Monthly Archives: December 2007

Why is one not necessarily cheaper than two

Not unlike many institutions, at some stage in their use of e-learning, CQU finds itself in the perceived position of having two learning management systems (LMS):

  1. Webfuse – a locally produce system; and
  2. Blackboard version 6.3 (yes, that old)

It has taken long for some very logical people to make the observation, “One LMS would be cheaper than two”.

Surely, no-one could argue with that logic?

Well, I’m going to take a stab at it.

That basic argument is, “Yea, one is cheaper than two. But getting there is going to cost a lot of money.”

Then there’s this argument about efficiency from Stephen Downes.

Are they the same?

I live on a farm and we run a few beef cattle. Let’s imagine that we have two cars.

First, let’s imagine we have two Holden Commodores. If we decide we can do without the second one, obviously it’s going to be cheaper. Less running costs etc.

Let’s imagine we have one Commodore for taking the kids around and a cattle truck. Is it cheaper to have one? Not if we still have to transport cattle around.

The assumption at CQU is that Webfuse and Blackboard are both LMSes and they are both the same. I’ve argued in many places (including here) that these two systems are not the same. Some of the things which Webfuse does which Blackboard doesn’t include:

  • Web interface to CQU’s student records system.
  • Assignment extension system
  • Academic misconduct database
  • Results uploading system
  • Blog aggregation management
  • An online assignment submission system/gradebook that actually scales to large classes and integrates with various CQU information systems

If Webfuse were dropped, then there would be an additional cost, either through loss of functionality and/or the cost of replacing the functionality in another system.

You can’t simply do everything in one system in the other. We can’t take cattle to market in the Commodore.

What’s the cost of changing?

Which brings us to the other major problem with the “one would be cheaper” statement.

It’s only cheaper if you drop one (or replace both with another one). The act of doing this is not going to cost free. The lost functionality is one aspect of this argument.

At CQU there are large numbers of staff and students who use both systems. These people have spent time and energy getting to know the system, to know how to workaround their limitations, to figure out how they do what they want to do with these systems.

Moving to another system is going to take them time and energy. It is going to cost money. It may not show up in the bottom line of some budget spreadsheet, but it will cost money.

Plus, if the solution is to drop both existing systems and replace them with a new system. For example, let’s say Sakai. Then the technical and user support staff will also have to throw away all their skills and experience and learn the new system. More cost.

Then there is the cost that while all these folk are learning the new system they won’t actually be able to improve their learning and teaching. They’ll be expending all their limited time and energy learning to re-create what they did before in the new system, before being able to upskill.

Then there is the whole problem with large scale IT projects. They generally fail. Are generally never on-time, on budget or deliver what they promise. They cost and they waste money.

New approaches to curriculum design informed by complexity

For a long time I’ve felt disquiet about the “super-rational” approach to design and development. I’m including both for information systems development (my original discipline and interest) and curriculum design/e-learning (my current job).

The “super-rational” approach is based on the assumption that human beings are supremely rational and that an expert can analyse the problem, identify an optimum solution and implement it. With curriculum design this can be seen in days spent by an curriculum design expert, usually in collaboration with a subject matter expert, going through a rigorous process of analysis.

My two big problems with this approach are

  1. Resourcing – this approach simply doesn’t scale. At CQU we have 2 curriculum designers and in 2007 offered 1000+ courses. The process I see at the moment is taking weeks of time. If we say it takes 3 weeks (much shorter than it actually takes). It will take 28.8 years to to a full curriculum design on all CQU courses, assuming of course there aren’t new courses introduced in that time.
  2. Irrationality – human beings are not rational beings. We’re pattern matching machines. Everyone’s rationality is bounded. This means that even with an expert, the quality of the solution is going to limited by the rationality of the participants. Even more likely is that the solution will match what the subject matter expert or the curriculum designer have done before.

What’s the alternative

This rambling is informed/sparked by two recent bits of information that I’m attempting to digest

  1. Complexity and Education: Inquiries into Learning, Teaching and Research – a book by Davis and Sumara. The first line of the book description on Amazon is “This book explores the contributions, actual and potential, of complexity thinking to educational research and practice.”.
  2. Sense-making and strategy a presentation/MP3 from Dave Snowden.

What I currently am taking from this (and remember my rationality is as bounded as anyone’s and more than most) is that traditional approaches to design (both IS and CD) are based on the unquestioned assumptions of science/mathematics from Euclid and Newton. Assumptions which include that there is one right answer to find, that we’re rational, that these things can be analysed and the answer found.

It appears that these approaches do not take into account some of the interesting insights which complexity science can bring nor do they take into account how the pattern matching nature of the human brain. One possibly interesting alternate approach to take would be approaches more informed by complexity theory.

One (of many) problems for this idea is that the traditional analysis approach is so embedded into everyday practice. It’s just not a habit that has to be broken but a fundamental way of understanding the world. It’s not something that I’ve personally succeeded in doing and I’ve been trying for a while.

One description/comparison of the difference is encapsulated in this mp3 by Dave Snowden. It is an excerpt from a presentation he gave in Helsinki on sense-making and strategy. In the excerpt he describes two approaches to organising a child’s birthday party. One based on traditional top-down approaches and another based on complexity.

Which one do you think is most appropriate.

What’s the mean in reality?

Who knows? I don’t. I’m still grappling with what it means for curriculum design in a form that I’m happy with. But an initial attempt, drawing on the Snowden excerpt, follows

  • Complex system.
    The design process has to take place within a “complex system”. For curriculum design this might mean a largish number of academics all going through the process, perhaps supported by appropriately skilled staff. i.e. not a one on one between designer and subject matter expert.
  • Draw boundaries.
    Not sure what that might be in curriculum design. Perhaps a delineation about what is being discussed….more work needed.
  • Catalyst/attractor.
    A simple and attractive activity or task which generates interest and starts the components of the complex system talking and interacting. In the last week or so I’ve been sent a question checklist which is designed to be used by a group of three people. The questions are about the design of a unit of curriculum. The questions are answered verbally by a single person and others grade it. I’m assuming discussion ensues.
  • Magnifying the positive and killing off the negative
    There needs to be some process by which the good is spread and the bad gotten rid of.

The catalysts have to be simple tasks that attract people but generate emergent, positive activity. Not detailed, top-down analysis plans and processes. You don’t start with curriculum mapping and other big up front analysis.

It appears to be “nice” in the limited, poor abstract manner outlined. How all that works in practice is another matter all together.

Perhaps now it is time to refer to the literature, like any good researcher should.