Several decades ago – long before the level of technological sophistication we experience today -- I was part of a movement begun by the late Julian Stanley, a psychology professor, and the Johns Hopkins University Center for Talented Youth (CTY) to save academically talented youth from boredom in the schools. The most controversial instrument to rescue them was a pedagogical practice called, rather prosaically, "Diagnostic Testing Followed by Prescriptive Instruction" or, shorthand, “DT>PI.” It was principally applied to the pre-collegiate mathematics curriculum and relied on just a few key assumptions and practices:
1. Students already know something about a subject before they formally study it.
2. Test students before a course begins and then just instruct them on what they don’t know.
3. Test students again when you as the instructor and they as learners believe they have competency in a subject.
4. Move immediately to the next level of instruction.
The DT>PI model was placed in a more generous context with the adaptation of Professor Hal Robertson’s (University of Washington) notion of the Optimal Match. Simply stated, pace and level of instruction should match optimally an individual student’s assessed abilities — with the caveat that those accessing that talent would always try to stretch a student beyond his comfort zone. The Optimal Match theoretically could apply to all students at any level of education.
When I used to speak publicly in a wide variety of settings — at colleges and universities, community colleges, schools, education association meetings, parent gatherings -- about what I thought to be the commonsensical notions of DT>PI and the Optimal Match, the reactions were pronounced and fiercely negative. My colleagues and I were accused of presenting educators with the dissolution of the structured classroom as we knew it then; forcing students unjustifiably to proceed educationally without sufficient instructional guidance; destroying the communal, cooperative imperative of an American education; and, producing social misfits because students would finish academic coursework before the schedule established (rather arbitrarily, I might add) by educational professionals for all students of one age at one time. Parents joined often with educators to decry such imagined alienation and damage to a child’s personality.
And then there was a change in 2013.
There are now two closely related pedagogies -- adaptive learning and competency-based learning -- that are embraced by a growing number in higher education as a viable component of educational reform. The Bill and Melinda Gates Foundation is awarding grants to 18 institutions to experiment with 10 different adaptive learning platforms, and President Obama has expressed support for these innovations and urged easing of regulations to make that possible.
In general, adaptive learning uses data-driven information to design coursework that permits students to proceed educationally at their appropriate pace and level. And competency–based learning allows students to be free of "seat time" and flexibly progress as they demonstrate mastery of academic content.
These definitions, when combined, delineate precisely the key components of DT>PI and that of numerous other experiments in self-paced learning over the last few decades. But now, while the naysayers are still out there, an increasing number of for-profits and nonprofits are turning to adaptive and competency-based learning as a component of the next stage of reform in American education.
Why now? Something must have changed in society to accept self-paced, individualized learning when only decades ago it was roundly rejected on pedagogical, ethical and psychological grounds. Those concerns are clearly not as inviolate as they were only years ago. Answering this question might well provide education reformers with insight into what is now possible — even expected -- from students for the learning platforms of the future.
There are at least three reasons why self-paced learning might be more popular now than it was only a few years ago: technological advances, financial exigency and a new self-profile of the learner.
Advances in technology that rely on advances in data mining and data analytics -- predicting future learning behavior by an individual based upon analysis of thousands of earlier learners — permit now a high ability to track, direct, customize, evaluate and advise student learning at instantaneous speeds. What in previous decades seemed to be an impossible task for a teacher or professor to manage in a single course — diverse learning points among students — is at least now technically feasible.
Many institutions are rather intent to find new strategies that will at once reduce their cost of providing an education. Adaptive and competency-based learning are thought to be such "disruptive" opportunities, although how accompanying data-driven, all-knowing and anticipating, high-touch technologies will reduce dramatically both cost and price (tuition) remains elusive.
And, lastly, students have perhaps finally realized the expectation of the self-esteem movement that has dominated instruction in our nation’s schools for several decades. Students might well now believe that they are the center of all activity — to include education — and that they are both the sole focus and the drivers of learning. All instructional effort exists for the purpose of fulfilling their desires.
This "power shift" makes learners, individually — not teachers or professors -- aggregators of knowledge by and for themselves. Any approach to education that places them at the center of learning activity accommodates their perspective on education. Adaptive and competency-based learning accomplish this masterfully. Self-paced, individually adjusted instruction, enhanced by “big data” technologies that guide student progress “lockstep” in a course and beyond, eliminates distracting elements to the individual control of knowledge. Primary among those distractions for students are faculty with their pesky, seemingly inefficient and irrelevant questions.
And thus, in 2013, what was not acceptable several decades ago is now thought a solution to crisis in American education. A combination of new technologies, financial emergency and a shift in who is at the center and in control of learning has caused this to occur.
But all is not settled. The changing circumstances introduce concerns that did not exist decades ago when students were not the arbiters of their own learning, self-paced instruction was not thought to be a solution for all students in American education but only the academically talented and big data did not exist to mine and anticipate every move in student learning.
