« Project Link update | Main | The Parable of the Fruit Trees »


Feed You can follow this conversation by subscribing to the comment feed for this post.

When I learnt LaTeX the three primary advantages were that
1) the files are all text, not proprietary files with special characters that don't work across platforms - propriety software could get corrupted and then be totally useless or potentially lost if the proprietary software file specifications change and
2) it did exactly as you told it too rather than having the document being taken over by the software e.g. ms word - with ms word changing something you don't want it too and then not being able to get back to how you had it and
3) ms word being totally hopeless at formatting long document - the longer the doc the more likely 2) would occur.

The Fake Latex paper certainly misses some important considerations about why Latex is useful.

mpledger -

The latex/microsoft trade-off is likely different for word v. powerpoint. With a document, one can type a whole paragraph without having to add any latex code, but with beamer every single bullet point needs coding (unless I'm missing some awesome shortcuts). So the coding/content ratio is much higher for beamer than for a word document with a comparable amount of technical language. And of course if you've got a latex document with a gazillion equations in it, it's far faster to import those equations into beamer than to try to reproduce them or screenshot them in powerpoint.

I did actually prep a beamer version of this presentation as well. The big advantage of beamer is that the more rigid format forces one to be clear and concise.
The big disadvantage of beamer is the sheer amount of time it takes to type stuff out. Here's a sample of some of my beamer code:


\item{Based on simple quantitative metrics, such as
\item{Reputational surveys}
\item{Research: certain types of publications, citations}
\item{Student satisfaction surveys (Macleans)}


Because beamer makes it costly to add bullets, there's a tendency to make a small number of long points, and also to avoid nesting. Which can be good or bad.

It's also much easier to draw diagrams in powerpoint.

Interestingly, it really helps to understand beamer when FakeLatexing powerpoint.

\item doesn't need braces for its content, so you can go (preferably with nested indentation, but Typepad is eating the spacing):

\item Heading
\item Sub-item 1
\item Sub-item 2
\item etc

That said, I'm in a mathematical field and LaTeX's advantage with equations is by far its best feature for me. Even the coding becomes a bit of an advantage via creating shortcut commands for commonly-used operators. Its second-greatest advantage is its cross-referencing capabilities: I never got section or equation numbering to work well for me in Word.

I'll readily agree that drawing diagrams is easier in Powerpoint, although from my brief usage Keynote wins out over both Beamer and Powerpoint. I have taught myself enough TikZ to draw basic figures (needing to annotate an image supplied by someone else), and while it worked I would not call the process 'fast'.

@Prof. Woolley,

Do you think TeX and Beamer still function as a signal to the profession? Other than when one is on the job market (or submitting to Econometrica), I'm not sure it's worth it. Then again my own work is empirical rather than theoretical. If I wrote theory, I'd use TeX for the equations.

jack pq - I wouldn't have bothered learning beamer if I didn't think it functioned as a signal. (Tip for people thinking of experimenting with LaTex/Beamer - sharelatex is great- intuitive, good documentation, no fussy installations, lots of templates).

Although I didn't actually end up using beamer, learning it was a good use of my time, in that my faking-up-beamer-in-powerpoint skills are now vastly improved.

Given that the above is all about LaTex, I'd like to get in a plug for running old versions of WordPerfect, under Wine.

Back to the topic of the article --- let me extend your McDonald's analogy. Would reverse engineering the fries actually make them yummy? If you can identify "key factors", or some other management speak, does that mean that you've actually found a secret recipe, or just that you've missed the vital element, perhaps in the combination of ingredients or in something minor that you overlooked? In a competitive marketplace, it wouldn't matter: the customers would know the fries don't taste as good. In higher ed, it would: the university administration would keep advertising its supposed excellence, while robbing the students of the advantages of a liberal education.

Let me give an example closer to the university. If we decide (for instance) that the purpose of a degree in (say) Russian literature is to communicate with potential business partners, learn critical thinking, and signal intelligence and hard work, then we're likely to replace it with a Berlitz course in "business communication", some sort of critical thinking seminar (where students reinvent an impractical water-purifier, or try to deconstruct YouTube ads), and a lot of marketing around how wonderful the institution is. The end result, however, will be less than the sum of its parts. It won't teach Russian language very well (it'll be instrumentally restricted to business language), won't teach critical thinking very well (it won't have been inculcated by the really hard work of analyzing Dostoyevsky) and won't signal hard work and intelligence (not honestly, that is). Besides, the students won't get to experience great literature.

