ActionScript as a teaching language

[Paul King]

On 2 Jan 2006 at 12:01, D. Hugh Redelmeier (D. Hugh Redelmeier <tlug-lxSQFCZeNF4 at public.gmane.org>) spaketh these wourdes:

> I have no idea where ActionScript is in this life cycle.
> 

ActionScript is into Version 3, now that it is no longer a Macromedia product 
(Adobe closed the deal to acquire Macromedia in mid-December). As of version 2, 
it is object-oriented to the point where you can make your own objects, but the 
scoping in version 2 was still, to me at least, a hodgepodge of scoping to 
timelines, graphics, curly brackets, with no unified place to place all your 
code. It appears as though for most things you need to click on an object or 
timeline, then Flash opens up a new editing buffer for it and you just write code 
for that object. If a visual object has a timeline, it is not clear to me when 
you attach the code to the object and when you attach it to the timeline. 
Everything is too tied to the interface to be sure of anything, especially when 
things get bigger and bigger. On the actionscript newsgroups, there appears to be 
general agreement that AS's biggest drawback is in how the editor handles the 
code, by segmenting it in the way I described.

To contrast with Visual Basic, at least if you are tying code to a button or 
other object, all of the code for the same form lies in a single source file. 
Flash only shows code with the object you selected, and hides the rest of the 
code from you. Visual Basic uses a single buffer for all of your code (unless you 
desire to use external code), but Flash has multiple buffers with a chunk of code 
here and a chunk of code there. You can use external code in Flash, also.

> | The LCSI link seems to be a good one. I like the fact that it says that these are 
> | *constructivist* tools, since constructivism is now the new educational "flavour 
> | of the month".
> 
> I'm not sure what "constructivist" means in education (I know that I
> could Google).  I can guess.  I bet it is what Papert has been
> talking about for 40 years.

Papert has been discussing teaching by "construction", but constructivism seems 
to be somewhat different. Constructivism can be applied to English, Social 
Studies, Art, Math, and anything else. Constructivism is based on the notion that 
children need to "construct" their own meanings and definitions and relationships 
out of the activities in front of them (which could be a poster board activity or 
whatnot, depending on what the content of the lesson is). They are asked to 
recognise patterns for themselves, and derive relationships for themselves 
through "hands-on" activities.

Constructivist theorists believe that they are supported by what is known from 
the workings of the brain. The more we learn, the more neural connections we 
make. We make the learning experience our own by recognising patterns, 
classifying, and so on. It is now believed by constructivists that the way we 
learn is by making associations. And when we do learn, new neural pathways are 
made to store the learning. And the more learning we do, the faster we will pick 
up on new learning, since many new neural connections are already there.

And as I said, if you are busy coding away, you are going to learn many things in 
a constructivist sense anyway, especially regarding things like computer math, 
and syntax rules, even when you are told in advance what the proper syntax is. 
And of course the big payoff comes when you see it all work.

> 
> LCSI was founded by Seymour Papert, his students, and others.  Papert
> is well known as one of the inventors of Logo and as the author of
> Mindstorms and other important avant garde educational stuff.  He was
> an actual student of Piaget.  So there is a pretty respectable
> heritage there.
> 

Very interesting.  I ought to read up some more on him I suppose.

> | I am curious as to how they see programming as being covered in a 
> | constructivist manner. I would have thought that programming itself is 
> | constructivist anyway, regardless of the language. CS is inherently hands-on, and 
> | when you have a bug or syntax error, you learn from your own mistakes. You can't 
> | get more constructivist than that.
> 
> Last I talked with LCSI folks (ten years ago), they were not dealing
> with programming as an overt goal.  Kids were using programming in
> Microworlds to accomplish other tasks.  I don't imagine that any of
> their material is directed at a programming course.  That did not
> prevent Microworlds being used to teach programming.

>From the website, it appears as though programming is still not an overt goal. 
But their approaches to science and math still interest me, since I can teach 
them as well.

> 
> Example (that I am making up):
> When a student is using Microworlds for (say) modelling a physics
> problem, Microworlds directs/enables them in a Constructivist way.
> 

This means that kids are directed to both ask and answer "the right question" 
about their activity, and not just ask any old question. While constructivism 
must be somewhat open-ended, you as the teacher have objectives to teach, and 
therefore things can't be too open-ended. All of the kids have to learn as a 
class. And there is usually a test at the end, so they have to all get exposure 
to the same questions, one way or another.

> High school students may find Microworlds' surface is aimed at too
> young an audience.  It may well put them off.  I know that there is
> serious stuff underneath, but that may not be apparent to the
> students.
> 
> One tough problem in CS is to find good motivating problems.  Maybe
> Constuctivists would focus on that.  

Yes, that has always been difficult. But I am not sure if constructivism per se 
will provide more motivating problems. Constructivism has to do with the 
student's individual learning style. Although now that I think of it, concepts 
such as recursion (another Grade 11 topic) could be well served by a 
constructivist activity.

> Teaching syntax for its own sake
> might be the opposite approach.
> 

You can't avoid doing that, anyway.

> For young kids, the Logo turtle was a wonderful come-on.  Maybe flash
> would do the same for ActionScript.
> 
> 
> I don't know enough about what you are hoping to do to recommend a
> particular language.  Mostly I've been trying to point out that
> Logo and LISP are not mistakes for the reasons you mentioned.

I was just fishing for language suggestions that have an impact on a Grade 11 
student similar to ActionScript without the syntax and editing horrors I 
mentioned in this and previous posts.

Thanks for your insights.

Paul King

--
The Toronto Linux Users Group.      Meetings: http://tlug.ss.org
TLUG requests: Linux topics, No HTML, wrap text below 80 columns
How to UNSUBSCRIBE: http://tlug.ss.org/subscribe.shtml