Problems and solutions for using computer (networks) for education

[mahmud34-729]

1.Introduction
Section:Next section
I will try to discuss the major issues involved in knowledge dissemination using computers (often called e-Learning) following my own experiences. This certainly introduces some strong bias. However, I do not try to avoid this since I have followed and experienced most developments over the last 50 years myself. Examples and anecdotes will prove some of the points, and will hopefully be both entertaining and useful.

Let me state clearly that this is not a research paper, but a survey of developments, many of them ending up in a dead-end road. I believe that this is what makes this paper also of interest for further research and innovation by showing that certain things do not work or work only if special attempts are made. To make it evident what lessons I have learnt and what one should never forget I will put such "aspects learnt" in boldface.

There is no doubt that Programmed Logic for Automated Teaching Operations (PLATO) was the first e-Learning system using computers, and for that time quite impressive, as it was used in an early timesharing system more than 50 years ago. It originated in the 1960s at the Urbana campus of the University of Illinois. Professor Don Bitzer became interested in using computers for teaching. With some colleagues he founded the Computer-based Education Research Laboratory. Bitzer collaborated with a few engineers to design the PLATO hardware. To write the software, he collected a staff of creative eccentrics ranging from university professors to high school students, few of whom had any computer background. Together they built a system that was at least a decade ahead of its time in many ways.

It is almost impossible to exactly define what PLATO was and how it developed, since it went through an endless process of improvements over the years. The paper (Bitzer, 1986) is an attempt to recapture some developments that I will not discuss in detail. But it is worth noting how powerful the system was, and how many of the good ideas in it around 1970 were only rediscovered decades later! Basically, PLATO around 1970 was still based on black and white terminals able to display mainly written text. Terminals were connected to a central computer presenting information to users whose identity was known and whose behaviour (every key pressing) was carefully monitored. Thus, it was possible to find out where the material presented was too difficult for students at a certain stage in their studies, how long they needed to go through a section. Further, multiple choice questions allowed to understand the level of understanding. This allowed to adjust learning material created with the authoring system called TUTOR. Thus, it was recognized at this early stage that unobtrusively obtained student feedback is important to make it possible to modify material for individuals or groups of individuals and that continuing testing of the level of understanding to provide alternatives and to provide encouraging feedback are crucial issues.

It is truly amazing to see that important issues such as "student feedback is essential" have been ignored in many systems developed later, and even today often not enough feedback is collected without bothering students and often with much too little evaluation of the feedback.

In the years up to about 2000 (when the internet started to be widely available) in North America three main streams are observable:

Plato's development continued and its deployment spread, including offering terminals with graphics and particularly important messaging, between users online at the same time or allowing communication of students at least asynchronously with instructors.
Systems to create sophisticated and animated graphics as part of the "courseware" were developed. One of the many such products was "Macromedia Director", a product to develop stand-alone (typically CD based) learning material. However, note that all those many stand-alone systems did not allow to collect specific feedback, except by using tedious questionnaires!
The emergence of hypertext/hypermedia systems used in timesharing systems typically in university settings provided the possibility to work highly interactively with linked material, including text, animations, graphics, even sound and video. All this was based on Ted Nelson's vision to develop a model for creating and using linked content he called "hypertext" and "hypermedia". Ted Nelson began implementation of a hypertext system named Project Xanadu in about 1960. However, his first and incomplete public release was finished much later, in 1998. He later worked with Andries van Dam to develop the Hypertext Editing System in 1967 at Brown University. In August 1987, Apple Computer released HyperCard, a powerful alternative to Macromedia Director, yet again with no networking facilities to speak of. Yet its impact, combined with interest in Brown University's Intermedia and similar undertakings that seemed to provide new and better platforms for e-Learning, led to broad interest in and enthusiasm for hypertext and new media. Yet all this did not provide a serious impact except as instrument to argue for more funding. This started to change a bit with the advent of the internet, particular the web.
The development in Europe from the late seventies to the turn of the century was somewhat different. In 1969, a British Engineer Sam Fedida proposed to equip TVs (everyone had one by that time, most with a remote control pad that could be used as simple input device) with a bit of additional electronics, called "decoder". This would be connect via phone-lines to a network of servers to allow users to retrieve information, to order anything offered and to write simple messages (akin to today's SMS). The first nationwide systems on that basis, called officially videotex (not to be confused with videotext!), were introduced in the seventies first in Great Britain and then successively in most European countries. In Austria, my home country, I was responsible to recommend whether to also take this approach. I recommended an alternative: to use the same basic idea, but add enough electronics in the decoder to turn it into a small programmable colour-graphic computer and equip it with a full keyboard. Thus, the MUPID, a colour-graphic networkable, programmable device was born, with all programs and data stored in network of videotex servers (today it would be called "in the cloud"). Due to the fact that the network of servers was run by the nations' telecom authorities messaging without spam was easy, senders of messages and information providers could be identified by the users, and micropayments were possible, since the amounts were just added to the monthly phone bill.

