13 April 2012

week 2 technological challenges

In the post for week 1 and during the ensuing online conference meeting, it was discussed how formal and non-formal learning are different. Also, it was pointed out how networked learning would offer a learning environment that is suited for non-formal learning. After all, networked learning environments are designed with the express purpose to serve non-formal learners. To be sure, this is not to say that other solutions may not exist, but networked learning is a sure candidate.

Environments for networked learning (learning networks) are online social networks. For such networks to flourish and function, tools are needed, that is technological artefacts. At the very simplest level, such technologies are a website with a web address (URL) that one can go to to meet others, one's social network partners. At a more sophisticated level, such tools should foster learning. They could for example induce a network structure which is best for learning, they could allow people to get to know each other, they could help to get answers from the network on content related questions, they could provide means for assessing oneself, etc. These tools may exist already, may consist of existing tools transformed to serve educational purposes, or they could have been custom-built. What tools to use and how to use them is not an easy matter to decide on.

Of personal and managed learning networks

As described in the Design of Learning Networks from an organizational perspective - Chapter 8 from Sloep et al., to be retrieved from the private Mendeley group - one should distinguish between two kinds of learning networks. Managed Learning Networks (MLN) are created to serve some organization's goal, unlike Personal (private) Learning Networks (PLN) which only serve the intersection (average, if you like) of the various individual goals that the network participants have. MLNs are sponsored by an organization, which sets up a network, for instance because it wants its personnel to share the knowledge they individually have with each other, perhaps even use it to create new, to the company strategic bits of knowledge. Organizations exist because they are a means to work more productively or more efficiently than a collection of individuals could. Organizations are successful because its products (goods or services) are better produced by dividing the labour of their production over various people each with their own special area of expertise. It thus makes eminent sense that organizations want to optimize this process of producing through a division of labour, by making sure people learn from each other. And therefore, organizations will tend to coerce their employees into participating in the network they have set up, they will spend money on setting up the network and on making it successful, for instance by telling employees what to do in the network. The second set of use cases discussed in Chapter 1 of Sloep et al. illustrate this quite well.

PLNs are different from MLNs in many ways. First, they exist because there are people who share some broad interest, not because some management wants them to. The first set of use case of Chapter 1 of Sloep et al.  describes them best. As indicated, there is no organization with an interest of its own that can act as a sponsor of the network. There are only the users and their interests. It is they and they alone who should see benefit in networked interaction. There is no greater good, although there will be a division of labour based on the various areas of expertise that the network participants represent. But there is no grand scheme, no organizational structure that has been set up to produce some service or product more effectively or more efficiently. There are only the interests of the individual users. This of course also implies that PLNs arise from the bottom up, solely on the initiative of the users; not, as in MLNs from the top-down, on the initiative of the management. PLNs have no official management, although users may adopt managerial (admin) roles. If they do so, it happens purely because of their merits. (Note: it is quite illuminating to look at the open source software movement and how in there networks emerge; a good overview is offered by Steven Weber, 2004:  The Success of Open Source).


There is another important difference between MLNs and PLNs that should be highlighted. From the perspective of the users' motivation to actively participate in a network, PLN users do so because they themselves want to. They are predominantly if not purely intrinsically motivated. If they loose their motivation, they will no longer participate and their is nobody who will, indeed could force them to do otherwise. In MLNs however, users may be motivated intrinsically, but their main motivation will be extrinsic, 'because my boss wants me too', 'because it is part of my job'. Either kind of motivation has its own problems that one should be aware of when designing tools for a network. Intrinsic motivation is powerful, but costly. People only have so much of it to spend and are therefore economical with it. This implies that PLNs are difficult to get started as in the beginning there is so little to prove that one should be motivated to participate (a kind of cold start problem). MLNs with their organizational sponsors do not have this start up issue. They are typically launched with a bang (party, event) and, certainly in the beginning, carefully moderated to generate traffic between users. This is often accompanied by the obligation 'to go there and do things'. The obligation is either translated in incentives to participate (bonus points, a good assessment, etc.) or in punishment (negative points, a negative assessment by management). The problem with this is that the networks stops functioning when moderation is withdrawn ('the network should now function by itself'). Also, people will tend to 'play the system' to get the rewards or avoid punishment without actually contributing to its greater goal (think of people who buy 'levels' in online games to move up the respect ladder quickly without having put in the effort). This situation is inherent in the two kinds of networks, one can only be aware of it and, for a PLN, try to avoid the cold start problem and, for an MLN, strive for more intrinsically motivated participants. The right choice of tools is also critical for this.

Technologies, tools and services

There is no point in going over a long list of existing tools as these tools will come and go as well as change and evolve rapidly. However, function categories of tools are more persistent. The following list of five only covers the most obvious needs that learning network tooling should address.

  1. content service - a service that points to written resources, such as texts, papers, presentations, collections of links, etc.
  2. peer-support service - a service that connects a help seeker with a potential help provider; if networks become large (from about a hundred people up) one cannot know everyone anymore and participants have to rely on software that matches their needs to the expertise and capabilities of others in the network; peer support covers content-related questions (not the trivial who, where, when, but the non-trivial why and how questions), queries for mentoring, for moral support, etc.
  3. coalition-formation service - a service that helps participants set up teams of collaborators, to write a paper, to investigate an issue, to study a topic, etc.
  4. online identity service - a service that collects online user data and allows recommender algorithms to do their work
  5. portfolio service  - a service that allows users to collect their achievements, show case them and perhaps even have them certified. 

Such services have to be implemented in concrete tools. To name but a few examples of content services, Wikipedia is a collection of lemmas as in an encyclopedia, Merlot is a repository of in a learning objects, Mendeley and dSpace.ou.nl are collections of scientific papers, Delicious is a collection of presentations, iTunesU is a collection of usually rich lectures, etc.  Peer support could be provided through Google's Aardvark service (acquired by Google in 2010, subsequently pulled from the market; cf. for details on its functioning Horowitz & Kamvar, 2012: Searching the Village:Models and Methods for Social Search), through LinkedIn or Facebook groups. I know of no publicly available coalition formation services, but that could only betray my ignorance. However, in the course of our research we have built a dedicated peer-support tool (cf Van Rosmalen, 2008: Supporting the tutor in the design and support of adaptive e-learning; SIKS Dissertation Series No. 2008-07) and coalition formation tools are subject of active research (cf. Sie et al., 2011: What’s in it for me? Recommendation of Peers in Networked Innovation). Online identity services do not exist yet, unless one counts Google Profiles as such (They are tricky, see two of my blog posts on privacy and on the problem). Portfolio services have been around for quite some time, but they have been usurped by formal education. An interesting new development are Mozilla's Open Badges.


The present discussion and the Chapter referred to, should all act as a starting point for Friday's seminar and, later on, the paper that is to be written. The following questions may be seen as starting points for the discussion.

  1. Are there other services than the four listed in the above?
  2. May Facebook, LinkedIn or indeed Mendeley or Academia.eu be classified as learning networks?
  3. Do you know of existing, publicly available tools that could be used to provide valuable services to a learning network?
  4. If Facebook were to be used as the infrastructure for a learning network, does its goal of making money through advertisements stand in the way of its function of a learning network. 
  5. How do portfolios differ from Mozilla's open badges?
  6. Although our group is too small to count as a learning network (if anything, it is a community of learners), yet consider the tools that we have been using. Are they adequate? If not, what should be replaced, what is missing?
Note: the post has been updated, with the addition of links and a few changes in the wording