The Third Bit

Gary Stager isn’t the first person to point out that we’ve been dumbing down computing education for the last 30 years—that we’ve gone from teaching kids how to program to teaching them how to use Excel to teaching them how to use iPads.  (My five-year-old didn’t need to be taught…)  What people mostly aren’t asking is why this has happened, but I have a theory.  I think teachers are teaching the computing that they use themselves, because that’s the most economical thing for them to do: they use Word to make hand-outs, and Excel to manage grades, so they’ve already invested in proficiency, so they can put together lessons in less time.  They don’t use Logo (or Scratch, or Python, or whatever) in their own work, so teaching it requires more effort.

Which makes me wonder how successful current “programming for everyone” efforts are going to be. I doubt most teachers are going to want to hack their classes’ web sites (and even if they do, their schools may not allow them to).  I’m even more skeptical of the idea that teachers will program in their day-to-day work any time soon.

(“But wait,” you say. “Most Latin teachers don’t speak Latin at home. And there aren’t a lot of physics teachers building lasers in their dining rooms.” True, but those are recognized, accepted specialties within teaching right now: they aren’t trying to gain ground, just hold onto what they have.)

Now let’s shift focus a bit. My 14-year-old nephew is on the web, creating content, almost daily, but “on the web” isn’t quite accurate. He’s actually on Facebook: along with GMail, YouTube, and Minecraft, that’s pretty much what the web consists of for him (at least that he’ll admit to his uncle). We already know that if we want to teach kids, going to where they are works better than bringing them into an artificial environment and giving them artificial tasks. I therefore think that if we want the 95% who aren’t already keen on hacking to care enough to do it, we need to teach them how to hack the places they already are. Having everyone build their own Facebook plugin would (a) take far too long to pay off and (b) be unsafe (imagine you’re Facebook’s director of security and someone tells you that 100,000 high school kids are about to build plugins for your site).  But what if we made a generic plugin that allowed people to build and run WebScratch programs inside Facebook, with access to (some) FB content?  I know, I know, there’s no such thing (yet) as WebScratch, but you get the idea: if we could create, test, and deploy this that makes it easy for them to build that, what would “this” and “that” be?

(“But wait”, you say, “Facebook is a closed pseudo-monopoly. We shouldn’t be supporting them!” I agree, but (a) going into a slum in Rio de Janeiro to teach the kids who live there doesn’t mean you support the idea of slums, and (b) they won’t understand why “closed” is bad until they understand more about the web in general, and open in particular, and they won’t learn either unless they care enough to learn in the first place.)

So, question #1 is, “What can we teach about computing and/or the web that will appeal to teenagers, but will also be useful to classroom teachers?”  And question #2 is, “What plugins can we build that will allow learners to hack where they are?”

Teaching

For every complex problem, there is a solution that is simple, neat, and wrong.
— H.L. Mencken

Over at The Atlantic, Philip Howard is trying to convince us that America’s education problem has a simple cause and a simple solution:

America’s schools are being crushed under decades of legislative and union mandates. They can never succeed until we cast off the bureaucracy and unleash individual inspiration and willpower… Successful teaching and good school cultures don’t have a formula, but they have a necessary condition: teachers and principals must feel free to act on their best instincts.

Mm. Jim Keegstra was acting on his best instincts—does Mr. Howard think he should have been free to continue doing so? Or what about teachers who ignore bullying: should they be free of bureaucratic “interference”? How about teachers who are themselves the victims of bullying or worse by school administrators: shouldn’t they have somewhere to turn? Yes, a lot of bureaucracy is pure empire-building, but a lot of it has been created after someone was badly hurt in order to prevent a recurrence.

I complained a lot about the pettifogging backside-covering red tape at U of T when I was there—in fact, it’s one of the main reasons I left.  I used to complain about having to wear safety goggles in the chem lab, too, until I met a guy who lost an eye when a beaker blew up. I agree that 90% of the time, 90% of the regulations are unnecessary. The problem is, we don’t know in advance which 90%. If Mr. Howard can convince me that he does, and that clearing away all that bureaucracy won’t make life easier for slackers, predators, bigots, and bullies, I will personally operate the shredder on his behalf.

