Many years ago, when LabKitty was just a LabKitten, computer games were text based. You played them by gathering around a teletype in the back of a ginormous windowless room housing a PDP-10 and enough air conditioning to turn the Sierra Madre into permafrost, usually after the normals left for the weekend and you got a stern lecture from the IT guy about the gruesome death everyone will experience should anyone trip the halon and, no, I don't care if it was an accident or not.
The hot title of the day was Zork, an adventure game in which you took on the persona of an intrepid adventurer and adventured forth seeking adventure. It was sort of Skyrim without graphics, or sound, or any kind of interaction beyond typed words. Skyrim played by someone in another room, and all you do is communicate with them via text messages.
Zork is still around in various incarnations, including at least one site (www.web-adventures.org) where you can play it in your browser. Which I did. Here's a few lines from the opening of the game to give you a feel of the Zork experience. Lines starting with ">" are what the player (me) typed, everything else is generated by the game:
Welcome to ZORK.
This is an open field west of a white house, with a boarded front door.
There is a small mailbox here.
> open mailbox
You open the mailbox, revealing a small leaflet.
> eat leaflet
I don't think that the small leaflet would agree with you.
> burn leaflet
What do you want to burn the small leaflet with?
> matches
What about matches?
> burn leaflet with matches
I see no matches here.
> bite me, ya PDP-11 wannabe.
Such language in a high-class establishment like this!
> *U$#*$*(GR*(FG!78
That's not a verb I recognize.
This is an open field west of a white house, with a boarded front door.
There is a small mailbox here.
> open mailbox
You open the mailbox, revealing a small leaflet.
> eat leaflet
I don't think that the small leaflet would agree with you.
> burn leaflet
What do you want to burn the small leaflet with?
> matches
What about matches?
> burn leaflet with matches
I see no matches here.
> bite me, ya PDP-11 wannabe.
Such language in a high-class establishment like this!
> *U$#*$*(GR*(FG!78
That's not a verb I recognize.
As you can see, the game often leads to enraged keyboard mashing. This got me thinking about homework.
Computer Aided Instruction was promised to be the Next Big Thing for decades, just as soon as we're able to make it work properly. I actually published my first paper on CAI (and later went on to a brief career in CAD and CAM before I realized I didn't have to do something I really didn't enjoy for lots of money when I could write a blog and be unemployed). I remember presenting the paper at some big engineering conference to an empty room. Nobody cared. If that isn't a concise summary of the current state of American education, I don't know what is.
Still, many continue to champion the CAI cause, taking up the banner in various formats and with varying degrees of success. Khan Academy comes to mind, which began as Salman Khan making simple YouTube videos of worked math problems and has since grown into an on-line educational juggernaut. Much has been made of their "flipped classroom" philosophy, an inversion of the traditional roles of class time and homework. However, we may summarize the benefits offered by their (astonishingly diverse) video selection rather simply: these can be viewed by the student at their own pace, with no penalty (academic or otherwise) for repeated viewings. An MPEG player does not judge you for rewinding.
Yet, for all of Khan Academy's well-intentioned lectures, they are just that: lectures. The interaction is one way, which is to say no interaction at all (what Neal Stephenson calls a "passive" in The Diamond Age). That is not to say they don't have value -- again, the flipped classroom model has much to recommend. But I claim CAI can do more. Something truly interactive. (What Stephenson calls a "ractive" in The Diamond Age. I'm currently reading The Diamond Age, I guess is my point here.)
Interactive instructional software started getting bundled with textbooks as soon as the PC became commonplace. To take but one example, many titles in the popular Schaum's Outline series include their "Electronic Tutor" -- a Windows program that acts as a sort of glorified graphing calculator guiding the student through selected worked problems in which you enter values and a plot related to the solution changes. This is a common motif in both stand-alone and on-line instructional resources alike, with some sort of graphical feedback attached to guided parameter exploration. Somewhat lazier offerings simply offer flash cards or randomized practice questions, which may lend a new sheen to an old standby but really does not exploit the technology much if at all.
As I see it, present efforts at interactive CAI invariably suffer the same fatal flaw: they are afraid to challenge the student. It's as if the instructional software industry doesn't understand their target audience. Because, clearly, that audience is willing and able to commit extended effort to complex problem solving via computer. Whether it's weeks spent beating Zork, or getting a Kerbal successfully to the Mun, or constructing a functioning Babbage engine in Minecraft, we have a platform that routinely engenders obsessive creative behavior. Why don't we exploit such obsession in CAI? That question is not rhetorical, and I believe the answer lies largely in that current CAI efforts spoon feed the student.
