Category: E-learning

The Evolution of Computer-Assisted Instruction

Jane E. G. Lipson, Albert W. Smith Professor of Chemistry, Dartmouth College

In 1973 Gotlieb and Borodin published Social Issues in Computing, which included (Chapter 8 “Computer Capabilities and Limitations”) a brief, but fascinating, five pages on the promise and issues associated with computer-assisted instruction (CAI).  From the remove of forty years it is striking to look back and see the extent to which the concerns they raised still resonate. In reflecting on how the view of forty years ago compares with the reality of today it is useful to distinguish – in an approximate way – two different areas in which CAI has evolved. One is associated with the development of curricular material that either focuses on instruction in computational skills, or makes use of computer technology to deliver educational opportunities that would not otherwise be possible (I’ll call this ‘augmented instruction’). The second involves the use of computers and related technical advances (for example, in graphics and delivery of video) to approximate or replace the kind of educational encounters that students have traditionally experienced in person, for example, lectures, tutorial sessions, various modes of testing, etc. (my shorthand will be ‘replacement instruction’).

With respect to the augmented instruction, Gotlieb and Borodin gave as an example the LOGO project at MIT. As described in Social Issues, LOGO provided an introduction to programming skills, and was used even early on by very young students.  It was successful at turning the abstraction of programming into concrete, tangible results – initially by allowing students to program the movement of a robotic turtle.  A good overview of the history of LOGO, formulated by a team at MIT, can be found at  LOGO ended up evolving very successfully; I was amused to learn that the version released in the 1990’s, called ‘Microworlds’, turned out to be the route by which my own two sons (now a university freshman and junior, respectively) were introduced to programming. One can certainly see a bright line from the programming-oriented computational tools being developed when Social Issues was written and those available now, LOGO and derivatives being but one example.

Other technical materials in the category of augmented instruction illustrate the spectacular expansion in opportunities that computers have provided, in large part because of the parallel growth in graphics capabilities, coupled with their plunging economic cost. The most dramatic examples involve training via computer-generated ‘virtual reality’, a category that includes flight simulators, combat training, and surgical training (particularly laparoscopic surgery).  Research is being aggressively pursued in this area, as the goals of providing a lifelike environment (be it a geographical or an anatomical landscape) remain challenging.

Now we turn to so-called replacement instruction, the ways in which computers have (or can) replace or supplement more traditional educational routes. Gotlieb and Borodin suggested that the most significant potential for CAI was in the opportunity it offered to tailor educational material to maintain the interest of students on the edges of the distribution of preparedness – those who are high achievers and those who have learning difficulties.  Within the context of what was available in the early 1970’s the authors discussed some of the obstacles associated with ‘individualized instruction': cost, the limited ‘cognitive’ ability of machines (an example they give is in dealing with spelling mistakes made by students answering computer-provided questions), and the need to operate within a structured school environment that would inevitably be slow to change.

Strikingly, Gotlieb and Borodin suggested that CAI “could come to be important if, as is considered desirable, schools become more decentralized and the educational process is extended over longer periods into adult life.”  What effectively made this prediction come to pass has been the astonishing availability of personal computers, which, arguably, no one anticipated in the 1970s.  As a result, learning has indeed become, to some extent, decentralized.  Students of all ages have effectively unlimited access to information, presented in both textual and visual ways.  One result of there being an online, enormous, ‘student body’ of potential individual learners has been production of a vast amount of tutorial-like material; this is wildly varying in quality, to be sure.  One particularly fine example in this category involves the Kahn Academy (, an educational organization/website created by Salman Khan, whose stated objective is to provide “A free world-class education for anyone anywhere”.  The content includes an impressive library of video tutorials (mainly math and science-related, but now branching out into the humanities and social sciences) and related exercises.  The production values are fairly simple, with the visuals involving an unseen lecturer (often Salman Kahn) speaking, while writing notes on a board.  Viewers can send in questions, which are listed (along with the answers) below the video.  The Kahn Academy has an inter/national reach, which is likely to expand further, as result of a partnership just announced  ( with Comcast.

We are thus now at the point that, instead of relying on help and guidance from teachers, librarians, and other educational professionals to supplement classroom material, students at all levels can access a huge array of resources, day or night, in order to enhance and amplify their understanding of almost any subject. I would argue that help in vetting these resources will continue to be a key role for educators, and that this point should be clearly made to students.  There can be a deceptive aura of authenticity and legitimate expertise associated with seeing concepts explained in videos and described in online texts, and misinformation is not always necessarily benign. The need to educate students in how to judge online credibility will only grow in importance.  That notwithstanding, the age of individual tutorial instruction has definitely arrived.

Now we turn to the last, and most controversial, topic, the role of CAI in replicating the experience of classroom instruction.  The relatively recent series of efforts to reach many thousands of students via online courses (or MOOCs – Massively Open Online Courses) has generated tremendous publicity.  In part, this is because some of these initiatives have involved individuals from highly acclaimed educational institutions, such as Harvard, MIT, and Stanford. My own views on this subject evolved into an OpEd piece that appeared earlier this year in the online magazine, Quartz (see  Of the various points I raised in that article there are two contrasting ones I would reiterate here:  On the one hand, it takes resources and proximity to attend university, and if the choice lies between a larger world of cyber-educated citizens or uneducated ones, I’ll take the former.  However, as a person who has been involved in research and teaching for over twenty-five years at one of the premier educational institutions in the U.S., I firmly believe that “The energy and excitement that animates a campus is generated by the creation, accrual and sharing of knowledge among a community of learners.  Only part of that occurs in organized and predictable gatherings. The sparks come from interactions that are informal and unplanned…”.