Teaching

I have a lot of respect for Heather Payne (founder of Ladies Learning Code): she’s working hard to get women into technology, and has been more creative about doing it in 12 months than I’ve been in 14 years. I’m therefore reluctant to disagree with her publicly, but her most recent post said this:

I find it pretty interesting to note that most of our Lead Instructors and Mentors are in a teaching role for the first time ever when they join us at a Ladies Learning Code workshop. And no one on the Ladies Learning Code team has a background in education. The funny thing about that is that it might be why what we’re doing works.

Respectfully, no. What LLC is doing works as well as it does because:

1. It has attracted a really talented bunch of inventive, open-minded, hard-working people.
2. It’s one day or one week, not four years.
3. Everyone in the room really wants to be there.

The first explains itself, so let’s look at the other two. A night out or a weekend away is very different from being married; similarly, one jam-packed day of new, new, new is a whole ‘nother thing than a four-year university program that’s trying to cover all the bases, including the ones learners don’t think need to be covered. (They’re often right, but not as often as they think.) The kind of energy and engagement I see in workshops (of all kinds, for all ages and backgrounds) simply doesn’t last for months or years.

Second, I believe that about a quarter of the students in any university program would rather be elsewhere (if not more). They’re in class because their parents told them they had to go to university, or because they’re worried about being unemployable if they don’t have a degree, or because whatever they’re doing seemed exciting two years ago when they picked their major, and they think that if they switch tracks now, those years will have been wasted. That, combined with peer pressure in high school teaching most teenagers never to appear keen about anything a grown-up is telling them (unless said grown-up is disreputable), sets the tone for the student side of most university classrooms.

The faculty side—well, as Eleni Stroulia said, we’re here to do research, they pay us to teach, we spend our time on administration. Most grade school teachers, and many university professors, are passionate about teaching—just look at the 60-hour weeks they put in. But many others have burned out, have discovered that they don’t like it as much as they thought they would, or (at university) have always regarded teaching as a tax they have to pay in order to do the research they love. And while I can’t speak for K-12 from first-hand experience, promotion in most universities really depends primarily on your research, no matter what’s in your employment contract about “teaching” and “service”.

So, do things like LLC work as well as they do (and they do work very well) because the people leading don’t have a background in education? I don’t think so. Would they work better if their organizers and instructors could ace the Audrey Test? I don’t know; I don’t think it would hurt, but I wonder if what formally-trained educators are taught is more relevant to the cross-country trekking their learners are expected to do than to workshop participants’ sprinting. What I do know is that we’d all benefit from more thoughtful analysis of what works and what doesn’t at various scales and for various learners.

Teaching

Once upon a time, about a quarter of a century ago, I went into a prof’s office in Edinburgh and told him that neural networks were the future, because, look, that’s how our brains work, right? Neurons and connections and firing potentials—those are the building blocks of human intelligence, so all we have to do to make AI work is, you know, do that in software. And if it isn’t working yet, well, that just means we need a bigger neural network, or maybe different firing rules. I was in my mid-twenties at the time—about the age that many of today’s ed-tech entrepreneurs are now—and when he tried to explain to me why things might not be that simple, I quickly filed him under “old fogey”. I didn’t bother to read any of the papers he pointed me at: what could those patient, painstaking, footnoted summaries of old news possibly have to say about the brave new world my friends and I were about to create?

Well, you all know how that turned out…

But now it’s 2012, and Audrey Watters’ post “What Should Every Techie Know About Education?” (a.k.a. “the Audrey Test”) is generating some interesting comments. Some are thought-provoking, like Ben Huffman’s that, “Every school superintendent, college president and graduate student in education would be able to easily pass the Audrey Test,and yet, the system is still broken.” Others, like SM’s, do a great job of show why techies need to do their homework before trying to “fix” education. (See BF’s reply for a dissection.)

The Audrey Test has also led to some interesting questions. The two I was asked recently are:

1. How much does one need to know before one solves a problem?
2. How much of the status quo is deeply embedded within the theories of learning and academia?

