The Ed Techie

Here comes everything

I've been reading Clay Shirky's Here Comes Everybody. In it he argues that organisations have costs that means they struggle to compete with masses because the masses can afford to have lots of failures, because the cost of failure is low, and the ease of organising is now drastically reduced. I made a similar argument in the Future of Content, by using natural selection as an analogy. Natural selection can afford to make lots of mistakes because it has thousands of individuals and millions of years to experiment over. An individual designer cannot afford to have so many dead-ends. But when it comes to producing complexity, this massively distributed process wins.

Shirky's argument is that social communication tools have lowered the threshold for organising. He gives an example of organising photos from an event - previously this would have required someone to organise it, to get in touch with all possible photographers, to collect and publish their photos. Now all that's required is that people stick them up on Flickr, and use a tag, independently of each other. The cost of organisation has collapsed overnight.

Things such as eduglu, sociallearn, loosely coupled teaching apps, and PLEs have been much on mind recently, so when reading Shirky's book I thought some of the same arguments he makes for organising people could be made for technologies. The 'cost' of organising, or integrating, applications used to be high, but through approaches such as widgets, RSS, web services, etc this cost has drastically reduced.

To make my point here are a couple of Shirky quotes which I'll then rephrase for technologies.

"By making it easier for groups to self-assemble and for individuals to contribute to group effort without requiring formal management, these tools have radically altered the old limits on the size, sophistication, and scope of unsupervised effort"

becomes:

"By making it easier for tools to (self) assemble and for applications to contribute to the environment without requiring integration, these approaches have radically altered the old limits on the size, sophistication, and scope of any individual to create their own environment"

And this Shirky statement:

"because the minimum costs of being an organisation in the first place are relatively high, certain activities may have some value but not enough to make them worth pursuing in any organised way. New social tools are altering this equation by lowering the costs of coordinating group action."

Can be recast as:

"because the minimum costs of being an integrated environment in the first place are relatively high, certain applications may have some value but not enough to make them worth pursuing in any organised way. New data techniques are altering this equation by lowering the costs of integrating all applications."

So, it isn't a case of here comes everybody but here comes everything. What's an educational tools? Whatever you want it to be.

A PLE - VLE continuum

There's been a bit of a PLE flurry of posts, generated by some Twitter discussion. Chris Lott posted that he couldn't see why people think you can't teach PLEs, or about PLEs. Scott Leslie argues that some don't like the acronym:

It’s personal, not monolithic” complaint with the term, which I get and agree with. My response is not to defend the “PLE” acronym but instead just say if it bothers you, come up with a different one, or don’t use a moniker at all, but more importantly, model model model it for the 95% of learners (and teachers) who are drowning in the tsunami of information and choosing to turn their backs rather than learning to surf

Meanwhile D'Arcy Norman maintains (with the aid of some cool diagrams) that one person's PLE will differ from another's, and it is the people that are important:

the exact technologies that are in use at any particular point in time don’t matter as much as the fact that it is people being connected through them. Tools come and go constantly, and the only constant is that the people are the important part of the equation.

The debate on twitter seems to have been about whether one can usefully talk about PLEs since, by their nature, everyone's is different. There is no The PLE.

I'd like to argue that the term is okay, it has context (that of VLE), and also there is some benefit in 'teaching' PLEs. But maybe we need to nuance the term a bit, as we may be lumping things together. If in doubt about an acronym, let's create some more! So we have:

• VLE/LMS - a centralised system that gives a consistent user experience to everyone
• TLE - Teacher learner environment. This is along the lines of Scott Leslie's loosely coupled teaching applications. Less centralised than a VLE, the educator determines the range of tools, e.g. a blog with specific widgets, but all students use the same.
• DPLE - Default PLE. In this novice users (could be educators, students, employees, etc), are given a default set of applications to constitute their PLE, but they have the freedom to switch them out over time. A minor example might be my recent conversion to iGoogle from the standard Outlook provided services. Except the option and ease of switching would be stressed more. Imagine a default Netvibes page, which people would soon customise.
• PLE - the type of thing we ed techies have accrued over time, and continues to evolve. Work might be required on getting these apps to talk to each other, but really the people who operate at this end don't need much help.

This gives us a continuum of personalisation:

One thing to appreciate here is that there is no value judgement attached by the placing on the diagram - being further to the right isn't 'better'. These different approaches to technology will be suited to different contexts  and audiences. That is probably obvious to everyone, but was something of a mild epiphany for me - PLEs and VLEs are NOT in competition necessarily.

[Update - that last sentence originally had the 'not' missing (such a small word). I've added it in, but now I'm not sure - which is truer for you - they are in competition or they aren't?]