Well-delivered, thoughtfully produced online presentations of material will certainly allow for dramatic reach in terms of potential students, and may even serve to enhance what is available  at resource-strapped institutions. However, it is difficult to imagine that they will be able to replace completely the impact of being an active and engaged member of a community learning together in ‘real time’. We are currently in an era of significant experimentation in this regard, and whether the results end up providing credit-driven educational opportunities for the masses or supplementary training for the few remains to be seen.

Near the end of their remarks the authors comment “The computer is only the latest candidate in a long line of technological devices that were supposed to revolutionize education”.  Going back to the ‘augmented instruction’ facet of CAI, mentioned at the start of these remarks, there can be no doubt that computers have created a revolution in education, particularly those aspects of education which use technology to boost technology competence, and which leverage the ability of millions of users to access individually all the resources generously available online. However, things are not as clear regarding the issue of ‘replacement instruction’; just as Gotlieb and Borodin suggested, we continue to grapple with what the role can and should be regarding computer-aided delivery of learning in a group context. The authors noted “the larger issues of the effects on the skills, motivation, and values imparted to students, when machines are substituted for teachers” and I would argue that this concern applies even when the machines are supplying knowledge delivered by remote but unknowable humans in lieu of in-the-flesh teachers.

We are social beings and learning has primarily evolved as a social enterprise.  The vast personal reach enabled by computer technology is individually empowering and socially disruptive. It is causing us to question notions of how to define community, connectivity, and social responsibility.  As our sense of human social networks continues to change, so will our understanding of what it means to be part of a learning community, and this will certainly result in a continuing evolution for the role of computers in education.

Jane Elizabeth Gotlieb Lipson is the Albert W. Smith Professor of Chemistry at Dartmouth College. She is a Fellow of the American Physical Society (APS) and an Associate Editor of the American Chemical Society (ACS) journal ‘Macromolecules’. Honours have included the Camille and Henry Dreyfus Teacher-Scholar Award, the Arthur K. Doolittle Award of the ACS, and service as Chair of both the Polymer Physics Gordon Conference, and the Polymer Physics Division of the APS.  She is very proud of her long-ago peripheral involvement in the production of “Social Issues in Computing”, for which she did literature reference work – as acknowledged at the start of the book.


What Aspects of E-Learning Will Have the Biggest Impact in the Next Decade?

Anant Agarwal, President, edX

In the 40 years since the publication of Social Issues in Computing, the area of computer-assisted instruction (now called e-learning) has developed in huge and unexpected ways.  Who would have imagined that college-level courses would be delivered via computer to students in every country? Or that this delivery would provide a means to generate and mine vast amounts of data for pedagogical research? The social implications of these developments are enormous, and not yet fully understood.  They are tools to democratize education through access, but can also, through research, enhance how education is delivered and packaged, online and on campus.  E-learning is in its infancy, with much unexplored potential. It is quickly evolving, with much more to come.

The next decade will bring continued, rapid change in the educational landscape, with the biggest impact in three areas: blended learning on campus, education research and continuous learning.

Early results of several blended learning pilots, where online learning activities are blended with in-person interaction on campus, suggest improved learning outcomes for students. Instructors too will benefit from using MOOC technology as a next generation textbook. Professors will have a choice to use multiple sources of content in their classrooms that best fit the topic, their teaching style and their students’ learning styles.  For example, MIT professor Michael Cima uses online assessments common in MOOCs in an on-campus class to see if more frequent assessment will improve learning outcomes. We expect to see more of this kind of blended learning on college campuses in the next decade, with SPOCs (Small Private Online Courses) becoming more popular.

In research, I anticipate we will discover more about how people learn, as we work with our partner institutions to mine MOOC’s Big Data.  For example, researcher Philip Guo studied student engagement as it relates to video length. By mining five million video viewing sessions, he concluded that six-minute videos were the ideal length.  In addition, a team of researchers at Harvard and MIT, led by David Pritchard and Lori Breslow, recently released their initial findings.  (Studying Learning in the Worldwide Classroom:  Research into edX’s First MOOC, RPA Journal, June 14, 2013, By Lori Breslow, David E. Pritchard, Jennifer DeBoer, Glenda S. Stump, Andrew D. Ho, and Daniel T. Seaton). One of their findings relates particularly to the social aspects of e-learning.  They found that a student who worked offline with someone else in the class, or with someone with expertise in the subject scored almost three points higher than someone working alone. Pritchard and Breslow’s group concluded that, “This is a noteworthy finding as it reflects what we know about on-campus instruction:  that collaborating with another person, whether novice or expert, strengthens learning.”  This finding will prove particularly insightful for colleges who incorporate online instruction into their on-campus curriculum in blended courses, discussed above.

And finally, I predict that we will see alternative education paths, such as continuous learning, start to emerge. With continuous learning, students may complete first-year subjects through MOOCs, then study for two years in a traditional campus setting—experiencing the magic of campus and group interaction–then enter the workforce for real-world skills, taking MOOCs or other courses as needed through their career in place of the final year.

Anant Agarwal is the president of edX, an online learning destination founded by Harvard and MIT. Anant taught the first edX course on circuits and electronics from MIT, which drew 155,000 students from 162 countries. He has served as the director of CSAIL, MIT’s Computer Science and Artificial Intelligence Laboratory, and is a professor of electrical engineering and computer science at MIT. He is a successful serial entrepreneur, having co-founded several companies including Tilera Corporation which created the Tile multicore processor, and Virtual Machine Works. His work on Organic Computing was selected by Scientific American as one of 10 World-Changing Ideas in 2011, and he was named in Forbes’ list of top 15 education innovators in 2012. Anant is a member of the National Academy of Engineering, a fellow of the American Academy of Arts and Sciences, and a fellow of the ACM. He hacks on WebSim, in his spare time. Anant holds a Ph.D. from Stanford and a bachelor’s from IIT Madras.