Skip to content

Openness: Why learners should know about, and influence, how decisions are made about their learning

Earlier this week, I delivered a presentation to TEDxEdmonton on why openness and learning analytics are critical for rethinking the future of education. The theme of the event was on open source culture and whether the promises of open source have been oversold.

My argument is that openness has not been oversold and that increased openness (of content, teaching/learning, analytics, policy, data, and technology) is really the only path forward for reform. Systems can be closed and blackboxed only once they are working well and the context in which they exist is stable. When everything is in a state of flux, we need opportunities for ideas to collide, innovations to be shared, concepts to be rehashed and mashedup, and iterative improvements to occur. Education today – at all levels – faces the challenge of tremendous change and unstettledness. Rigid systems break in periods of flux.

Openness in response to change

The Xerox 9700 is the nemesis of, and catalyst for, the open source revolution.

Richard Stallman, in 1980, encountered a printer at his MIT lab that didn’t allow end-users access to source code. Until this point in the very young computer science world, openness and sharing was the norm. However, with vendors seeing opportunity for revenue by locking down software and denying end-users access to source code, the ethos of openness was challenged. Stallman, in response, launched the Free Software Foundation.

Stallman is a slightly prickly fellow (see his comments after Steve Jobs died). Stallman’s radical and uncompromising views on free software resulted in the formation of the more moderate open source software movement. This movement, which included Eric S. Raymond, author of the near-manifesto The Cathedral and the Bazaar, provided more palatable views of openness than Stallman’s FSF. As a result, the concepts behind the open source movement soon spread to other areas: business, government, and social systems. The principles of open source include openness, democracy, access, iterative improvements, and collaboration. Zuckerberg calls this the hacker way and details the principles as:

The Hacker Way is an approach to building that involves continuous improvement and iteration. Hackers believe that something can always be better, and that nothing is ever complete. They just have to go fix it — often in the face of people who say it’s impossible or are content with the status quo.

Hackers try to build the best services over the long term by quickly releasing and learning from smaller iterations rather than trying to get everything right all at once…

Hacking is also an inherently hands-on and active discipline. Instead of debating for days whether a new idea is possible or what the best way to build something is, hackers would rather just prototype something and see what works. There’s a hacker mantra that you’ll hear a lot around Facebook offices: “Code wins arguments.”

Hacker culture is also extremely open and meritocratic. Hackers believe that the best idea and implementation should always win — not the person who is best at lobbying for an idea or the person who manages the most people.

He then distills the hacker way down to five key principles:

  1. Focus on impact
  2. Move fast
  3. Be bold
  4. Be open
  5. Build social value

Reform, rhetoric, and change

In education, we have decades of reform rhetoric behind us. I have never heard someone say “the system is working”. There appears to be universal acknowledgement that the system is broken.

Classrooms were a wonderful technological invention. They enabled learning to scale so that education was not only the domain of society’s elites. Classrooms made it (economically) possible to educate all citizens. And it is a model that worked quite well.

(Un)fortunately things change. Technological advancement, coupled with rapid growth of information, global connectedness, and new opportunities for people to self-organized without a mediating organization, reveals the fatal flaw of classrooms: slow-developing knowledge can be captured and rendered as curriculum, then be taught, and then be assessed. Things breakdown when knowledge growth is explosive. Rapidly developing knowledge and context requires equally adaptive knowledge institutions. Today’s educational institutions serve a context that no longer exists and its (the institution’s) legacy is restricting innovation.

Digital networks antagonize planned information structures. Planned information structures like textbooks and courses simply can’t adapt quickly enough to incorporate network-speed information development. Instead of being the hub of the learning experiences, books, courses, and classrooms become something more like a node in part of a much broader (often global) network. The shift to networks is transformative in how a society organizes itself (see Wellman’s Little Boxes, Glocalization, and Networked Individualism – .pdf).

All of this reform talk, however, presents a challenge. How do we know what’s working? How do we know what our students are doing and how those actions contribute to their success as learners or to developing their sense of self-efficacy?

Social networks provide a natural way of self-evaluation and error-correction. But social structures have scale limitations – peak social, if you will.

Many of our interactions today produce a digital data trail that is ripe for analysis. Companies are collecting this data and making decisions, on our behalf (see here). We often don’t have recourse to alter these automatic data collection and analytic systems. Given the growing influence of algorithms in deciding what we will read, watch, each, and what we will do next, it’s increasing important that we are able to see what is happening behind the scenes with analytic tools.

Knowing how schools and universities are spinning the dials and levers of content and learning – an activity that ripples decades into the future – is an ethical and more imperative for educators, parents, and students.

The need for open learning analytics architecture

A few months ago, we released a concept paper on open learning analytics (.pdf). The goal of this paper is to draw attention to the need for algorithmic transparency in order to ensure that context and the needs of individual learners are reflected in teaching and learning.

