I came back stimulated and with some new ideas from CAL97, so on the most important criterion the conference was a success for me. The most interesting thing was meeting Martin Gilbert from UCL who is developing devices for dynamic student feedback from minute to minute from students to lecturers with the expectation of replicating Dutch data showing large improvements in exam results. Another was a long conversation over the conference dinner about the factors that really determine organisational change in teaching methods (which controls whether CAL will actually make a difference). The third was Jonathan Briggs' emphasis on the notion of "push", which got me thinking about what concepts we need to think about databases and broadcast television in a single framework of information sources and uses.
In contrast to the informative unplanned conversations with delegates, the plenary "keynote" speakers, with the sole exception of the after dinner speaker, seemed to be completely out of touch with a) the experiences reported in the refereed papers, b) educational ideas and educational practice, c) all knowledge of the history of technology in education. They were all uncritical techno-enthusiasts, with the further exception of Sir Geoffrey Holland who concentrated mainly on the pressures for change in education rather than on technology as a panacea. None reported on achieved educational benefits and none seemed aware that all the promises they were repeating had been made before and not fulfilled for other technologies. In fact we live in interesting times, but not interesting in the way the keynote speakers made out. If computers and communications technology are going to make a difference to teaching and learning, then we need a theory of why they are different from radio, television, cinema etc., all of which failed to do so although all were novel worldwide communications technologies. If CAL (or the internet) is to make much difference it must raise learning quality and quantity while reducing costs. So far there are almost no studies in which this has actually occurred. Typically it can raise quality for increased costs, or reduce costs while reducing quality. The latter happens when people who believe the plenary speakers and their ilk replace current teaching by machines of one kind or another only to discover that they hadn't identified all the different functions the human teachers were in fact covering. The interesting current challenge is to find the combinations that work well: but few successes have so far been reported by workers in the field, while the visionaries haven't even identified the issue yet. Thus the keynote speakers seemed rather old fashioned: still talking about technology-led advances, whereas the lessons of the last few years' work seems to be that innovation needs to be education-led and evaluation-supported.
The contrast I experienced in where I found the most stimulation (in the unplanned contacts not in the organised speakers) is perhaps not unusual, but it seemed particularly typical of this conference where nearly everything at the time went very well, although most things about the advance organisation were poor. The facilities were mainly excellent: on one site, comfortable accommodation and good food, a high helper:delegate ratio to overcome problems, and the ideal conference bar (on the spot, only one bar so you only had one place to go, large enough, and cheap). However although things worked out well in the end, the omens had not been good especially for presenters. There was a discount for early registration, but decisions about paper acceptance (at least after the demanded corrections) were not made in time for this. The start time of a Sunday afternoon was the worst possible: both rail and air travel are much worse on Sunday, so almost no-one could arrive in time on the first day, but those who were thereby forced to come a day early had a wasted half day. Given that travel was made unnecessarily inconvenient, this was compounded by the lack of advance information about when each presenter's slot was. In fact presenters were not even told how long their slot would be: thus having submitted a proposal for a workshop that ran well in a half-day slot at a previous conference, I only discovered the day before I travelled that it was scheduled for a one hour slot (actually 50 mins, given the need to move between buildings). And I was the only speaker I talked to who had even that amount of advance information by dint of repeated phoning and searching web pages in the computer science department. The timetable put events in different buildings but no time to move between them, forcing even chairpersons to cut one or other session short so they could sprint to their next commitment. The name badges were unreadably small: many conferences are bad at this, but these were the worst I've seen. Most of the meeting rooms were much too small. 200 delegates and 6 parallel sessions means you must expect an average (not a maximum) of 30 per session, but the rooms were organised to seat about 15-20, and every session apart from the plenaries I went to was disrupted by overcrowding. Most telling of all, the "feedback form" had no place for commenting on any of these issues, suggesting just how far from the organisers' minds they were. It is interesting to see how many of those listed as on the "programme committee" are not listed on the delegate list: they would have been in the best position to know about the organisation and apparently didn't fancy it, just as only one of the keynote speakers bothered to attend even half the conference.
Still, it turned out that such pessimism was misplaced: real human dynamics made for a good conference despite all the poor planning. Perhaps there is a lesson for CAL here: CAL and technologically based teaching requires far more accurate advance design and organisation than face to face teaching. A much more reliable recipe is the one the conference embodied: keep the numbers down, get a good bunch of people, make them travel so there is real face to face contact, a lot of it not planned or scheduled, and have lots of on the spot assistants running round solving problems as they come up. That is what the conference showed actually works. If the keynote speakers had stayed to attend the conference (or of course thought about their own education), they might have discovered that too.