The reusability paradox. How can reusability be bad?
When first presented with this concept last year, I must admit I really did struggle with it. As a techhie, every fibre of my being compels me to focus on reuse. Hence, the paradox. After some weeks of struggling with the reusability paradox, it did start to make some sense, emphasis on ‘some’.
I have recently revisited this concept, both in discussion with my (to be) PhD supervisor, but also in my day-to-day work as an Educational Developer/Lecturer/Educational Technologist. My revisit has prompted this blog post as a way of recording some connections I have made to real-world examples of this phenomenon, and how this impacts my thinking about technology (re)use. This thinking is far from crystalised.
David Wiley explains the reusability paradox in the context of reusable learning objects, and more broadly, the open content movement. When this concept was initially presented to me, it was already positioned in terms of technology. I find it easier to start with the original context in learning design.
What is the reusability paradox?
David explains it quite succinctly as:
A content module’s stand-alone pedagogical effectiveness is inversely proportional to its reusability.
He explains that the more contextualised a learning object is made, the more meaningful it becomes to that context. However, it also means the learning object becomes less reusable to other contexts. We have a trade-off situation – effectiveness (in learning) vs. efficiency (in scalability). David concludes:
It turns out that reusability and pedagogical effectiveness are completely orthogonal to each other. Therefore, pedagogical effectiveness and potential for reuse are completely at odds with one another, unless the end user is permitted to edit the learning object. The application of an open license to a learning object resolves the paradox.
I don’t think an open licence alone will resolve the paradox, but that is a discussion for another post.
The reusability paradox in the wild
So enough of abstract concepts – how does the reusability paradox play out in the wild and in other ways besides learning objects?
I often see the reusability paradox when working with lecturers – conceptually the same as David Wiley explains, but at a higher level. My particular experience relates to the contention of reusing units of study between different awards/degrees. This is pretty typical in the STEM areas – in my institution we refer to them as service courses (units). I work with a science school, and a key foundation unit of study taught from the school is anatomy and physiology. There would be a dozen or more degrees that require students to have a sound knowledge in this area.
Conventional management wisdom seeks to reuse anatomy and physiology units for health related-degrees. This is efficient use of resources, right? And “why re-invent the wheel?”
But before I explore those questions, let’s first take a step back for a moment.
The key criteria for reuse is applicability to other contexts. If there is sufficient overlap or congruence with another context, then a reusability factor could be considered high, thus worthy of reuse. Learning is very contextual, particularly when you factor, as David does, the underpinning of constructivist learning theory. Learners construct new knowledge, upon their own existing knowledge. This is very individualised, and based on each learner’s past experiences, and ways of thinking.
Learning designers have some tricks to help deal with such diversity, such as researching your cohort, conducting a needs analysis, and ultimately categorising learners and focusing on the majority. Clearly, this is flawed – but this is how massification of education works. For instance, if you are preparing a unit of study for nursing students, then you can make some reasonable assumptions about those students motivations (i.e. they want to become a nurse); their prior formal learning (i.e. previous units studied within a structured nursing curriculum); and even down to smaller groups such as pathways to study (i.e. were they enrolled nurses – ENs or school-leavers). These assumptions of course aren’t always correct. Nevertheless, the key point is that this unit of study is reused by all nursing students studying for the Bachelor of Nursing degree. A more or less reasonable trade-off between effectiveness and efficiency.
So let’s return to the example of an anatomy and physiology unit of study. In this instance, we see different discipline areas, albeit health related, attempting to reuse a unit of study. Despite all being health related, a paramedic student’s needs aren’t the same as physiotherapy students’, or medical science students’. And while some disciplines hail from within the same school, others disciplines are situated elsewhere within the organisational structure. Now, consider the diversity of the cohort.
So to cope with this type of diversity, I typically see three approaches:
- Make the unit of study as abstract (decontextualised) as possible making no assumptions about learners or their backgrounds, and “teach the facts”.
- Design the unit to cope with the highest represented context (i.e. the discipline with the most students).
- Design the unit of study to address multiple contexts, in an attempt to make it meaningful to multiple disciplinary groups.
In other words, make it meaningful for no-one; make it meaningful to the biggest group, and nobody else; or, try to make it meaningful for everyone.
Approach 1 is obviously ineffective, especially considering constructivist thinking. You end up with students asking “why do I need to know this?”, or “that course was so dry and boring.”
Approach 2 while not quite as flawed as approach 1, can be less than ideal. Particularly when the highest represented group is small compared to the entire group. In such cases, the other groups feel marginalised, “I want to be a Paramed, not a Physiotherapist.”
