Re: Agenda for W3C Most Important Priorities [Education] call, 9 January

Dear All,
My thoughts, inline, on the education priorities.

On January 8, 2015, Marcos Caceres <marcos@marcosc.com> wrote:
> 
> 
> 
> On January 6, 2015 at 11:32:42 PM, Jeff Jaffe (jeff@w3.org) wrote:
> > > 2. Online Learning Experiences
> > > 
> > This subjects would include questions like :
> > What is a learning object ?
> > How to propose an active reading great experience vs passive
> > reading (Digipub) ?
> > 

I think we need to be careful not to get into prescribing what good educational software looks like. In my opinion, very little has yet been developed - but the way this is going to happen is by creating a platform that allows innovation, and not by prescribing best practices, which will be based on a poor evidence base.

I agree that "what is a learning object" is a good question. I think the answer should be framed not in terms of what it *is* but in terms of how it relates to the rest of the ecosystem; in particular in terms of how a learning object is launched and how it is closed. Such a definition would suppose that the term was being used as an analogy for a programming object, and would not include passive assets such as pictures. Define the terms of integration and interoperabilty and leave developers the freedom to create innovative solutions to fit into that framework.

The consequence of active learning experiences will be outcome data (and probably complex launch parameters and/or asynchronous communications). The ability to share data between instructional activities and learning management systems is in my view the critical piece in the development of useful ed-tech.
> > 
> It would be good to have examples and definitions of active VS passive learning. In particular, I would be extremely interested in seeing "exemplars" of active+digital learning experiences (i.e., apps or ebooks that have been shown scientifically to provide superior learning experiences to passive content) and how they compare with passive learning experiences.
> 
> This would allow people creating the technology on which these active systems run to see what functionality is missing on the Web. I assume Kahn Academy is one platform that provides such an experience - noting that, at least in iOS (iPhone), one can only passively watch videos, while on the website one can actually do active exercises.
> 
> > Can we import pedagogical nature of learning in openweb platform
> > ?
What do we mean by "pedagogy". My straw man proposal would be "the design and sequencing of learning activities and the management of feedback". As per my first comment, the design of learning activities is out of scope for this group - but the development of interoperable specifications for the sequencing of learning activities is in scope. Previous attempts have been IMS Simple Sequencing and Learning Design, both of which failed. Part of the problem, as identified by extensive discussions at LETSI between 2008 and 2010 was the poor definition of the bricks in the wall - i.e. the learning objects. Define what is a brick - then consider how you might use your bricks to build walls. Then let the wider community get baking.


> This assumes that the Web can't already do this. This needs to be rephrased as:
> 
> Does the web platform impose limitations on the pedagogical nature of learning? If so, how?
> 
I would say that the barriers - or perhaps the missing pieces (broadly speaking, runtime interoperability of structured data) - are common to all platforms. But if a solution can come out of the web, so much the better.
> 
> Additionally, how do other platforms compare? Which platform does better and why? This can include paper or a combination of paper and digital forms etc.
> 
> > How to manage links between teacher and pupils ? Links in the Education
> > Community ?
> > 
> This would indeed be interesting... again, in particular if hyperlinking, or websites and social networking sites like Facebook, are somehow insufficient.
> 
> In particular, I would be interested to see examples where using HTML-based systems did not serve the purposes of educators and they found themselves having to create native applications.
> 
Again, I would say that there is no intrinsic deficiency in HTML-based systems, on their own or in combination with native applications. The deficiency is in the absence of common data-sharing protocols that allow the accumulation of a critical mass of applications, combining instructional learning objects with pedagogical management tools.
> 
> For instance, Apple's App Store has a huge education section where one can find things like "BlackBoard Mobile" (which was, and probably remains, traditionally a HTML-based learning platform). Kahn Academy is also now a native application.
> 
> There is also native applications for creating pedagogical content, like "Explain Everything".
> 
> With regards to connecting parents/student/teachers, see the QSchools app created by the Australian Government.
> <http://deta.qld.gov.au/about/app/qschools-app.html>
> 
> (It just takes web data and puts into a native app context).
> 
> 
> 
> > We should also cover the difficult subject of cognitive mapping.
I don't think we do. Let developers implement the theory (much of which in education is as yet unproven) in innovative ways that will produce evidence of the truth of the underlying theory. We should not try to tell them how to do that - but rather create a platform from which they can do so in unexpected ways.
> > 
> > 
> 
> Ok, but so long as it relates to the Web platform somehow.
> 
> > ----
> > 3. MOOCs
> > 
> > As one of the current possible online learning experience, this
> > is a subject to cover.
> > 
I think we should be cautious about spending too much time on MOOCs, which in my view have shown little evidence of effectiveness and, perhaps more seriously, are poorly defined. Most applications of MOOCs are no longer "massive". Perhaps what would be useful would be to describe exactly what pedagogies MOOCs actually use - generally a combination of digital exposition and peer-to-peer social networking - and then consider the implications of those pedagogies. In other words, look at the MOOC phenomenon at the technical/pedagogical level, not at the level of commerce and marketing.
> > 
> > ----
> > 4. Collection of Data and Privacy
> > 
> > Web Technologies will be used for Education purpose. We must
> > be more than serious on that or our Kids data could be "stolen".
> > 
> Again, this needs to look at the limitations of what the platform can provide (e.g., with respect to TLS). However, encrypted client-side storage might be useful here.
> 
> > This subject is linked to MOOCs and
> > adaptative learning matters. There is a serious need for standardization
> > here.
> > 
> Be nice to see if there are a list of abuse cases that have occurred that we could learn from. What was stolen, how, and from who?
> 
> > May be we could need a "data handling description language"
> > allowing the codification of consensual agreements, in short
> > enabling students to agree to reasonable data sharing procedures.
> > 
> Maybe, but often these kind of agreements are one way: like EULAs, they leave the end-user (in this case, students) with only a [x] check box to click and with no way to negotiate terms or any legal leg to really stand on.
> 

The "data handling description language" was my suggestion. Echoing my position on application development, I don't think it is our job to second guess the legally desirable "balance of power" between learners and education providers. Formal education is, after all, compulsory in most of the world and knowledge of learners' performance and capability is essential to teaching. In the context of a prescriptive education system, data handling rules may be prescribed by governments. In a market system, students may have freedom to choose their education supplier based on their approach to data handling, inter alia. All we can do is to ensure that the behaviours can be described accurately, whether in the context of government regulation, consensual standardisation of best practice, or consent forms offered at the level of individual organisations. Clear, machine-readable description will support automatic comparison and automatic application of rules by management software systems.
> 
> What would be nice is to have better assurances on the client side about what things only students are able to access (affording students more privacy and protection from teachers, parents, system admins, or anyone else they might not want looking at their stuff).
> 
This assumes that the student has the authority to control what the teacher sees of the student's learning data. I don't think such an assumption would reflect practice in the great majority of formal education systems. Nor should it, in my opinion - but the more important point is that we do not have any standing to tell anyone what the rules should be.

Looking forward to the discussion tomorrow.

Crispin.