A defining element of DT>PI was that students must not just study what is the next logical step in a course, but they must through the exhortations of a teacher or professor attempt to go beyond what was thought statistically possible — they must stretch themselves intellectually at every point. Professor Stanley used to constantly quote the line of the poet Robert Browning that one’s reach must always exceed one’s grasp.
Questions remain whether in the absence of a live instructor exhorting a student who is not necessarily academically acute and motivated, students will extend their reach or settle for statistically generated achievement delivered by an electronic adviser (I am referring here to traditional-aged undergraduates, not working adults who are propelled by substantial motivational factors). Such absence of exhortation could be extremely damaging to the majority of American students who often do not naturally attempt to achieve to the levels of which they are capable without personal mentorship.
And one traces in those who are enthralled with "big data" and "data analytics" for solving the maladies of American education a disturbing belief. Student will achieve through data-enhanced technologies the perfectibility of education — perhaps life itself -- by eliminating all resistance, frustration, indecision, trial and error, chance and expenditure of time. For example, Harvard University social scientist and university professor Gary King is quoted in a May 20, 2013 New Yorker article entitled "Laptop U." as saying, “With enough data over a long period, you could crunch inputs and probabilities and tell students, with a high degree of accuracy, exactly which choices and turns to make to get where they wanted to go in life."
And yet, there is growing commentary that it is precisely the absence of frustration, resistance and associated imperfections in a so-called “Me Generation” and its aftermath that is compromising contemporary students' learning and preparation for a life. By educators blithely accepting students’ assertion of self-determination without legitimate maturing experiences (that will include failure and self-doubt) and by arranging learning electronically so that they will make no wrong decisions, they are granting them little ability to deal with inevitable disappointment and frustration in life.
Students are educated without gaining resilience and that is hardly an education of which a nation can be proud or secure, regardless of the utopian promises of the big data enthusiasts. All this reminds me of a call I received decades ago from an entrepreneur who wanted me to comment on his idea of developing a school basketball court that would have the hoop move electronically with the ball so that no student would ever miss a shot and thus, in his words, "suffer humiliation."
So while I am delighted that self-paced education in the form of adaptive and competency-based learning is finally a more generally discussed component of reform in American education, I urge that those advancing it think long and hard about some of the humanly-damaging consequences of learning platforms so perfected by technology that students are offered a Faustian bargain – the comfort of non-resistant and frustration-free learning in exchange for the ultimate loss of a resilience needed for a satisfying life after schooling.
William G. Durden is president emeritus and professor of liberal arts at Dickinson College, and operating partner at Sterling Partners, a private equity company.
"Frenzy" may be the best way to describe what’s currently happening in higher education.
On one hand, there’s MOOC (massive open online course) mania. Many commentators, faculty creators, administrators, and public officials think this is the silver bullet that will revolutionize higher education.
On the other hand, there is the call for fundamental rethinking of the higher education business model. This is grounded most often in the argument that the (net) cost structure of higher education is unaffordable to an increasing number of Americans. Commentators point out that every other major sector of the economy has gone through this rethinking/restructuring, so it is only to be expected that it is now higher education’s turn.
Furthermore, it is often claimed that colleges and universities need to disaggregate what they do and outsource (usually) or insource (if the expertise is really there) a re-envisioned approach to getting all the necessary work done.
In this essay I focus on the optimal blending of online content and the software platforms underneath.
Imagine how transformative it would be if we could combine self-paced, self-directed postsecondary learning (which has been around in one form or another for millennia) with online delivery of content that has embedded in it both the sophisticated assessment of learning and the ability to diagnose learning problems, sometimes even before the learner is aware of them, and provide just-in-time interventions that keep the learner on track.
Add to that the opportunity for the learner to connect to and participate in groups of other learners, and, to link directly to the faculty member and receive individualized attention and mentoring. What you would have is the 21st-century version of do-it-yourself college, grounded in but well beyond the experienced reality of the thousands of previous DIYers such as Abraham Lincoln, Frederick Douglass, and Thomas Edison.
A good goal to set for the future? No. The great news is that we already have all the components necessary to make this a reality in the near term. First, it is now possible to build “smart” content delivered through systems that are grounded in neuroscience and cognitive psychological research on the brain mechanisms and behaviors underlying how people actually learn. The Open Learning Initiative at Carnegie Mellon University, which creates courses and content that provide opportunities for research for the Pittsburgh Science of Learning Center (PSLC), is an example of how research can underlie content creation.
Such content and systems depend critically on faculty expertise, in deciding exactly what content is included, in what sequence, and how it is presented. Faculty are also critical in the student learning process, but perhaps not solely in ways we have traditionally thought. That is, it may not be that faculty are critical for the actual delivery of content, a fact we have known for millennia given that students obtain content through myriad sources (e.g., books) quite successfully.