My answer to the first is, “A lot more than I did two years ago (or even three months ago).” For example, watch this interview with John Sweller. I don’t agree with everything he says, but my disagreement is based on gut instinct, i.e., on a poorly-examined mental model based on (probably inaccurate) recollections of personal experience. He knows more about how learning actually works than I do; that doesn’t mean he’s right, but if I don’t take some time to learn what he and the constructivists he’s criticizing are arguing about, the odds are that I’ll repeat mistakes they made and moved past twenty years ago. Yes, there’s a tradeoff between learning the past and building the future—there are only so many hours in a life—but if I could send email back in time to May 2010, I’d tell myself to spend an hour a day, every day, learning from others

The second question is harder for me to answer. On the one hand, many of the people I’ve found since I started doing my homework are much more deeply opposed to the status quo than I am. (In fact, I didn’t realize how different from the status quo it was possible to be until Audrey’s article pointed me at Paolo Freire, which kind of reinforces my answer to question #1.) On the other hand, every big institution’s prime directive is to perpetuate itself, and one strategy that academia uses well is to keep people so busy talkin’ radical to one another in safe little sandboxes called “departments” and “journals” that they never actually have any effect on the real world. Which brings us back to the Ben Huffman comment quoted earlier…

I don’t know what the answer to #2 is. But here’s what I currently think:

1. Technology has changed every aspect of our lives (see Cat Valente’s thought-provoking “Life With and Without Animated Ducks: The Future Is Gender Distributed” for commentary on this in a related context).
2. But that doesn’t mean that technology changes everything, every time.
3. So if someone dismisses previous work in education without knowing what it was, they’re probably doing so to give themselves an excuse not to do their homework.
4. Which means they’re probably going to frame questions naively, reinvent a lot of wheels, and waste a lot of time. That will undoubtedly keep them busy, but that’s definitely not the same thing as being productive.

To boil it down: if someone came to you and said, “I’ve never looked at the source code for Django, Rails, Struts, Zend, or any other web application framework, but I don’t need to, because this new language I’m using is so different from everything that’s come before that they’re all obsolete,” you’d admit that there was a chance they were right, but would you bet on it? Particularly if they’d never talked to anyone with, you know, actual real-world web application programming experience? I don’t think it would matter how smart or hard-working or charistmatic or well-funded they were—I think you’d put your money elsewhere. So why is education different?

Later: today’s post by Mark Guzdial is another example of why people like me need to know more about stuff like this. I respect Mark tremendously, and Hal Abelson’s record is as good as anyone’s; if they disagree, I think I have an obligation to do a little learnin’ of my own before venturing an opinion. See also “Scottish Verdict“.

Teaching

As I wrote a few days ago, I asked Audrey Watters to put together a “Joel test” to assess how much someone technical knew about education. She has now posted her response; if I were grading it, I’d say, “Exceeds expectations,” because what she has done is explain why a simple 12-point yes-or-no test wouldn’t be enough:

What you do with your code or your content or your users’ data—while incredibly important—is only be part of what I “test.” … If I were to really formalize such [test], I think it would also have to involve what you know, what you think about education, about teaching, about learning, about politics and theory and practice—its history, its present, its future.

A few questions are US-centric, but I’m sure it would be easy to swap in Canadian (or British, or Estonian) acronyms for FERPA and the like. What they’re all triangulating, I think, comes down to one crucial question:

Do you know what’s been done before, or are you a (willing, and possibly even proud) victim of ed-tech amnesia?

I can answer about two thirds of Audrey’s questions; I think that missing third is another reason why my P2PU course on teaching free-range learners fizzled. On the upside, her questions present me with a wonderful learning opportunity

So how about you? How much do you know about where we are, how we got here, what’s worked in the past, and what mistakes we don’t need to make again?

Later: one of the first comments on Audrey’s post said, “All the other questions are in my opinion irrelevant except the last two. No one but the ‘experts’ care about all that.” (By “other”, the poster meant everything except “what is the purpose of education?” and “what is the purpose of educational technology?”.) When someone says that, what I hear is, “I’m going to choose various options without realizing (or without acknowledging) that there were alternatives.” That’s the best way I can think of to perpetuate the status quo.

