist/programmer etc etc.  Indeed, no longer any need for many tradesmen, as modular design removes the need for measuring and cutting - the house will be square, the dimensions will be known, and the module you buy will just slot in nicely.

So the dystopian world where there are a small number of experts, and most everyone else has "jobs" that involve being nice to people.  You know - the smiling waitress, the chatty hairdresser, the inspiring personal trainer, the attentive therapist.  And, to go with this, the great many people who won't have jobs at all.  Right now, at least in Australia, these people are demonised, and every attempt is made to blame them for their own plight and pay them as little as you possibly can.

With this in mind, I questioned university education.  Why are we bothering to educate anyone except the very, very best students in physics, maths and engineering?  In fact, why aren't we identifying these students and sending them straight to Caltech?

But that is where todays revelation comes in.  You see, to come up with a fantastic new alloy, you don't have to be a genius.  Or even very good.  You have to be persistent, and clever enough to realise when you've found something interesting.  You will need to understand your field very well, but there are many, many fields.  And each one is quite narrow, so while it is hard to understand your field well, many, many people can do this.

And our world relies on technology.  Someone designed the glass that the iPhone uses.  People are working flat out designing better photoelectric cells.  People are exploring the intricacies of how light can manipulate atoms, and vice versa.  There are literally millions of things in this world that need to be explored, because they might be the next LED light.  The humble LED light that now saves millions of tonnes of CO2 from going into the atmosphere.

So my less dystopian future goes like this.  Train lots and lots of physicists, chemists, engineers, programmers, materials scientists, biologists, mathematicians.  And (shock horror) through higher taxes employ them.  Employ them on promising ideas, and on impossible ideas.  Employ them on the problems of the day, and on arcane research that will almost certainly never bear fruit.  And sure, right at the top in the Apples and Googles of the world, you'll have the best and brightest pulling together all of the discoveries made into the wonderful products of the future.

Or we could continue down the dystopian path.  The low tax path where there really are only meaningful jobs for the best and the brightest.  Where we rely on Apple and Google to decide what ideas are worth pursuing  Where the skills and abilities of a great many are not used.  And most of us fight for the crumbs from the oligarchs tables.   And the oligarchs?  For every Elon Musk, there will be 10 who just want the biggest yacht and the prettiest mistress.

Let me say that again.  In a low tax environment, the skills and abilities of a great many will not be used.  And we will all be poorer for that.  Genuinely poorer, because great discoveries will not be made.  And it is happening now.  In Australia we have CSIRO, our nations flagship scientific research organisation.  Its funding is being steadily cut, because we refuse to tax enough to pay for it.  An astronomer in CSIRO developed Wifi.  How many future inventions are not being found because of funding cuts?  How many future inventions are not being found because the need for an immediate return on investment is destroying research motivated by pure curiosity?

And I'll go one step further.  We are also wasting the talents of our best writers and artists.  The BBC and other national broadcasters are squeezed.  Squeezed so that the idea of something as original and amazing as Monty Python appearing now is laughable.  And again, we are all the poorer for that.  Not just the lucky ones who get to make amazing shows or works of art, but all of us who never get to revel in them.

The low tax, small government world is a step backwards.  If you wish to sum it up succinctly, you may say it is a world where, in order to placate the very rich, we take away opportunities from many, and impoverish the world as a whole.

Saturday, October 29, 2016

And Warren Mundine is a fool.

Warren Mundine appeared on Q&A, along with some guy who pushes equality.  Maybe it was Sir Michael Marmot, I can't remember now.  And the bloke who pushes equality argued that rising inequality was bad, and that we needed to do something about it.  I think he may have even suggested raising taxes on the rich.

And Warren Mundine laughed at him, as though he was a total moron.  A fantasist.

Which makes Warren Mundine a fool.

TV shows need to be consistent

OK, this is definitely a 1st world problem.  I like English murder mysteries.  Agatha Christie,  Father Brown, Midsomer Murders, Death in Paradise, Hercule Poirot, etc etc.

I particularly like these ones because they don't pretend to be real.  They are fantasy worlds, and the people who inhabit them aren't real.  They are too rich, beautiful, angry, unreasonable, nasty, good, kind, vindictive, to be real.  Just to see Miss Marple having tea with an old friend transports me to a half dreamed world that I've never known.

But there is another sort of British murder mystery.  The one where the description contains the word "gritty".  One understands that these are more harrowing and realistic in their portrayal of the people and events.  Except, and this is the bit that grates on me, they are not consistently realistic.  So you'll have a former policewoman returning to the force, and resuming their career as though nothing happened.  No interviews, no filling in forms with HR, no, they just suddenly start working again.  And then, when chasing a serial killer, they look through a window and see someone tied up in the way that the serial killer does.  Common sense, let alone police training, says that before you go in, you make a quick call to HQ, give them the address, and ask that they send backup as a matter of urgency.  But no, some lazy writer decides that our cop will just waltz in there and (not at all surprisingly) get whacked over the head by the serial killer lying in wait.  So shows like this are gritty and realistic when it suits them, but drop all pretensions of reality when they can't be stuffed working out a good plot.

You see, I can forgive all that in the fantasy shows.  Because the whole thing is fake.  But in a show pretending to be real, I expect them to go to a lot of effort to make it compelling.  Like Breaking Bad.  When watching Breaking Bad, a couple of times I thought I'd spotted a plot flaw.  But each time, it was put there deliberately, and it had logical consequences later on.  You thought, "But other people are going to realise blah blah blah", and later on in the show other people did realise, and it played out beautifully.

So, British gritty crime dramas, how about you don't just rub our noses in poverty and nastiness and intrigue, but actually put a decent effort into writing truly great plots.  Or at the very least, believable ones.