Thus, in addition to stand-alone "e-Learning Computer Labs" (at this point in time under abbreviations like computer assisted learning or instruction, computer-based teaching or training and others) videotex and MUPID offered networked variants. This is documented in Videotex (1982), Longley and Shain (1983), Maurer and Sebestyen (1984) and particularly in Maurer (1985, 1986). It allowed the use of colour and of different types of animation. Figure 1 shows four screen dumps from information downloaded with MUPID from the nationwide network. The top row shows that even full-fledged dictionaries were already available at the time, and small pages charges (1P in the example) could be collected after warning the user (a feature still missing in today's internet!). The second row shows two game applications: it was possible to play chess synchronously and asynchronously with arbitrary many persons, at the same time engaged in a (written) chat with other players. The second row shows that games like exploring parts of the universe, including (limited) animation were possible. Indeed, one of the gaming applications proved popular to the extent that the system collapsed under its load during Xmas 1984!

However, only 50,000 MUPIDs were produced in Austria and Germany. The typical IBM PC was getting more accepted, so SW and protocols had to be adapted to PCs and to whatever networks were available, reducing the possibility to use full colour and powerful communication facilities, including central supervision of learners, feedback between learners and courseware supervisors and between learners. It is possible to argue that because of this, first attempts of truly networked e-Learning with all kinds of communication facilities were delayed by almost 20 years until the internet was starting to become accepted and inexpensive enough to allow its use on a large scale.

Nevertheless, one rather unique commercial e-Learning undertaking was started around 1986 as joint work with CDC (using an authoring tool similar to PLATOs Tutor, mainly developed by the late John Garrat) called Computer Supported Teaching of Computer Science (COSTOC) making use of colour and animation as mentioned above. At some stage over 300 one-hour lectures were available and where used in a number of labs in Europe and two in the USA: one at the University of Texas at Dallas under the directorship of Fillia Makedon, and one by the University of Denver under the directorship of the late Peter Warren. Figure 2 shows a few samples. In the first row you see the cover of a brochure on the system and next to it a multiple choice question: it is remarkable that even at this stage the system was capable of handling spelling errors, thus the wrong spelling Autria of Austria was recognized! The bottom two pictures come from a university course on sorting algorithms by Hof bauer and Maurer (1988). Particularly the left picture, showing the sifting down of a value as part of heapsort makes it clear that animation was indeed a powerful tool.

For more literature see Makedon and Maurer (1987a, b), Maurer (1987, 1988), Koegel and Maurer (1987), Makedon et al. (1987), Huber et al. (1989). This very partial list shows the immense interest in what was then called "presentation type CAI" with colour graphics, animation, and some feedback and testing facilities.

However, with the exception of Austria and Germany the COSTOC lessons could not be downloaded from a nationwide network, but at best from some university network, reducing the important feedback to courseware developers. Communication between students and tutors usually required an extra component tailored to the local circumstances and interrupting the learning stream. Hence, most efforts in e-Learning between 1985 and well past 2000 were based on stand-alone or only very locally networked groups of PCs or workstations. That is, they were limited to local e-Learning labs, or even just to e-Learning on an isolated machine with material available via some external storage device ignoring the lessons made already wit PLATO. We will discuss this period in the next section.

Source: Google