An audit on where stand with PLEs

1999 A.D. Learning
by donaldtheduckie

(via Scott Wilson)

Back in 2006 I posted some of my reservations about PLEs. At the time I thought these issues were insurmountable, and largely being ignored by the PLE advocates. Now, I'm more convinced of the possibility, and worth of a PLE, I thought it would be good to see where they stood against my original set of reservations.

This of course comes with a massive pre-question 'what do you mean by a PLE?' (as Alan Levine said in a twitter post, it seems to be a spoke diagram with lots of web 2.0 logos - what can he mean?). And maybe the notion has softened somewhat from a purely personal learning environment to a more distributed learning environment built on a range of applications. Some of these will be teacher selected e.g. 'we'll be using this wiki', and some user selected e.g. 'post to your own blog'. Anyway, here is my initial set of reservations, with some random indication of progress.

1. Support. My initial argument was that VLEs allow centralisation of support which becomes unmanageably complex if every user has different applications. While this is true, the need for support has diminished I feel because a) there is a greater level of familiarity out there (although I am not naive enough to think that even amongst the digital natives everyone is computer savvy) and b) the sort of third party tools available are easy to use and robust. People don't take courses in how to use Facebook after all.
   Progress: 50%
2. Quality assurance. My argument was that universities could not ensure a certain quality of provision with PLEs. I feel that given the superiority of many third party applications over institutional systems, the reverse may now be true. But it still remains that it is much more difficult with a PLE type world to point at something and say 'that's what student A is doing at time X'.
   Progress: 75%
3. Suitability. My argument here was that users may adopt or use the wrong technology. This brings us back to what we mean by a PLE, since it might not be the case that they are left completely to their own devices (pardon the pun). And in addition I don't think I envisaged the way widgets would come in to play, so your blog, Netvibes or iGoogle page is easily populated.
   Progress: 60%
4. Negotiation of activity. My argument here was that if everyone is using a different tool it means that every activity starts with some negotiation about what you will use, instead of just getting on with the task using the default VLE services. I still think there is something in this, but if we take the slightly looser interpretation of a PLE, then it may not mean that every task is negotiated, but rather that the educator has a choice of tools to choose from.
   Progress: 40%
   
5. Technological complexity. This is the big one. Since 2006 the rise of RSS as a common data standard, the use of widgets and the development of standards such as openID makes a lot of the potential complexity less of an issue. Look how easily a user can embed different tools, video content, feeds, etc. But at the moment the different applications are dumb to each other, and this is where work will focus over the next few years (through things like Social:Learn and eduglu).
   Progress: 50%

That's not bad progress so far I reckon. What we need now is to move away from the spoke diagrams to some real applications.

My personal work/leisure/learning environment

Following on from the previous post I was thinking about how I have accrued a set of technologies around me. This wasn't a deliberate policy - I didn't think "I need to construct a personal working environment. Here are the functions I need..." Rather I have come across them, experimented with some, rejected some and kept others. It was only when considering what tools I couldn't afford to lose that I realised I had 'constructed' something that might loosely be termed a working environment.

I've mapped it out in Compendium, and here is the image (click on it to open up in a new window):

I suspect my set of tools is smaller than a lot of people's. My criteria for inclusion were:

1. I use it regularly
2. It's outside of my formal work provision (no Outlook, Moodle, FirstClass, etc)
3. It is something I have input to, not just a resource e.g. Guardian online (YouTube and Flickr are a bit borderline here).

A few things to note about this:

1. The boundaries are blurred - it is simultaneously a work/leisure/learning environment
2. The accrual over time is a different approach from setting out with a feature specification, and is much more organic.
3. There is little integration between the services.
4. Most of the tools are very easy to use - apart from some knowledge base queries in Typepad and the instructions for creating a slidecast I don't think I've had to consult a manual or seek help for any of them.
5. Following on from 2) I didn't know I wanted some of these until I saw them. Take Slideshare for instance, I wasn't sitting there thinking 'I wish I could share my Powerpoint files easily', but when it came along I took to it instantly.
6. I can't really say why I've taken to some applications and not others. I could rationalise my choice, but ultimately it is a complex mix of personality, emotional response to the interface, perceived need, timing, motivation, etc. This does mean it is a personal collection.
7. There is too much Google in my life.

It would be good to find an easy means of sharing these, with clouds so we could see which tools are used regularly. I'm sure someone must have an idea for doing this (Cohere, the soon to be released web 2.0 extension to Compendium might be an idea).

PLEs and the institution

Following on from all our VLE discussion, this post from Scott Wilson caught my eye. He has another good diagram (I've borrowed his Future VLE one before - with acknowledgement Scott!). I have had my reservations about PLEs, and particularly how they interface with the institution. I like the idea of assembling your tools over time, but there are some real issues about how these interact with institutional systems (authentication in particular - are you a student on this course, what support do you have, how do we verify course records, etc). Scott goes some way to addressing these. I think a telling quote is:

Originally I thought that a personal system could manage all the variety of all connected institutions, but more realistically there does seem to be a real need for a more concrete coordination system sitting between the personal system and the enterprise, for handling jobs such as initial rendezvous and peer association.

