In the following I reflect on my aborted and half-baked attempts at harnessing design patterns within the practice of e-learning at universities and wonder whether it was a lost opportunity and/or a project that was destined to fail. This is written in the light shed by the work of a number of other folk (Google “patterns for e-learning”), including the current JISC-emerge project and, I believe, the related Pattern Language Network.
I think I’ll end up contending that it was destined to fail and hope I can provide some justification for that. Or at least that’s what I currently think, before writing the following. Any such suggestion will be very tentative.
Way back in 1999 I was a young, naive guy at the crossroads of software development and e-learning, I was wondering why more academics weren’t being innovative. Actually, the biggest and most troubling question was much simpler, “Why were they repeating the same mistakes I and others had made previously?”. For example, I lost count of the number of folk who tried to use email for online assignment submission in courses with more than 10 or 20 students. Even though many folk tried it, had problems and talked about the problems with additional workload it creates.
At the same time I was looking at how to improve the design of Webfuse, the e-learning system I was working upon, and object-oriented programming seemed like a good answer (it was). Adopting OOP also brought me into contact with the design patterns community within the broader OOP community. Design patterns within OOP were aimed at solving many of the same problems I was facing with e-learning.
Or perhaps this was an example of Kaplan’s law of instrument. i.e. patterns were the hammer and the issues around e-learning looked like a nail.
Whatever the reason some colleagues and I tried to start up a patterns project for online learning (I’m somewhat amazed that the website is still operating). The why page” for the project explains the rationale. We wrote a couple of papers explaining the project (Jones and Stewart, 1999; Jones, Stewart and Power, 1999), gave a presentation (the audio for the presentation is there in RealAudio format, shows how old this stuff is) and ran an initial workshop with some folk at CQU. One of the publications also got featured in ERIC and on OLDaily.
The project did produce a few patterns before dieing out:
There’s also one that was proposed but nothing concrete was produced – “The Disneyland Approach”. This was based on the idea of adapting ideas from how Disney designs their theme parks to online learning.
I can’t even remember what all the reasons were. Though I did get married a few months afterwards and that probably impacted my interest in doing additional work. Not to mention that my chief partner in crime also left the university for the paradise of private enterprise around the same time. That was a big loss.
One explanation and a “warning” for other patterns projects?
At the moment I have a feeling (it needs to be discussed and tested to become more than that) that these types of patterns projects are likely to be very difficult to get to work within the e-learning environment, especially if the aim is to get a broad array of academics to, at least, read and use the patterns. If the aim is to get a broad array of academics to contribute to patterns, then I think it’s become almost impossible. This feeling/belief is based on three “perspectives” that I’ve come to draw upon recently:
- Seven principles for knowledge management that suggest pattern mining will be difficult;
- the limitations of using the Technologists’ Alliance to bridge the gap;
- people (and academics) aren’t rational and this is why they won’t use patterns when designing e-learning and
7 Principles – difficulty of mining patterns
Developing patterns is essentially an attempt at knowledge management. Pattern mining is an attempt to capture what is known about a solution and its implementation and distill it into a form that is suitable for others to access and read. To abstract that knowledge.
Consequently, I think the 7 principles for knowledge management proposed by Dave Snowden apply directly to pattern mining. To illustrate the potential barriers here’s my quick summary of the connection between these 7 principles and pattern mining.
- Knowledge can only be volunteered it cannot be conscripted.
First barrier in engaging academics to share knowledge to aid pattern mining is to get them engaged. To get them to volunteer. By nature, people don’t share complex knowledge, unless they know and trust you. Even then, if their busy…. This has been known about for a while.
- We only know what we know when we need to know it.
Even if you get them to volunteer, then chances are they won’t be able to give you everything you need to know. You’ll be asking them out of the context when they designed or implemented the good practice you’re trying to abstract for a pattern.
- In the context of real need few people will withhold their knowledge.
Pattern mining is almost certainly not going to be in a situation of real need. i.e. those asking aren’t going to need to apply the provided knowledge to solve an immediate problem. We’re talking about abstracting this knowledge into a form someone may need to use at some stage in the future.
- Everything is fragmented.
Patterns may actually be a good match here, depending on the granularity of the pattern and the form used to express it. Patterns are generally fairly small documents.
- Tolerated failure imprints learning better than success.
Patterns attempt to capture good practice which violates this adage. Though the idea of anti-patterns may be more useful, though not without their problems.