Still later: Dan Hickey’s post “Some Things About Assessment Badge Developers Might Find Helpful” (and the longer version here) are useful reading as well. I have a lot of catching up to do…

Teaching

The final meeting of my P2PU course on teaching free-range learners how to program took place this morning. I enjoyed chatting with the people who showed up, but overall I was disappointed with how things went: less than a quarter of those who signed up back in January were still with us at the halfway mark, and less than 10% were in at the end. Partly, I think, this happened because the demands of day-to-day life trump good intentions every time. I also blame myself: sending out a couple of chapters of reading material at the start of the course sent a clear signal that this was going to be top-down and centralized, rather than bottom-up and peer-to-peer, and that turned a lot of people off.

That mistake is a symptom of something more fundamental: I have never actually completed a course online myself [1]. Think for a moment about what that means. I have probably spent more than twenty thousand hours in classes of various kinds. I wasn’t just learning algebra and history during that time: I was also learning what made a class good or bad. How did the teacher handle interruptions and disruptions? How big were the assignments, and how easy were they to understand? Was there some overall arc to the course, or was it just a disconnected jumble of facts? Consciously or otherwise, the first time it was my turn to teach, I drew on that experience.

But online is different, and a lot of what I soaked up in those hours and years doesn’t apply (or worse, seems to, but is wrong). I’m pretty fluent with email, Skype, chat, bulletin boards, wikis, and so on, but using those things to socialize or to coordinate a software development project isn’t any more relevant to online teaching than knowing the difference between keeping a bunch of undergrads engaged and just blocking their view of the whiteboard.

As many people have pointed out, the greatest and most common weakness in Silicon Valley’s approach to education is a thundering lack of relevant experience [2]. I’d therefore like to propose that the following question be #1 on the Audrey Test:

If my answer is January had been “more than zero”, the people who signed up for my P2PU course would probably have gotten a lot more good out of it. And if the techies who want to revolutionize education by internetting it could say “more than zero”, I think society would get more good out of them.

[1] I did the first six weeks of an online course about online courses through Ryerson University back in 2010. It was awful: there was no discussion on the bulletin boards, I got A’s on the first four assignments by turning the instructor’s point-form notes into sentences, and so on. I’m now lurking on a MOOC that George Siemens is running, but—well, see the comment in the main post about day-to-day life vs. good intentions…

[2] See also Kanyi Maqubela’s “The Reason Silicon Valley Hasn’t Built a Good Health App“.

Teaching

Audrey Watters’ writings are a must-read for anyone interested in education and technology. As she’s commented many times, many of the techies who want to revolutionize education don’t actually know much about it: they’ve been through school themselves, but either aren’t aware of, or haven’t bothered to investigate, the decades of painstaking research that’s been done on how people actually learn. In response, she’s compiling a “Joel test” (which I think should be dubbed the “Audrey test”) for ed-tech, i.e., a dozen quick questions you can use to tell if someone knows their stuff or is just bluffing. There are already lots of interesting replies to her original tweet; my five suggestions for her list are:

1. Stereotype threat.
2. The unreliability of self-reporting.
3. The importance of self-efficacy.
4. The teacher’s enthusiasm matters more than the teaching method (which is why every new idea initially seems so promising, then does less well when scaled up).
5. Students can learn to do well on drill exercises without actually learning fundamental concepts (e.g., Eric Mazur’s students could get good grades on midterms but still not think in Newtonian physics).

What would be on your list? What are the dozen things you really, really want every technologist to know about education?