Scott suggests a course coordination space where some of this is sorted out - maybe this is similar to my concept of the educator choosing loosely coupled systems. The danger is that maybe too much time gets spent in coordination for every course - which may make the default option of the VLE seem attractive. The response could be 'sure these other tools are better, but we spend the first two weeks of every course thrashing out what we are going to use. The VLE tools are good enough.'

Stringle implications

The last in this trilogy of posts...

Several things occurred to me when playing with Stringle yesterday. I'll try to elucidate them here:

Course design - increasingly I feel that writing course resources is a redundant activity. I think course design will move towards creating activities around as yet unknown content. For instance, we added in a feed for any content that is tagged elearning and evolution. Now it's a reasonable bet that this will contain some decent (and not so decent material). But it helps future-proof the course. I don't think you could quite rely on such resources but the ratio of educator derived material to external resources is surely going to decrease.

Why does HE develop tools - I know, I've commented on this before, but looking at the range of tools that are just out there, and the ease with which you can pull them in to a system made me wonder it all over again. I would like to propose Weller's rule of educational software development which goes something like 'if an educational establishment wants to develop a tool then by the time it has reached the functional requirements stage a better one will be freely available'. Allright, it's not a snappy one, but you get the idea.

PLEs are possible - I've been a bit of a critic of PLEs, partly because they seemed obsessed with clients and partly because they underestimated some of the challenges I felt. But I can see how a mid-ground between the 'it's all down to the user' and 'central control' models could be found, by providing default tools, and allowing users to pull in their preferred ones if they wish.

Openness is the new king - if the struggle to get Vista shipped hadn't convinced me that the proprietary model of software development was dead, then this has. I kept thinking of the time I had spent getting a course ready for the UK e-university, who spent somewhere between 10 and 18 million pounds on developing a VLE. It wasn't very good. Imagine what you could do with something like a Stringle approach for a tenth (even a hundredth) of that money.

Anyway that's enough on Stringle.

Stringle 2.0

Having shown you Stringle, I thought it'd be interesting to think what else it would need to be a usable tool.

Here's my thoughts:

Tools –

A good range of tools that users can select from. Tony's working on this, so you should have a range to select from and can simply drag them into your toolset and give them a label (thus avoiding all that del.icio.us and URL stuff).

Easy way of adding new tool - if you come across a tool then you want to be able to add it simply by clicking a button. If we were providing students with a university social bookmarking tool, then this could be done easily.

Toolset related to resource - I think this would be quite powerful. If you select a particular tool then it has some metadata associated with it which automatically loads in the appropriate tool(s). Thus if you're doing a timeline activity the only tool available is a timeline, if you're doing a collaborative activity then maybe a discussion board and IM tool are displayed.

Content

Easy add - as with tools, you don't really want people to mess around with OPML managers, so you need a neat way of adding content in.

Icons - next to the links in the grazr widget it would be good if you can identify different types of resources, e.g. activities had an icon or were styled a particular way.

Enterprise -

Authentication - you're going to need some way of allocating roles and resources to individuals, which means all the tedious but mission critical stuff around linking with student record systems, assigning roles, security, etc.

Tracking - this is essentially dumb integration. If you really want to do tracking you need to know what people are doing in each of the tools. This might be insoluble if you want to allow an open system.

Usability - it's no oil painting at the moment and definitely could benefit from some good interface design.

Of course, you may be asking yourself 'why bother? Why not just use Pageflakes?' (or something similar). There may be something in this, but what impresses me is the way a system can be configured quite easily. More on this in the next post.

Stringle - almost a web 2.0 PLE?

Tony Hirst visited me yesterday in Cardiff and we spent the day going through his String n Glue VLE (Stringle). I'll split up my posts on this, so in this one I'll do a quick users guide, then I'll look at what else it needs in another post, and finally some thoughts.

Before we get stuck in I ought to just say that there is a bit of clunkiness about this, and the immediate reaction might be 'most students are not going to do this.' Which is true, but a lot of this clunkiness could be removed with a bit of resource and programming (e.g. creating a button that automatically adds a resource). I'll come back to this in the next post, but for now, assume it could all be made easy.

You will need a del.icio.us account and an OPML manager (e.g. from http://www.opmlmanager.com/) in place before you start.

Go to Stringle at http://ouseful.open.ac.uk/stringle2.php

You'll see a list of tools across the top like Docs, whiteboard, etc. Clicking one of these will load the tool in to the Tools tab in the central frame.

