Panel
=====
Participants:
	Manuel Freire		MF
	Paul Cristea		PC
	Alexandra Cristea	AC
	Craig Stewart		CS
	Declan Dagger		DD
	Harrie Passier		HP
	Paul Libbrecht		PL
	Michael Yudelson	MY



**
DD	'Pedagogy driven authoring' intro, most AEH work from base level up, this produces quite technical environment. Why not try to do a top-bottom approach.
HP	my exprience is that almost all aplications, are not developed from user's perspective but from technical possibility, when learning is the most difficult process, it's very important to understanding and bear it in mind developing educ. systems.
PC	there's a ;arge variety of types of content presentations, you must create multiple versions of the content and it should be presented differently for diff ppl. Such richness of presentation cannot be achieved singlehandedly. We need to get gather and reuse e/o resources. Otherwise, we will lose the real meaning of personalization.
AC	Main critics of bottom-up approach, they do not know what to expect. Ppl, authorities aren't aware of possibilities in AEH. It is the responsibility of technology to push for 'more diversity'.
DD	We should abstract and speak their language.
MF	We proabbly try to come up with 'partially' adaptive approaches, t make the transition smooth and educate the burocracy about AEH.
PL	We cannot instrumentalize 100% computer education, just part. How far do we want to go there? How should we motivate for adaptivity.
DD	We should create a repository of 'gold standard' methodology in AEH to share and cross implement and reuse it.
xx	most school admins they just don't know these things exists. and they are afraid of embrasing a new technology.
AC	yes, we need to make small step in educaing ppl about adaptivity, and show how they can benefit from it. There's also a q. of roles. You need to have Manbager of LMS, authors, test-users. And those roles also break down further. But there's no 'out of the box' solution and that point shoud be maid clear.



**
HP	LMS, AEH systems are complex. We have to think of how can we support the developers. We have to 1st define what is good, develop representation.
MY	... about 'shared conceptialization' lexicon of AH
AC	this type of common representation is very beneficial, cross interoperability of systems and system data. the theoretical approach to this problem is very valuable. There is already some stantard classifications of things in AEH (e.g. Brusilovsky) we should take it further and come up with higher level reusable language of AEH. So that you can implement and Exchange them across systems. Again adaptivity should be 'stage by stage' low, med, hi.



**
MF	"Visualization of tools and models". Authoring is a bottleneck. Reconcile authoring with mental model. E.g. 'Table of Contents', 'road map'.
HP	suppose you visuzlise the cheching of consistency of UM for e.g.
AC	for a common lang it's like having both the ingredients and the recepy, w/o the recepy it's fine, but with visuzlization you have to show things in visual form.
MF	I meant not only the debugging. You can actually get inside the system and avoid many errors. I personally wanted to use Pentri Nets for verifications
CS	'Zigzag' brainstorm of Ted Nelson: multidimentional spreadsheet - repr of UM
PA	here we have a different dimencion of viz-n, we should adhere to human vizualisation capacities. it'a about getting  a better insight.
AC	'building blocks' stuff...



**
CS	Converotrs. The actual use-base of systems is fragmented. it'll be ideal to draw together everyone to have easy and transparent interoperability. A single core-program or API for conversions will be far more appropriate and very early. 
MY	SCORM does that.
CS	any standard takes time, if you come up with core API, core group of AH designers, you dont have to go through massive standard thing.
MF	what is the difference between the standard and core
AC	the difference in the size of the community to work out things. One thing I noticed about SCORM is that it constraints many things.
CS	I'm not against comon language, I just doubt the speed and still you can't avoid conversions.
MF	what if instead of making programs conversible/convertable, why don't we have a common storage format, reusing e/o stuff. 
AC	yes it is useful, but our mission should be to find good ways to exchange 'the recepy' too.



**
PA	Evaluation driven authoring: a lil overrated. Let's make a disctinction b/w teaching process as itself and evalusation of students' performance. Qualitative & Quantitative evaluation. Here We're describing the quality in quantitative measures. What's the purpose here is to discuss how can you bring the feedback to the authors to augment the course material. Students' grades can describe both the students and the grading process (or course material quality). In AEH fedback towards srudents should go very fast, but feedback to teachers about courseware should be slower, but still should be there.
AC	There is reasearch about such feedback, for e.g. Cristobal Romero. But the reasearch is very small.
MF	Feedback loops: small-stoodent, med-teacher, big - system/'author of courseware'.