Sunday, July 27, 2008

Raymond Lister's paper

After the Gold Rush: Toward Sustainable Scholarship in Computing (pdf, 16pp) by Raymond Lister, Faculty of Information Technology, University of Technology, Sydney, Australia

Mark Guzdial, on his blog, described this as a "great paper" and then uses it to promote Cognitive Load Theory and its critique of constructivism or constructionism (link to Mark Guzdial's post)

Cognitive load theory keeps coming up again and again as an alleged refutation of constructivism. So, I read the Lister paper with that in mind. I have previously written a rebuttal of the Kirschner et al paper which claims that minimal guidance during instruction does not work (one of the relevant paper cited by Lister)

The Lister paper is about the authors journey from being a bad teacher of computer programming to a better teacher and some of the attitudes and pitfalls encountered along the way. It's an admirable paper from that point of view, of someone becoming aware of what emerges as rather extreme deficiencies of many teachers and deciding to begin to tackle this seriously.

Here is a brief summary of some of the problem attitudes identified, at length, by the author:
  1. the assumption that all students learn in similar ways (one mould fits all)
  2. the reluctance of those who have been successful within a system to question deeply the way that system does things
  3. more years of teaching means you teach better
  4. the most outspoken individuals often dominate curriculum change
  5. teachers don't criticise other teachers, the culture of silence
  6. Anecdotal flag planting, I tried this and it worked
  7. Lack of real evidence
  8. Systemic separation of the theory of teaching and learning from particular disciplines
  9. Teachers resist researching their own teaching
  10. Teachers blame their colleagues at lower level levels when their students don't know stuff
  11. If students fail then blame the student, not your teaching. Students might be lazy, spend too much time in paid work or are genetically deficient (common belief amongst teachers when it comes to learning how to program)
  12. the need to move past the industrial model paradigm to a more ecological model, in harmony with the teaching and learning environment
For the author the wakeup call was too many of his students failing and students not learning how to program.

He describes some ways in which his teaching has become more interactive with his students and various papers he has read in his search to become a better teacher and also to understand the reasons for the decline in IT enrolments. He takes the trouble to find out the real reasons why students are obtaining wrong answers to "easy" questions and explores some interactive techniques, such as asking students to explain "in plain English" what a short piece of code did.

So far, so good

But along the way the author has quoted from some papers on Cognitive Load Theory and agrees with those authors that constructivism is deficient. He goes onto say that the Australasian University Teaching and Learning (T&L) communities are heavily influenced by constructivism and contrasts this with "discipline based academics" who tend to focus mainly on their disciplines at the expense of teaching and learning theory. Towards the end, however, the author repeats his call for the bringing together of theory and practice based on a "social constructivist view of the world"

The authors view of constructivism is taken uncritically from the Cognitive Load Theorists (link to the original Kirschner, Sweller and Clark paper) and is then oversimplified further by him to equate constructivism with problem solving:
On the basis of that definition, computing education has used constructivist approaches for decades. For example, many of us introduce students to programming via the problem-solving approach, which McCracken et al. (2001) defined as an approach where we provide students with a problem description, and then require them to decompose it into sub-problems, implement them, test them, then assemble the pieces into a complete solution (Lister, p. 6)
Hence, Lister misunderstands the issue of what the Papert version of constructionism is because he relies on the inadequate definition emanating from these critics of constructivism. (Rather than researching what supporters of constructivism are saying.) As I pointed out in my earlier rebuttal there is only a single Papert reference in the Kirschner, Sweller and Clark paper even though Papert is recognised as an authority and has authored many papers, books and supervised many PhD theses. Now we see this harm being spread to Lister and then onto Guzdial.

Most tellingly, many of the positions that Lister puts as problems to be overcome are in fact similar to issues that the constructionist Papert has long ago identified as problems to be overcome. For example, Lister explains at some length about the problem of constructivist theory (T&L departments) and the practice of disciplines being kept separate. But Papert has famously said:
You cannot think about thinking without thinking about thinking about something
In their paper, Software Design as a Learning Environment (1990), Idit Harel and Seymour Papert, focus on and discuss in detail four important issues in developing constructionist learning environments, in this case, in the learning of fractions:
  • Development of Concept - the need to move beyond rigid, particular and isolated understanding to more flexible, generalised and connected understanding
  • Appropriation of the Project - taking personal ownership enhances learning
  • Time Frame and Rhythm of Work - School time is organised in rigid and fragmented segments, whereas experts and an apprenticeship model has a totally different feel to it
  • Metacognitive Awareness - thinking about ones own knowledge and understandings is an important part of learning
Plenty of scope for interactive teaching here.

One final point, briefly. Having now read Marvin Minsky's book The Emotion Machine, I can now see ways to improve my rebuttal of the Kirschner, Sweller and Clark paper. This does connect to Lister's concerns about the relative inadequacy of "folk pedagogy" and citing Cognitive Load Theory as an antidote. I can see now that Cognitive Load Theory might just be another form of folk pedagogy with pseudo scientific terms like "working memory" and "long term memory", concepts that sound scientific but have yet to be explained. See Minsky, page 243.

At any rate, to suggest, as the Lister paper does, that Papert's constructionism is a theory that eschews a deep study of interactivity with the learner, that it means something like give them a problem and walk away, is a gross misrepresentation. (although, to be fair, this may well be true of some of the University based Teaching & Learning faculties, so muddied have the constructivist waters become)

3 comments:

artichoke said...

Bill you might enjoy the Matthew Tabor's Blog (if you don't already read it) his ability to critique ideas reminds me a lot of the intellectual honesty in your posts

Bill Kerr said...

I left this comment on Mark Guzdial's blog:

[quote]
Raymond Lister's paper misrepresents constructivism as just problem solving, in part, by taking its definition from Cognitive load theory. More here:
http://billkerr2.blogspot.com/2008/07/raymond-listers-paper.html

To teach algebra by just giving students problems, walking away and letting them struggle with it? I'm not aware of anyone suggesting that.

Improve the quality of worked examples? Possibly, everyone would agree with that. Constructionism recommends minimising instruction, not doing away with it entirely.

The typical low level instructionist paradigm would be for the teacher to rely on the worked examples and then set problems out of a text book

A constructionist example could be to explore line gradients using logo, eg. how could you emulate these line gradient using logo and then connect that up to the linear equation in some way.

The point I'm making is that a real discussion of the issues does not connect very well to the valid but restricted points of Cognitive Load theorists, that our working memories are limited. The constructionist response to that is, take your time, learn it in mind size bites.
[/quote]

Bill Kerr said...

thanks arti,

the pieces at matthew's blog about my friend gary stager's one liners and about principal's not grasping maths were valuable and challenging, so I see what you mean :-)