The model is listed below:

The analytics techniques range significantly, but the intent is to develop modularized analytics methods that educators can experiment and share for others to use and tweak. A few examples of those techniques:

The principles driving this project include:

  1. Algorithms should be open, customizable for context
  2. Students should see what the organization sees
  3. Analytics engine as a platform: open for all researchers and organizations to build on
  4. Connect analytics strategies and tools: APIs
  5. Integrate with existing open tools
  6. Modularized and extensible

Discussions about open online courses, related technical trends (mobiles, social networks), and economic trends (reduced funding for public universities) over the past decade have created a “a boiling point for universities”.

This is the duplication model of academic value: if something can be duplicated with limited costs, it can’t serve as a value point for higher education.

Content is no longer a value point.

Teaching and accreditation still are, but to a lessor degree than only a decade ago.

Individual assessment, teaching, one-on-one consultation and mentorship – those factors that can’t be scaled – serve as the foundation and premise of tomorrow’s education model. Learning analytics serve to give educators information on what’s working and what’s not working. For this reason, analytics tools must be open, embodying the principles of open source movements or the hacker way: iterative, hands on, democratic, open, and transformative.

5 Comments

  1. Frances Bell wrote:

    I agree with so much of what you say George – elements of process being in flux can generate innovation.
    But!! You say “My argument is that openness has not been oversold and that increased openness (of content, teaching/learning, analytics, policy, data, and technology) is really the only path forward for reform.”
    It’s the analytics and data elements that need to be unpicked IMHO. Do you really mean that data, even personal data should be open? There is a significant difference between a learner being able to see and check their own personal data and it being available publicly. It’s a matter of granularity and attribution. A class of students might know what the overall scoring was for their class evaluation and know what how they scored it themselves but not know what each other classmate scored.
    In the UK we have Data Protection legislation where individuals can find out what is held about them and to whom it is passed on http://www.ico.gov.uk/for_the_public/personal_information/how_manage/access_info.aspx
    I think we are in danger of confusing openness about process (who stores what? to whom do they pass it on?) with openness of actual data. Open knowledge (depersonalised evidenced data) will thrive in the public domain – personal data is more tricky – we shouldn’t release that of others – and even our own can be subject to abuse/ misuse.
    Openness sounds like motherhood an apple pie but share their multiple interpretations.

    Saturday, February 11, 2012 at 2:27 pm | Permalink
  2. Dear George (and Francess),

    In the surge for personalisation, I think we are about to cross a boundary to get from personalisation 2.0 to the new 3.0.
    The 1.0 version consisted in mere task oriented functional design, requiring designers to consult the future end users about their needs and wants, whilst keeping an eye on the effectiveness of technical solutions.
    The 2.0 version consisted in ‘user centered design’ in terms of handing over influence and control to the professional end user; the result of this move were technical solutions that would allow end users to control their own set of tools in order to personalise their intranetportal or build their own PLE or PLN. But actually, most users refrain from the painstaking personalisation, except from some minor pimping, thereby mainly using standard features and functions. The development of API’s and possibility to create accounts anywhere by using a facebook, google or twitter account (‘Bring Your Own Profile’) has somehow lifted the burden of keeping track of a myriad of internet accounts and lowered the threshold for end users.
    The 3.0 version of personalisation consists of openness and transparency on the part of the end-user, enabling smart software to adapt and personalise any interface (search engine or otherwise) on whatever smart device whenever the end user introduces himself. Within the same evolvement, user centered design is replaced by community centered design thereby broadening the design focus to social affordances that enable interaction, deliberation.

    Google’s much criticized ‘social search shift’ is an example of the 3.0 version. And Facebook is twisting the 3.0 version to Zuckerbergs lock-in vision of “Bring YourSelf, We’ve Already Got Your Entire Profile.”

    Either way, the way forward seems to be paved with openness and transparancy, to the point where …

    Sunday, February 12, 2012 at 7:23 am | Permalink
  3. gsiemens wrote:

    Hi Frances – individual student performance data should *not* be open. However, learners should be able to see what the organization sees. In particular, I’m advocating for open and transparent sharing of algorithms and weighting measures that influence the types of content a learner sees or the types of social connections the are exposed to. If our learning lives are going to be algorithmically driven, then I’d like to know the decisions being made on my behalf…

    Sunday, February 12, 2012 at 10:25 am | Permalink
  4. gsiemens wrote:

    @Paulo – yes, there are multiple generations of personalization, and google’s social search is one of the more recent examples. Personalization in learning is different from personalization on the web. In a formal course, the knowledge that I as a learner need to master is clearly defined and represents a combination of research and best practices in that field. Personalization in this context provides me with opportunities to fill the gaps in my knowledge. i.e. personalization is “nudging” me toward the knowledge that others have deemed important in that discipline. Personalization on the web doesn’t have as clear an end-target.

    Sunday, February 12, 2012 at 10:29 am | Permalink
  5. Jago Brown wrote:

    Dear George,

    Thank you for the following work & proposal: http://solaresearch.org/OpenLearningAnalytics.pdf

    I’ve been looking in this direction for a while, & now there is something to point to on the horizon.

    Request: I would like to see student friendly dashboard mock-ups for k-12. Plus example data sets that we could generate within existing MIS & LMS (Moodle) systems…

    Kind regards Jago Brown

    Saturday, February 18, 2012 at 8:41 am | Permalink