Teaching

Someone asked me earlier today if it would be worth Mozilla’s while going to SIGCSE (the big annual computer science education conference in the US). I said, “That’s where the teachers are.” They said, “But we’re teaching non-programmers.” To which I should have replied, “So are most people at SIGCSE, most of the time.” Many (maybe even most?) of the people in “Computing 101″ classes at colleges and universities have never done any programming before, and the grade school teachers who show up are always working with beginners as well. One difference between Moz and regular classroom teachers is that many (most?) college and university instructors focus on students who are likely to become full-time programmers. Another (probably more important) is the setting: free-range vs. battery-farmed.  I’m not sure what to do with these observations, but I think they’re worth pondering. I also think there’s a lot both groups could learn from each other…

Teaching

Software Carpentry has 110 PowerPoint files, each containing between 20 and 120 slides—call it 5000 slides in total. I’d like to convert them to HTML5 for use with Slide Drive, the deck.js+audio slideshow tool that David Seifried is building. Here’s the breakdown:

• A few bugs in the slide decks need to be fixed—it’s a fairly small job.
• I already have audio recordings narrating the slides. A few will need to be redone to sync with bug fixes in the slides, but that’s a fairly small job.
• I also have transcripts of those recordings. They’ll need corresponding edits, and reformatting, but again, that’s fairly small.
• I don’t have time-marks synchronizing the audio with the slides. I’m sure that information is embedded in the Camtasia project files I created when making the current videos, but I don’t know how to get it out (yet).
• Exporting the main text in the slides (the bullet points) is straightforward, though a fair bit of manual touch-up will be needed to reformat it.
• Ditto for the code samples (which don’t show up with the main text, since they’re in separate text objects).

And then there are the diagrams… Roughly a third of the slides have diagrams of some kind, which makes about 1500 in all. That’s too many to redraw, and anyway, I shouldn’t have to: they’re stored in PowerPoint in some kind of vector format, so I should be able to “export as SVG” or the moral equivalent thereof.

But “should” and “can” aren’t the same thing. I can save my PowerPoint in a lot of different raster image formats, but not in a vector format like SVG. However, I can select elements of a diagram, copy them, and then paste them into a vector drawing tool like Inkscape, which indicates that something in Windows understands how to do the required format conversion.

Doing that 1500 times would be very tedious, though. What I really want is a way to automate that process, i.e., save each slide in a deck as a separate SVG that I can later open and edit. Googling turns up a couple of possibilities, but:

• The VeryDOC Powerpoint-to-SVG convert completely drops the embedded text (i.e., the captions).
• Ditto OpenOffice.org Impress—slides must be saved one at a time, and only the graphical elements come out (not the captions).
• Davisor Publishor doesn’t even export all the graphics.
• docx4j‘s pptx4j sounds like it ought to do the job, but (a) it’s a library, not a tool, so some Java would have to be written to create an actual converter, and (b) it’s not clear that it does actually do the job.

Please tell me I missed something…

Later: in response to a suggestion, I downloaded a Windows clipboard viewer and copied a diagram out of PowerPoint. The display looked like this (click the image for full-sized):

It appears to be a reference into the source file, rather than an actual representation of my boxes, arrows, and text.

Software Carpentry, Teaching

[reposted from the Software Carpentry blog]

Back in 2007, Jon Udell observed that screencasts facilitate accidental knowledge transfer in a way that more traditional media don’t. As I said yesterday, we’d therefore like to start recording short screencasts of programmers thinking aloud as they solve small problems using their preferred tools. The aim is to show learners how to program—what order to write things in, how to debug, when and how much to test, and so on. Everything will be covered by the same Creative Commons license as our other material, and made freely available for remixing and other use.

If you’d like to help, please:

1. Volunteer to be recorded by mailing us. We’ll help you install a screen recorder (if you don’t have one already—you might be surprised to find that you do), give you a small problem, and edit the video you produce so that you don’t have to.
2. Volunteer to edit video for us, so that we can put our energy into organizing people .
3. Volunteer to work the floor at PyCon in March. We can’t attend (workshops to run, etc.), but it would be great if we could get a dozen or more “here’s how I do it” recordings done during the conference.

Remember, as an open source project, Software Carpentry depends on your help to survive and thrive. If you have wanted to help, but have worried that creating and recording lectures would be too much work, this is a way for you to help that will take half an hour or less. We look forward to hearing from you.

Software Carpentry, Teaching