Still, effective and efficient student learning has always depended critically on how well faculty master both these content steps as well as the other parts of the learning process, as evidenced by the experience with faculty who are experts at doing it and the ease with which learning seems to happen in those situations.
Second, these “smart” systems exist in a context of sophisticated analytics that do two things: (a) monitor what the learner is doing such that it can detect when the learner is about to go off-track and insert a remedial action or tutorial just in time, and (b) assess what the learner knows at any point. These features can be used to set mastery learning requirements at each step such that the learner cannot proceed without demonstrating learning at a specific level.
Ensuring mastery of content has long been a major concern for faculty, who used to have to spend hours embedding pop quizzes or other learning assessments into their courses, set up review sessions, set office hours during which students may (or may not) attend, and implore students to contact them is they encountered difficulties. The dilemma for faculty has usually been figuring out who needs the assistance when and how.
The sophisticated analytics underneath content delivery systems help take the guesswork out of it, thereby enabling faculty to engage with more students more effectively, and, most important, design the engagement to address each student’s specific issue. Better student-faculty interactions will likely do more to improve student learning than most any other intervention.
Third, the platforms on which these “smart” systems are built and delivered include ways to create virtual teams of learners (both synchronously and asynchronously) and to include faculty interaction from one-on-one to one-on-many. This tool will make the long tradition of having students form study groups easier for faculty to accomplish, and enable students whose physical location or schedules may have made it difficult previously to participate in such groups to gain their full benefit.
Fourth, the creation of these “smart” systems has resulted in much clearer articulations of the specific competencies that underlie various levels of mastery in a particular field. As evidenced by the various articulations and degree profile work done in the U.S. and internationally, and by the development of specific competencies for licensure by several professional associations, faculty play a central role.
Fifth, the specification of competencies makes it easier to develop the rubrics by which learning acquired prior to formal enrollment in a college/university or in other ways not otherwise well-documented can be assessed, and the learner be placed on the overall continuum of subject mastery in a target field or discipline. Although faculty have always played a central role in such assessments, standardization of assessment has proven difficult. However, with the inclusion of faculty expertise, assessments such as Advanced Placement exams and learning portfolios can now be accomplished with extremely high reliability.
All of this could have enormous consequences for higher education. To be sure, we need more research and development of a broader array of content and delivery approaches than we currently have. In the meantime, though, three steps can be taken to meet students’ needs and to increase the efficiency with which colleges and universities provide the educated citizens we need:
Define as many postsecondary credentials as possible in terms of specific competencies developed by faculty and practicing professionals. This will provide the bases for developing as many “smart” systems as possible for improved content and learning assessment, and for assessing prior learning.
Meet students at the edge of their learning. Each student that arrives at a college/university is at a different spot along the learning continuum. Previously, we made at best very rough cuts at determining where students should start in a course sequence, for example. But more sophisticated prior learning assessment means we can be much more precise about matching what the student knows and where s/he should connect to a learning sequence. Not only would this approach minimize needless repetition of content already mastered, but it could also provide faster pathways to credentials.
Design personalized pathways to credentials. Better and clearer articulation of what students need to know for a specific credential, plus better assessments of prior and ongoing learning, plus more sophisticated content, plus the opportunity for faculty to engage individually and collectively with students in more focused ways means we can create individual learning plans for students to complete the credentials they need. In essence, a learning gap analysis can be done for each student, indicating at any point in time what s/he still needs to know to achieve a credential. Faculty mentorship can become more intrusive and effective, and a student’s understanding of what and why specific knowledge matters would be deeper.
Institutions that have greater flexibility to address these steps will be the most likely to succeed. I am heartened by the many professors and administrators who are creating the innovative approaches to make the changes real, and to embed them in the culture of their respective institutions. They provide students with superior advising and clearer pathways to achieving the academic credentials students seek. In the longer run, those institutions are likely to see cost structures decline due to more efficient progress through academic programs.
The technology-driven changes described here may well enhance student learning, and help us reach the goal of greater access to higher education for adults of all ages.
But it raises a crucial, and largely unaddressed, question that gets lost in debates about whether costs can be reduced using such technology or whether it will result in fewer faculty jobs.
We have not yet adequately confronted the definition of “faculty” in this emerging, technology-driven environment. Although a thorough discussion of that issue necessarily awaits a different article, suffice it to say that just as technology and costs have changed the job descriptions of people in most other professions, including health care, it has also created new opportunities for those in them. For instance, even though the rise of nurse practitioners has changed key aspects of health care delivery, the demand for more physicians, whose job descriptions may have changed, remains.
In any case, the best part is that these new approaches do not replace the most important aspect of education — the student-teacher interaction. Rather, they provide more effective and efficient ways to achieve it.
John C. Cavanaugh is president & CEO of the Consortium of Universities of the Washington Metropolitan Area.