Network-based Education

This video from Sir Ken Robinson presents a modern critique of public education.  “The Problem”, as he puts it, is that the modern system of education is out-of-date.  It was conceived in the 19th century to serve the Industrial Revolution, and as such it is predicated on a factory metaphor.

The realization that we are now (in the United States) in a post-industrial revolution — i.e., the Information Revolution — helps reveal the deficiencies of the industrial educational model in an largely interconnected society.  A new model is called for, that is conceived as a network, rather than an assembly line.

This video explains the historical transition from the reductionism and specialization that emerged from the Enlightenment (exemplified by the metaphor of a tree) to that of a network (or the metaphor of a web), and suggests that the current climate requires instead people with the ability to communicate across specialist boundaries.

To understand a model of education that is suited to a network-based understanding of knowledge, we must understand the types of interactions that take place with regard to sharing knowledge.  There are three types, and the first isn’t really “inter”-personal at all:

  1. Self-learning.  This is the type of learning mythologized by Matt Damon in the movie Good Will Hunting.  The disadvantage is that students will make lots of mistakes — i.e., their learning will be inefficient.  The advantage is that the knowledge students discover for themselves is likely to be more profound and longer-lasting, and prepare them to create new knowledge thru research.
  2. Learning from a teacher.  This is the model that most US students become familiar with in their K-12 experiences.  In this model, the teacher identifies the learning objectives, designs the exercises, and assesses student progress.  The advantage is that a good teacher will discover what is most efficient and effective, thereby accelerating student progress.  The disadvantage is that teachers are expensive, especially when learning is inefficient despite following the teacher’s lesson plans.
  3. Peer-to-peer learning.  Particularly at residential colleges, students will often form study groups where they learn from one another.  The advantage is that peer-to-peer interaction often helps students of all aptitude levels.  The disadvantage is that group work might be perceived as violating the frequent admonishment to “do your own work” or risk being accused of cheating.  Moreover, where student grades are based upon relative performance, there are disincentives to assist each other.  Lastly, peer-to-peer learning can be inefficient without expert leadership, and getting several people together at the same place and time can be problematic.

All three of these types of learning are typically expected in many college classrooms.  However, they do not all relate to advances in information-communication-technology (ICT) in the same way.  For example, in peer-to-peer and self learning, costs are coming down rapidly.  Ironically, in Higher Education (which is dominated by the teacher-led model), costs continue to rise.  This may be in part because teacher-led instruction has hardly been changed a whit by ICT.  Although more lectures than ever are posted on the internet, most models of on-line instruction involve minimal interaction between the teacher and the on-line students — at least compared with traditional classes.

Traditional teacher-led instruction already uses a broadcast model, where one expert is delivering a monologue.  While it’s now widely recognized that the Internet multiples the potential audience by thousands of times, the basic broadcast model is not necessarily changed.  The economics of this model of teaching places students in direct competition with one another.  That is, the addition of one more student to the class diminishes the attention that other students can receive from the expert.  The only consolation is that once a class gets pretty big (say, more than 40 students), the marginal loss to each existing student by adding another is very small, because few students were genuinely expecting quality interaction with the teacher anyway.  Thus, the economics of on-line courses dictate extremely large classes, where the cost of adding additional students is small and the loss of quality in the experience is minimal.

However, the rapid expansion of ICT has resulted in an explosion of self-learning by making instructional materials extremely inexpensive.  Students now have unprecedented access to archival material, including instructional videos such as those published by Khan Academy, and on-line courseware such as MITx.  Unfortunately, self learning does not result in the credentials necessary to communicate mastery to society.

Peer-to-peer instruction has similarly exploded, and not just via Skype or other technologies that allow distance tutoring, but through websites such as or that allow students to share solutions to homework problems.  Moreover, the videos on Khan Academy do not exist entirely unto themselves — they also invite comments from viewers that can ask questions, write comments that provide answers, or clarify aspects of the video that are unclear.

It is in peer-to-peer learning that Higher Education might find the greatest potential for improvement, albeit at the expense of narrow, albeit cherished, ideals of academic integrity. This is because peer-to-peer instruction is subject to what is called the Network Effect.  Unlike a traditional classroom, addition of new people in a network makes the network more valuable to everyone else already on the network.  Massive economies of scale result.  So in a peer-to-peer learning environment, more students means more knowledge, more interaction, and a better experience for everyone.

Whereas peer-to-peer learning used to be limited by the number of people that could fit around a table, ICT now allows thousands of people to interact without having to be in the same space at the same time.  What hasn’t happened yet (at least not in a formal sense) is a hybrid model in which an expert teacher provides the learning objectives, structures some of the learning activities, moderates the interactions, validates the credentials, but peers conduct the assessments.  We do see the beginning of peer assessment (such as rating comments on YouTube), but we have yet to see a formal course offering that masters assessment in such as way that captures the network effect.

Nevertheless, this hybrid model seems to be the future of Higher Education.  We will likely see fewer Instructors, larger classes, much, much lower costs, and an enormous effort to integrate ICT and social media into college courses.  Most importantly, the expectations of the students and faculty have lagged the technology.  That is, students generally do not feel like part of a learning ecosystem just because a course incorporates social media.  For example, they don’t necessarily feel a responsibility to post, or comment, or answer questions from other students.  Where these options are already available on-line, they’re almost always exploited by intrinsically motivated individuals (except perhaps for some top-rated Amazon reviewers, who do get perks), where 20% of the people do 80% of the work.  But that could change.