A cynical response might be this is the current state of the art in CAI because this is the current state of the art in CAI. Additional sophistication requires advances in artificial intelligence -- solving hard AI problems such as natural language processing. But Zork certainly has not solved any hard AI problems and it provides a rich interactive environment that encourages -- nee demands -- reflection, ingenuity, and imagination. In short: it develops complex problem solving skills. Why not use the extant Zork algorithms to construct a new breed of educational software?
The best illustration of the concept I can conceive is a simulation. Consider a topic from first semester physics (this would probably be used within the context of a physics course, where the relevant topics have been recently presented in lecture):
-- Welcome to ZfIS --
A missle launcher fires a rocket horizontally and recoils up an inclined plane, rising to a height of 4 meters. Find the initial speed of the rocket.
> show answer
Nice try.
> use newton laws
I see no newton laws here.
> examine inclined plane
The plane is smooth and inclined at an angle theta.
> examine launcher
The launcher is mounted on wheels that roll without friction.
> examine rocket
The rocket is made by the Acme Company, but is otherwise unremarkable.
> find the initial speed of the launcher
I don't know how to do that.
> show equation
I don't know what equation you mean.
> show equation of motion for uniform acceleration
I know three such equations:
1) v = v0 + a • t
2) v^2 = v0^2 + 2 • a • x
3) x = x0 + v0 • x + 1/2 • a • t^2
> use equation 2 on rocket launcher
I don't know how to do that
> solve equation 2 for v0
Rearranging equation 2 we have
2b) v0 = √ (v^2 – 2 •a • x)
> put v = 0 a = 9.8 and x = 4 into equation 2b
That results in the square root of a negative number.
> put v = 0 a = -9.8 and x = 4 into equation 2b
You find v0 = 8.85 m/s
> show equations for conservation of momentum
The conservation of linear momentum states: p_initial = p_final.
> define momentum
Momentum (p) is the product of velocity and mass.
> weigh rocket
You weigh the rocket. The result is 110 kg.
> weigh launcher
You weigh the rocket launcher. The result is 4400 kg.
> solve (4400)(8.85) = 110(x) for x
The result is x = 354.
> the initial speed of the rocket is 354 m/s
Correct!
A missle launcher fires a rocket horizontally and recoils up an inclined plane, rising to a height of 4 meters. Find the initial speed of the rocket.
> show answer
Nice try.
> use newton laws
I see no newton laws here.
> examine inclined plane
The plane is smooth and inclined at an angle theta.
> examine launcher
The launcher is mounted on wheels that roll without friction.
> examine rocket
The rocket is made by the Acme Company, but is otherwise unremarkable.
> find the initial speed of the launcher
I don't know how to do that.
> show equation
I don't know what equation you mean.
> show equation of motion for uniform acceleration
I know three such equations:
1) v = v0 + a • t
2) v^2 = v0^2 + 2 • a • x
3) x = x0 + v0 • x + 1/2 • a • t^2
> use equation 2 on rocket launcher
I don't know how to do that
> solve equation 2 for v0
Rearranging equation 2 we have
2b) v0 = √ (v^2 – 2 •a • x)
> put v = 0 a = 9.8 and x = 4 into equation 2b
That results in the square root of a negative number.
> put v = 0 a = -9.8 and x = 4 into equation 2b
You find v0 = 8.85 m/s
> show equations for conservation of momentum
The conservation of linear momentum states: p_initial = p_final.
> define momentum
Momentum (p) is the product of velocity and mass.
> weigh rocket
You weigh the rocket. The result is 110 kg.
> weigh launcher
You weigh the rocket launcher. The result is 4400 kg.
> solve (4400)(8.85) = 110(x) for x
The result is x = 354.
> the initial speed of the rocket is 354 m/s
Correct!
I expect there would be much more exploration and dead ends which I've kept to a minimum here, but this sample demonstrates the concept: Computer aided instruction offering a truly interactive experience, and one that is not afraid to challenge the student -- unapologetically, substantially, (perhaps even insurmountably) -- if it means the rewards are made all the more meaningful.
This example is from mechanics but extensions to other topics in physics are obvious. Beyond that, there is fertile ground to be explored in chemistry, engineering, mathematics, and elsewhere. A library of interactive exercises, available to all for strengthening problem solving and creative thinking.
ZfIS. Zork-flavored instructional software.
No comments:
Post a Comment