I’ve spent a few hours reading and thinking about pedagogy and LARGE student numbers in anticipation of my upcoming work on a Commonwealth of Learning quality assessment team headed for University of South Africa (UNISA). For those not familiar with UNISA it is the world’s oldest, exclusively (single mode) distance education university and a mega-university (330,000 students).

My UNISA friend Paul Prinsloo referred me to a couple of recent works by Thomas Hulsmann. Thomas is a German academic who spent a few years post-retirement at UNISA. Hulsmann has an extensive publication record that focuses on cost effectiveness of various DE delivery. He has published a COL review (Hulsmann, 2016) The Impact of ICT on the Costs and Economics of Distance Education: A Review of the Literature in which he examines the costs of distance education (in many forms from correspondence to e-learning).

I was intrigued to find in his literature review the identification of the shift in my perspective from earlier days when Randy Garrison and I developed the Community of Inquiry model.  He notes my shift from a focus on “traditional” student-teacher interaction (constructivist paradigm) to are more pragmatic “any interaction can work” – as in my Interaction Equivalency Theorem. He even highlights the only time that Randy and I shared (in print) differences of opinion about  “big distance education“. Anyways, I am pleased to know that my PhD supervisor, colleague, Dean and co-author Randy Garrison and I remain good friends – even if we have diverged in pedagogical terms,

In this review Hulsmann revisits his earlier work identifying interaction as the key cost variable and identifies two distinct models of interaction. Type I is interaction with Information (or content – from books, to web sites, to simulations and everything in between). Type C is for Communications with humans (usually focused on expensive student- teacher interaction). Type I interaction is scalable in that variable costs do not rise with additional students. In contrast Type C communications costs with teachers increase with every student enrolled.

I then moved to a 2016 article Hulsmann wrote with Shabalala (Hülsmann & Shabalala, 2016) to look at cost issues associated with ten “signature courses” developed at UNISA.  These ten courses added formative feedback (time consuming for tutors) and attempted to support computer conference interactions amongst students.  By North American standards the course numbers (each tutor responsible for 3 classes of 60 students each) were very high and the workload for mostly part-time tutors to complete both conference moderation and marking far exceeded what they were paid for.  Hulsmann and Shabalala again use his Type I and Type C lens to show that interaction (of any type) is expensive and not always possible in educating very large numbers of students – at an affordable cost. They do note (and I heartedly agree) that the use of multiple-choice quiz questions, with feedback, could have provided the formative assessment without increasing tutor workload, as is done routinely in MOOCs.

They also hint, but don’t develop the possibility that students could provide their own network support.  The course designers were influenced by Heutagogy thinking and assumed (hoped??) that students would be able to become self-directed knowledge seekers and builders. However the primary technique used was reported as having the tutor pull back from active ‘teaching’. Not a challenging assignment for overworked tutors!

The signature courses had some success. The courses resulted in higher grades and higher completion than normal UNISA courses, but the sustainability of the model is questionable.

This discussion of interaction costs took me back to a chapter I published in 2008  (Anderson, 2008) Social Software technologies in distance education: Maximizing student freedoms. I wrote this chapter partially in response to Hulsmann’s model. I introduced an emergent type of interaction that I labelled as type S – for Social Interaction. It was entertaining to review this 12 year old work and my overview of the promise (without today’s perils) of social networking as a type of interaction that goes beyond Hulsman’s type C. Social networking thrives on many-to-many interactions and benefits from networked effects, crowd evaluations, persistence beyond a single course and the near ubiquity of relatively low cost availability- even in many developing countries.  I illustrated the place of Type S below.


In the nearly decade that I and Jon Dron developed a type S system (Athabasca Landing) we never really won the institutional (or many teachers’) support that we had hoped for, but the system persists today.

I remain convinced that we need to develop learning activities and the skill and motivations of both teachers and learners to exploit the potential of type S interaction. Equipping students with the tools and the skills to develop networks for both learning and personal support is of critical importance – not only for cost effective distance education, but also for effective life-long learning.


Anderson, T. (2008). Social Software technologies in Distance Education: Maximizing Learning Freedom. In Evans. T, M. Haughey, & D. Murphy (Eds.), International Handbook of Distance Learning (pp. 167–184). Bingley, UK: Emerald.

Hulsmann, T. (2016). The Impact of ICT on the Costs and Economics of Distance Education: A Review of the Literature. Retrieved from http://oasis.col.org/handle/11599/2047

Hülsmann, T., & Shabalala, L. (2016). Workload and interaction: Unisa’s signature courses – a design template for transitioning to online DE? Distance Education, 37(2), 224–236. https://doi.org/10.1080/01587919.2016.1191408