One the left there is a list of content. Clicking on one of these will open the content in the Web tab in the centre (or you can add in a URL in the box at the top and it will display it).

And that is essentially it - you have your tools at the top and your content down the side. Both are configurable, so let's start with the tools.

Say I want to add in a scheduler tool, like Remember The Milk. You need to bookmark the site using del.icio.us (any social bookmarking tool could be used but for demo Tony has opted for the best known one). You need to tell Stringle it is a tool, and to do this you use the tag stringle:tool You can create clusters or categories of tools, which you can load in to Stringle at different times. So for example you might have a collection of collaborative tools. Lets say I group my Todo tool in with admin tools. I would then add in the del.icio.us tag by leaving a space and adding 'admin'. Now I want to tell it what text to put along the top in Stringle, so I add another tag, say 'stringle:Milk'.

Now I want to load my admin tools in to Stringle. To do this I add in some text after the URL - ?t=del.icio.us name/category name. So in my case I would add in ?t=mjweller/admin to give the url http://ouseful.open.ac.uk/stringle2.php?t=mjweller/admin

Okay I said it got a bit clunky, but this could all be automated. Now let's add some content in to the left hand panel, which comes courtesy of Grazr. To do this you'll need to go to your OPML manager. If you are in OPMLManager, then right click on the main folder and create a sub-folder. Let's say I'm creating a course on Evolution, so I'll add in a folder with this name. As I discover different resources I can add them in here using the add link function. These can be sites, rss feeds, podcasts, etc.

Once you've built up your collection of resources, you want to import it in to Stringle. This requires URL jiggery-pokery again. This time add ?oURL for OPML feed so in my case it would be ?o=http://www.opmlmanager.com/opml/mweller.opml giving the URL http://ouseful.open.ac.uk/stringle2.php?o=http://www.opmlmanager.com/opml/mweller

Now we can of course combine the two to give my toolset and my content -

http://ouseful.open.ac.uk/stringle2.php?o=http://www.opmlmanager.com/opml/mweller.opml&t=mjweller/admin

It may be a bit buggy still (especially in IE Tony tells me), but you get the idea. That's the boring stuff, I'll look at what else could be done to it in another post.

Outsourcing learning technologies

This article in the Economist Consumer technologies are invading corporate computing (via George Siemens) outlines how some universities are effectively outsourcing a lot of technical development, using the Google Apps for your domain bundle. The guy says that "Compared with the staid corporate-software industry, using these services is like “receiving technology from an advanced civilisation”, says Mr Sannier."

A couple of things on this - firstly there is often a delay between the rhetoric and the reality when new ideas come along. The whole service oriented approach was one of these. I found myself wisely telling everyone to think of everything as a service, without really being sure what this meant. Google have continued to demonstrate what this really means, and this is a good example. It's not about buying a big bit of kit for your IT guys to install, you simply say 'all our students get the Google services.' Secondly, it really does mark something of a watershed that universities are now formally recognising these consumer, real world apps and not insisting that somehow education is special and needs to develop its own tools. Thirdly, what sort of impact does this have on institutional practice? As a small example, at the OU we have struggled with a calendar project for a while (partly it isn't a calendar project but a multiple database problem), but at what point would it be effective to say 'hey, go with Google'. What the institution needs to concentrate on then is making data available in the right format. It's kind of analagous to educators needing to shift their focus from creating content to supporting understanding.

Stephen Downes' visit

Stephen Downes came to the OU today and gave a talk on PLEs. We had a chance to chat beforehand, and his talk was, as ever interesting and thought provoking. I felt that his vision of a PLE, although it steered clear of the client based talk I have seen in other ones, was very much based around an individual and their collection of resources. I didn't see much room for collaboration in it. I queried him on this and he responded that the resources should include community and peer resources and tools such as Skype would be included. I can see how this would be neat (and as I have blogged before, I like the netvibes personal portal approach to collecting tools) - however it still makes collaboration difficult if we all have different tools. If you have Skype and I have Trillian, how do we conduct a collaborative activity?

VLEs are often criticised because they provide a blanket provision, but the benefit of this is that you can get on with doing the task in hand, safe in the knowledge that everyone is using the same system. This is part of my problem with PLEs - the problem they report to solve is not so great as to require the massive technological and cultural change it would require. For instance in order for a PLE to be most effective we need interoperability between all tools, and for universities to become a thing of the past, or at least how we think of education to alter radically. I'm not sure either of these is likely to happen in the near future. What I do like about PLEs is that they force us to think about learners as being in the technology world, and not technological tabula rasa. I don't think higher education has even begun to grasp what it means to have the net generation entering their ranks. But I still have my doubts about PLEs - I think I may just be lacking that visionary thing.