Sean - thanks for bringing the discussion back on track.

The McDonald's analogy, as you may know, is actually a true story. Ray Kroc bought the rights to franchise the McDonald brothers' hamburger stand, but when he tried to reverse engineer the fries, doing exactly what he saw the McDonalds brothers doing, he couldn't get the taste right. Turned out that the secret of those fries was that they were fried in beef tallow (refined beef fat). All good, until customers started dying of heart attacks!

But let's think about reverse engineeringt in the context of Russian literature, as you suggest. I agree that writing business memos in Russian is not going to give a student the same appreciation of Russian culture and language, or the same skills, as analyzing Dostoyevsky. But it's easy to fall into the line of thinking that goes as follows:

Some students learn a great deal from some elements of the present curriculum
Some students learn a great deal from the present curriculum
Students learn a great deal from the present curriculum
Therefore change is bad/unnecessary

When I taught third year public finance, half of the students wouldn't don't show up to any given class. 75% never or rarely came to office hours. A good chunk didn't do the readings. Either the course content, or the course delivery, or both, was failing to engage and interest the students. Some of those students have gone on to do great things. But I really doubt that those who put in the bare minimum required to scrape through the course got a great deal out of it.

So the question becomes: how to make things better? Do we employ a team of pedagogical experts to come up with ways of getting students really engaged in the concepts of equivalent variation and compensating variation? Would pedagogical experts be able to figure out how to use clickers and experiential learning to make students excited about Pareto optimality and social welfare functions? Perhaps they could, but I'm pretty sceptical about some of what's taught in faculties of education - I'm not sure how many of the pedagogical methods advocated are backed up by,, say, large scale randomized controlled trials (though probably some of it is). And it's not obvious that it's possible to figure out how to teach something without truly, deeply, understanding the underlying concepts. Plus I figure that, at least in the case of economics, a great many students have little interest in learning much of what is taught - they basically want skills that will help them get a job.

So I share your skepticism about what happens when university administrations try to make things better. But I guess I have a less rosy view of the current status quo - perhaps because I'm not teaching or studying Russian literature.

Thanks. I'd actually heard the McDonald's story before, somewhere, and mean to get some beef tallow to try it out one of these days.

I'd suggest that part of the reason that few students came to your class was that they'd been inculcated in a transactional view of education. Everyone from parents to guidance counsellors to pundits have always talked to them about education in terms of career qualifications, so they do the minimum to get the credits, which they're already paying for, anyway. (Why pay twice, with both effort and cash?) As a result, though, they don't get much out of the course.

It's a cliché to say that you get out of a course what you put in, but it becomes a paradox once we start thinking of education in terms of outputs: if you're only interested in what (say) job-skills you'll get out of a course, you probably won't get any.

I share your pain, BTW, but more for my first-year (required) English courses.

Perhaps students should receive some type of collateral, or pay into an insurance scheme for when they graduate - and don't get a job.

Teaching people to read books is more important than income rising as rich people in Canada live in large low quality-of-life cities. Most important is to potentially activate those precuneus epigenetics that use technology for good, such as neuroimaging existing military nuke war training for stopping AI and neuroimaging those who can deal with other technological risks.
Money has made universities research AI and robotics way to much, at least for 3-10 more years. I'm hoping after a low-moderate attack, universities don't fund such and fund entangled microwaves to find some labs, and fund a line of a million entangled oxygen atoms/molecules in a waveguide to optically detect slight decoherence when a geoneutrino goes through them all: they will shift a bit assuming flavour changing or some other lab-imagible feature affects the waveguide laser. Gets all labs. Structural health monitoring using entangled photons in a waveguide or layers of quantum dots doesn't work well with artillery vibrations but can detect some hacking. Putting a wire metalens in ice gives you a radio arrow to a fist of nuclear material or scaled to a spaceship running away.
Dion though about truth detection AI in 2003 and Obama for terrorist propaganda in 2007 and both rejected it because it was too hard (would've learned pattern recognition); that's why I have the 3 best PMs from my hometown (ahead of Garneau) as working to speak is useless compared to a NORAD curricula improved. Money is so easy too worry about until a McLuhan experience of a Tet offensive shocks people back into a Cold War mindset: but the neutrino detector is 20 years away without a neutri-hattan Project and we don't have the time: I don't have a GAI and Bostrom didn't focus in 2005 to AI or robots.

The comments to this entry are closed.

Search this site

  • Google

Blog powered by Typepad