<html><head><style type="text/css"></style></head><body><div style="font-family:times new roman,new york,times,serif;font-size:12pt"><div>Hi Tony 1. Most interesting learnable fields have numerous aspects, and many of these require various kinds of skill learning.. The Kokorowski stuff could possibly be of great help in one part of the larger picture. For example, very composer and artist does problem solving, but the center of their arts is not problem solving. Yet, learning how to deal with problems within an art is an important part of the larger picture. Similarly, most forms of instrumental music playing require "chops" (technique) to be developed, and this requires an enormous amount of drill and practicing (and computers these days can provide very helpful tools here). But as Francois Couperin said in his book about learning to play the harpsichord "I would rather be moved than amazed" -- in other words, musical performance is not primarily about chops. The big difficulty in many "educational" schemes, particularly in a pragmatic society in a pragmatic age is that the lesser is substituted for the greater and the greater is redefined to be the lesser. 2. But, I also believe that educational technologies like the OLPC must have a new kind of "mentoring user interface" which is "better than no teacher and better than a bad teacher" (and as much better than these as possible). This has been a 50 year dream and goal (since McCarthy's "Advice Taker" proposal), and it is philisophically possible to achieve (pragmatically, on the scale of Kennedy's "putting a man on the moon within the next decade" goal of the 60s). In other words, "the Dynabook which can help its user learn what is needed" is likely the only reasonable way to scale and distribute the educational quality of experience that the billions of children in the world need. And, this is another example in which the all too human tendency to claim victory before achieving it will be an immense problem in really doing the job well enough. </div><div style="font-family: times new roman,new york,times,serif; font-size: 12pt;"> Cheers, Alan <div style="font-family: arial,helvetica,sans-serif; font-size: 10pt;"><hr size="1">From: "forster@ozonline.com.au" <forster@ozonline.com..au> To: Alan Kay <alan.nemo@yahoo.com> Cc: gregsmithpm@gmail.com; iaep@lists.sugarlabs.org Sent: Monday, June 29, 2009 9:37:47 PM Subject: Re: Re: [IAEP] Comments on David Kokorowski, David Pritchard and ''Mastering'' Educational SW I share Alan and Edward's concern that we are selling kids short, if physics education is all about just getting the 'right' answers but want to break down the question posed by Greg into some sub-questions: 1)Despite the limitations of education which focuses on getting the 'right' answers, more structured or formulaic instruction may still have a place in education. 2)If (1) is accepted, is there a useful role for programmed instruction by intelligent tutors? 3)If (2) is accepted, can intelligent tutors give feedback by Socratic questioning? 4)If (3) is accepted, what are the implications for the Sugar OS? 1)  There is good evidence that more traditional education which focuses on getting the 'right' answers to rather repetitive application of principles learnt in worked examples, results in knowledge which is inert, which cannot be used in real world situations. Real world problems are ill defined, multi disciplinary, with poorly defined goals and multiple or no solutions. Rote learning of standard solutions does little to advance real problem solving skills. If real world problem solving is accepted as the goal, this 'higher order thinking' still requires a repertoire of basic skills including numeracy and literacy. These basic skills may be best taught by traditional rote methods. Arguably Newtonian mechanics could be included in this basic skill set. 2) Programmed instruction depends on how smart the intelligent tutor is.. Masteringphysics claims that through extensive trials, common misconceptions have been identified and the intelligent tutor can react appropriately. Given the millions of ways that it is possible to misunderstand, I retain mild scepticism about their claims but am prepared to accept them for the moment. 3) Socratic questioning exposes inconsistencies in a learner's understanding or mental model: “If you believe A then B flows as a consequence, but B is inconsistent. Do you still believe A?” A good method of 1:1 tutoring, but is the intelligent tutor really smart enough to get into the learner's head space and expose the inconsistencies of their mental model? 4) Should something like Masteringphysics be reproduced in Sugar? Given the immensity in doing the task well, I think no. Hopefully, if we wait, somebody will produce similar materials on the net which are in the public domain. Sugar then would need  reliable network access, a compatible browser and the ability to display multimedia. These issues have already been identified as Sugar priorities. Does Sugar offer the opportunity to do this stuff in a better or different way? I see 2 features of the Sugar environment which make it different and potentially better, to accessing web based materials with other OS's. It is potentially reprogrammable by the student. Something like Masteringphysic's intelligent tutor could be reprogrammed by students. Unfortunately there is usually a disconnect between the programming skills and the physics skills. That is, to reprogram a physics sim suitable for year 9 physics requires year 12 programming skills. Collaboration is built in. Collaboration is not new in IT education. Forums, blogs, wikis, Moodle all facilitate collaboration. Sugar just takes it a little further by allowing easy collaboration at the activity level. It facilitates peer tutoring, which I see mainly as an alternative to intelligent tutors, but maybe there are synergies that I an missing. Tony </div></div></div> </body></html>