Approach 3 can also ineffective because you can end up with a study unit that is incredibly complex. This group of students does this, that group does that. As the lecturer, you have to manage the mixture. The students too can become confused about requirements. You can also run into “equity” type policy constraints, such as “all students must do the same assessments.” This is an important point. If you end up with such complexity, you really have to ask the question, “why not just have separate units of study?”
But solving this challenge isn’t the focus of my blog post.
The Reusability Paradox as it Applies to Education Technology
So does the concept translate to technology? Yes it does! And similar issues arise as a result.
Recall the three approaches I see people use to deal with the challenges of reuse for multiple contexts?
- Make abstract
- Contextualise for the largest group
- Contextualise in multiple ways for multiple groups
Let’s consider Approach 1
David Wiley says of the reusability paradox:
The purpose of learning objects and their reality seem to be at odds with one another. On the one hand, the smaller designers create their learning objects, the more reusable those objects will be. On the other hand, the smaller learning objects are, the more likely it is that only humans will be able to assemble them into meaningful instruction.
I think this statement has some “translatability” to an education technology context as:
On the one hand, the smaller developers create their learning technology tools (e.g. programming libraries rather than complete systems), the more reusable those tools will be. On the other hand, the smaller learning technology tools are, the more likely it is that only developers (and not designers) will be able to assemble them into functional learning technologies.
To make learning objects maximally reusable, learning objects should contain as little context as possible.
To remove context is to make something more abstract – to take away intrinsic meaning or specific function. Indeed this makes things more reusable, it also requires re-contextualisation. In the context of technology, abstraction leads to dependence on the developer.
Let’s skip to approach 3. With this approach, we end up with technology that attempts to do everything for everyone. These technologies become so complicated to use, that people simply don’t use them. My favourite example of approach 3 is the Moodle Workshop activity which is a “powerful peer assessment activity”. I consider myself to have a reasonable grasp of technology, and yet after 45 minutes of tinkering with the workshop activity in Moodle, I gave up. I have only seen 1 person at my institution use it. It has so many options, too many options, because it tries to account for all the different ways one might attempt to embed peer assessment into their course.
So what about approach 2? We reuse a learning technology without change – meaning it is focused on the majority of requirements (however that might be determined). This is typical of COTS (commercial off the shelf) solutions. This inevitably leads to functional gaps – “the system does this, but I want to do that.” If the gap is substantial, it can lead to workarounds.
Does technology need to be reusable?
This is where I struggled last year with the reusability paradox. If you can’t reuse a technology, then isn’t that a serious limitation? Management are constantly looking to replicate successes – “This worked so well, so how can we use this in other areas?”
When I am creating/adapting/augmenting technology for others, I have to demonstrate “bang for buck” in terms of my time invested. Does what I create, at least have to pay for itself in affordances? I normally look for economies of scale, and the obvious way is through reuse – it is usable by X number of people. Management/decision-makers get this – easy. However, technology can offer other economies. For instance, depending on the technology, it may instead allow a specific group of people to do something much better, quicker, cheaper, or if its very innovative, something they couldn’t do before. But that something might be very specific, so specific that it isn’t very reusable, and limited to a small audience. Yet, if it still yields a net gain, is that bad?
What if a technology is so specific, it’s designed for just one person – yourself?
At some stage in our lives, we have all had to engage with some form of workaround to get from A to B. Not just in terms of technology but life in general.
If you create a workaround, does it need to be reusable? Perhaps not. But what if you want it to be? How can you go about it?
This is where my time (and thinking) ends for now.
Damo, Much of interest here. A couple of pointers/thoughts.
I think Wiley’s four choices of what to do has some interesting application here. i.e. one approach is that when developing something (content or technology) that aim shouldn’t be too make it abstract and reusable. Instead, make it suit the context, but use choice #4 “allow and enable for contextual modification of the learning object”.
Give other people (both teachers and learners) the ability to modify the object/content/technology. Make it as easy as possible and use technology to scaffold the process. Mike Caufield’s work and thoughts on federation speak to this in terms of content. Allowing people to see what others are writing, adopt and adapt that, and allow others to see what you are adopting and adapting.
In terms of the type of problems you cite around the STEM subjects, I wonder whether Caufield’s thinking around choral explanations might offer a way forward.
But back to technology, what about a federated approach to small, loosely joined tools. Take the UNIX command line approach, combine it with APIs, client-side scripting, and federated version control/tracking/management to allow and support choice #4? Not everyone needs to be able to do the development, but allowing them to see what others are doing and make choices to adopt that.