Thursday, August 22, 2019

Creating an environment for Learning Technology development.

In a recent post "A supportive environment for developing learning technology" I discussed two Universities with specialist departments that allow them to develop cutting edge learning technology that addresses their needs. At the end of that post I said I'd suggest ideas of how other Universities without such specialist departments might get the same benefits, without substantial costs. The key is recognising the factors that are actually important, and then finding a way to provide them without committing to the cost of a full specialist department. However, slotting learning technology into a business systems department does not work.

Learning Technology development requires embracing risk, experimentation, and occasional failure - so needs a different approach from business systems. Learning technology will always be used in ways that are unpredictable, and on a multitude of different devices, and needs to fit the users requirements. Learning Technology development is a complex interdisciplinary process that involves research and scholarship. Business system development is very different: business systems are typically static and inflexible - no one is doing research on improving them, and every user has almost identical needs. More importantly, the end user is trained (and forced) to use them correctly for the designed purpose. Because they have very constrained requirements, and long periods between updates they can be run with very traditional, bureaucratic processes and separation of teams. External cooperation, essential for open source Learning Technology, is also largely alien to the business systems world.

This means that the business system development section of a central IT services department is not the right place: the required culture is quite different.

The core things that are needed for Learning Technology development are:
  1. Expert developers with the right knowledge and skill sets.
  2. A quiet productivity space for the developers.
  3. Close communication between the experts and decision makers.
  4. Sufficient access to suitable servers for development, testing and piloting
  5. Close collaboration between developers and teachers who will be end users.
  6. Access to a wide range of devices for testing.
Looking at each of these in turn:

1. Developers

A Learning Technology Developer is a programmer with specialist knowledge. As a programmer they need to be a generalist, able to work in a variety of languages, but they should also be familiar with the key specialist knowledge for learning technology: pedagogy, interoperability standards,  accessibility, data protection etc. Programmers need to be good at self-directed learning, so much of the specialist knowledge could be learnt on the job given suitable guidance, but if starting with a team of one, someone with the right experience and expertise is essential. The ideal candidate would have a track record of both educational software and academic publications, and experience of working with IMS and other standards. Teaching experience would also be a nice extra, though maybe not essential.

2. Space 

Space is an issue at the University where I work, and I suspect at most others, however space at a University follows a variant of Parkinson's law - "Work expands to fill the time available for it's completion". In my experience in each University department, space requirements expand to fill the available space, so the issue is not so much a shortage of space, as a shortage of space in the right places. The needs of productive programmers are well researched and documented (DeMarco & Lister, 2017)*, though largely ignored. For a university they shouldn't be a huge problem as they're similar to the needs of a lecturer - a quiet private office, large enough to have one-to-one meetings and with a large desk. Importantly, there is no need for close co-location with colleagues unless they're working on the same project, so this office can be anywhere on campus, as long as it's quiet. (The usual argument for co-location, being available to just quickly chat, is a major productivity drain for programmers.)

* Anyone making decisions about where to locate programmers (or other knowledge workers) should read Peaopleware part II, "The office Environment" first. The impact on productivity of inappropriate office space is quite staggering.

3. Communication  with decision makers

One of the factors that is makes the Connected Intelligence Centre at  UTS successful is that the head of that department reports directly to the DVC Education, meaning that the experts and decision makers have close communication. In my hypothetical department-less version the same would have to be achieved by having the developers line managed by someone in direct regular communication with the VP/DVC for Education. A related aspect of CIC is that it is also a community of experts, and the benefits of that may be harder to emulate in my department-less alternative, but maybe my solution to points 5 & 6 also solves that.

4. Suitable servers

This one should be easy - just give the developers a prepaid credit card tied to Amazon EC2 or Microsoft Azure, and provide a quick way of allocating DNS names to pilots. While it should be easy, in the past when I've had the research income to run Amazon servers with a pre-paid card finance have blocked it, but I expect a supportive VP could easily make this happen. 

5 & 6. Collaboration with teachers, and access to test devices

Again, easy in some ways, but in my experience getting the right people together happens more by chance than design. However, I've got an idea about this that also solves another problem.

I recently read a couple of papers (Porter & Graham 2016, Porter et. al. 2016)  that were looking at what staff felt was needed to support them adopting blended learning. For those that are confident with technology it seems that the main perceived requirement is evidence (from peers) that the new techniques work, but the less confident staff would also like tutorials and by one-on-one assistance with the technology. From my own experience I also know that sometimes even the most confident early adopters also sometimes want access to more specialist equipment for content development, a wide range of devices for testing, or just a place away from their office to avoid disturbances.

My idea to solve this, and also create the serendipitous meetings needed to get good collaboration going, is to have a drop in content development and testing suite, with a Learning Technologist available to provide one-to-one assistance. Group training could also take place in the same location. Importantly, the Learning Technologists from around the University would take turns providing support, so all departments would feel ownership, and the Learning Technology Developers would also take turns in this role so that they met and talked to teachers.

Ideally this would be dedicated space, with a range of workstations, separate sound proof booths for recording videos and audio, and more specialist equipment like a lightboard, microphones and cameras, and good quality graphics tablets. However that gets back to the whole finding space at a University issue, as that would require a bit more than an empty office in an obscure quiet corner. A temporary solution might be to allocate a teaching cluster to this purpose outside normal teaching times, and if that was successful the argument could be made for dedicated space. 


DeMarco, Tom, and Timothy R. Lister. Peopleware: Productive Projects and Teams. Third edition. Upper Saddle River, NJ: Addison-Wesley, 2013.
Porter, Wendy W., and Charles R. Graham. ‘Institutional Drivers and Barriers to Faculty Adoption of Blended Learning in Higher Education: Drivers and Barriers to Blended Learning Adoption’. British Journal of Educational Technology 47, no. 4 (July 2016): 748–62.

Porter, Wendy W., Charles R. Graham, Robert G. Bodily, and Daniel S. Sandberg. ‘A Qualitative Analysis of Institutional Drivers and Barriers to Blended Learning Adoption in Higher Education’. The Internet and Higher Education 28 (January 2016): 17–27.