- The way we know things is not the way we report we know things.
Even if you are given a very nice, structured explanation as part of pattern mining, chances are that’s not how the design decisions were made. This principle has interesting applications to how/if academics might harness patterns to design e-learning. If the patterns become “embedded” amongst the academics “pattern matching” process, it might just succeed. But that’s a big if.
- We always know more than we can say, and we will always say more than we can write down.
The processes used to pattern mine would have to be well designed to get around this limitation.
Limitations of the technologists’ alliance
Given that pattern mining directly to coal-face academics is difficult for the above reasons, a common solution is to use the “Technologists’ Alliance” (Geoghegan, 1994). i.e. the collection of really keen and innovative academics and the associated learning designers and other folk who fit into the left hand two catagories of the technology adoption life cycle. i.e. those to the left of Moore’s chasm.
The problem with this is that the folk on the left of Moore’s chasm are very different to the folk on the right (the majority of academic staff). What the lefties think appropriate is not likely to match what the righties are interested in.
Geoghegan (1994) goes so far as to claim that the “alliance”, and the difference between them the righties, has been the major negative influence on the adoption of instructional technology.
Patterns developed by the lefties are like to be in language not understood by the righties and solve problems that the righties aren’t interested and probably weren’t even aware existed. Which isn’t going to positively contribute to adoption.
People aren’t rational decision makers
The basic idea of gathering patterns is that coal face academics will be so attracted to the idea of design patterns as an easy and effective way to design their courses that they will actually use the resulting pattern language to design their courses. This ignores the way the human mind makes decisions.
People aren’t rational. Most academics are not going to follow a structured approach to the design of their courses. Most aren’t going to quickly adopt a radically different approach to learning and teaching. Not because their recalcitrant mongrels more interested in research (or doing nothing), because they have the same biases and ways of thinking as the rest of us.
I’ve talked about some of the cognitive biases or limitations on how we think in previous posts including:
In this audio snippet (mp3) Dave Snowden argues that any assumption of rational, objective decision making that entails examining all available data and examining all possible alternate solutions is fighting against thousands of years of evolution.
Much of the above applies directly to learning and teaching where the experience of most academics is that they aren’t valued or promoted on the value of their teaching. It’s their research that is of prime concern to the organisation, as long as they can demonstrate a modicum of acceptable teaching ability (i.e. there aren’t great amounts of complaints or other events out of the ordinary).
In this environment with these objectives, is it any surprise that they aren’t all that interested in spending vast amounts of time to overcome their cognitive biases and limitations to adopt radically different approaches to learning and teaching?
Design patterns anyone?
It’s just a theory
Remember what I said above, this is just a theory, a thought, a proposition. Your mileage may vary. One of these days, when I have the time and if I have the inclination I’d love to read some more and maybe do some research around this “theory”.
I have another feeling that some of the above have significant negative implications for much of the practice of e-learning and attempts to improve learning and teaching in general. In particular, other approaches that attempt to improve the design processes used by academics by coming up with new abstractions. For example, learning design and tools like LAMS. To some extent some of the above might partially explain why learning objects (in the formal sense) never took off.
Please, prove me wrong. Can you point to an institution of higher education where the vast majority of teaching staff have adopted an innovative approach to the design or implementation of learning? I’m talking at least 60/70%.
If I were setting the bar really high, I would ask for prove that they weren’t simply being seen to comply with the innovative approach, that they were actively engaging and embedding it into their everyday thinking about teaching.
What are the solutions?
Based on my current limited understanding and the prejudices I’ve formed during my PhD, I believe that what I currently understand about TPACK offers some promise. Once I read some more I’ll be more certain. There is a chance that it may suffer many of the same problems, but my initial impressions are positive.
Geoghegan, W. (1994). Whatever happened to instructional technology? 22nd Annual Conferences of the International Business Schools Computing Association, Baltimore, MD, IBM.
David Jones, Sharonn Stewart, The case for patterns in online learning, Proceedings of Webnet’99 Conference, De Bar, P. & Legget, J. (eds), Association for the Advancement of Computing in Education, Honolulu, Hawaii, Oct 24-30, pp 592-597
David Jones, Sharonn, Stewart, Leonie Power, Patterns: using proven experience to develop online learning, Proceedings of ASCILITE’99, Responding to Diversity, Brisbane: QUT, pp 155-162