This is a relatively old newsletter from David W. Johnson, Roger T. Johnson (2 of the greatest names in this field) and Edythe Holubec.
The first point they made is to urge us to abandon the "Individual Assumption". They wrote:
The individual assumption is that instruction should be tailored to each student's personal aptitude, learning style, personality characteristics, motivation, and needs. Computers were originally viewed as an important tool for providing individualized learning experiences. The attempts to individualize instruction through technology clarified that so much variation exists in learning styles and personality traits and so little agreement exists on how to translate differences in learning styles and personal traits into instructional prescriptions that attempting to do so is exorbitantly time consuming and expensive (requiring considerable development and hardware costs). Even if it were possible it individualize, working alone limits the resources available to learn and the isolation tends to result in boredom, frustration, anxiety, and the perception that learning is impersonal.
I agree whole-heartedly to their argument. To provide a complete individualised program for "solo" learning, we have to exhaust all possible potential responses from the learners and cater for each and everyone of them in different combinations. Take a very simple example. If we are to provide feedback to students based on their answers to three simple multiple choice items (each with 4 possible choices), we would have to provide a set of feedback to cater for 4x4x4 possibilities (64 in this case). If we allow customising items based on the answers to previous items, the feedback "space" will be much bigger.
I have long given up on this. Rather I prefer to depend on peer to peer feedback - collaboration/co-operation.
I also do not like the idea of providing a closed choice learning environment (i.e. a learning environment where all the possible responses are anticipated by the designer and learners are basically faced with a glorified multiple choice). I prefer open answer learning environment where learners provide response using their own words and imagination. The mental activities between picking a choice out of a few and making up a response are qualitatively different, the latter being much more valuable in real life situations. Unfortunately, up to now, computers are still not very good in processing open responses although some exciting development is taking place. The solution to this problem, ironically, is based on providing "human intelligence" rather than "artificial intelligence".
They continue to suggest that
There are two ways that technology-supported cooperative learning may be used: (a) cooperation around computers and (b) cooperation through computers.
computers can be based on software that is designed originally for solo use, but we introduce creative learning strategies to let groups of learner to learn around the software.
computers assume the computers are connected and they become a tool for learners to reach out to a wider world than they normally can. This is where collaboration tools come in. Today, we see a great many such tools. One I routinely use is Skype (http://www.skype.com) which allows me to talk to anyone who is also on the Skype network free. Based in Melbourne, Australia, I am able to work with my US customers continuously throughout their whole working hours via Skype. Without it, the exuberant price of Australian telecommunications will have sent me bankrupt :-) .
I still hold the view that collaboration tools are great by themselves. For effective learning, we need to have clear protocols (or structure) so that learners can work efficiently and effectively. The distinction between co-operative learning and collaborative learning, to me, lies in how such protocols were negotiated and formed.