Personalized Learning: It’s not the Algorithm, It’s the Learner
We live in a time of porous boundaries between human intelligence and machine intelligence (rightly called “artificial intelligence”). We need a Turing Test to decide whether an entity is human or not. If you apply for something online, you may have to prove, through Captcha, that you are not a machine. And, when it comes to the challenge facing education -- how to provide quality education for large numbers of students at a reduced cost – the temptation to cross the machine-human boundary and let machines (that is, algorithms) do the heavy lifting is almost irresistible.
That temptation needs to be resisted.
As much as information technology can bring efficiencies to education, in the end the best learning is not machine-driven but human-driven. Machines (that is, digital technologies) can remove physical obstacles and provide helpful guiding information but in the end humans learning with other humans in real-life situations is still the best way to learn.
Machines provide information faster than anyone could have imagined but learning is making sense of information and discovering its meaning, the real goal of learning, and something machines can’t do (yet).
Digital Technologies Provide a Leg Up
Learners can benefit from the guidance of algorithms that point the learner to online tutoring systems, for example, that are proving as effective as human tutors. learners can learn methods and approaches from the online tutors that then help them along their own learning paths. Their own learning paths. That’s the point: adult learners (that is college-age learners) learn best when they themselves create learning paths; the online tutor can provide a leg-up, but they cannot be the whole of the learning experience.
The Promise and Peril of Adaptive Learning Technologies
Adaptive learning technologies, online learning analytics used to create learning paths for learners based on their performance, might help some learners but cannot, in most cases, provide the opportunity for deep and lasting knowledge about how to learn. The machine, in adaptive learning technologies, has taken over: the algorithm is creating learning pathways, not the learner. This approach could be understood as an attempt at “semi-passive learning.” This is not to say there are not uses for adaptive learning technologies, but it is to say that this approach can only be one element in a human-driven learning path. (I did participate, as a consultant, in a project tangentially about adaptive learning technologies that was funded by the Gates Foundation; we are all experimenting with how best to partner with digital technologies but I think the golden rule has to be “thou shalt give the reins to the learner.”)
The Age of Personalization
But personalized learning, in comparison to adaptive learning technologies, is a very broad concept, encapsulating our entire culture’s experience with digital technologies. “Personalization” could describe the major effect that digital technologies are having on our culture. My car remembers how the driver’s seat should be configured for me and how the outboard mirrors should be positioned for best viewing from that seat positioning. My car is personalizing part of my driving experience. Our Blue Ray player remembers exactly where we were in a show we were watching the previous night; Google “knows” you and personalizes your search experience. Personalization affects all parts of our lives.
But only a human can truly personalize everything she or he does. It is the age of personalization but that only means assisting each of us to spend less time on details and more time on important human activities, such as imagination, creativity, discovery, integrating, intuition, taking leaps of faith. Personalization by digital technologies only frees us humans to better personalize our lives (that is, find our own ways).
Personalization in Learning
One of the most important personalizing technologies, maybe the most important of personalizing technologies for learning, is the eportfolio. Why? Because not only does it help you create your own personal identify on line, but using an eportfolio frees learners to follow whatever learning path they wish while still documenting (with a smart phone) that personal learning path and making it part of their learning history. ePortfolios allow learners to take the lead, make their own mistakes, or serendipitous discoveries, and learn as humans learn best.
This may be the age of digital technologies, but it is also the age of unleashed learners: machines are only good if they enhance the lives of humans. They are not good is they take over the lives of humans or take over learning experiences.
Our whole effort in using digital technologies to improve learning should not be to impose new restrictions on curiosity and discovery by finding new ways of leashing the learner to the machine, but to free learners from the limitations of previous technologies and of other physical limitations on learning. Algorithms, like good teachers, need to be guides on the side and not new sages on the stage.
The AAEEBL Annual Conference, July 27-30 in Boston, includes personalized learning as one of the main themes of the conference.
The New Media Consortium Horizon Report and Personalization
This year’s New Media Consortium Horizon Report includes a section on personalized learning.
From the abstract to this section:
“Personalized learning refers to the range of educational programs, learning experiences, instructional approaches, and academic-support strategies intended to address the specific learning needs, interests, aspirations, or cultural backgrounds of individual students.”
So far, so good. This definition is broad, as it should be, describing personalized learning as a general trend in education, in recognition of the realization that the more active that learners are on behalf of their own learning, the deeper and longer-lasting their learning will be. But in subsequent paragraphs in this abstract, personalized learning seems to move more and more toward machine-driven and not human-driven.
“While there is a demand for personalized learning, it is not adequately supported by current technology or practices. The increasing focus on customizing instruction to meet students’ unique needs is driving the development of new technologies that provide more learner choice and allow for differentiated instruction. Advances such as online learning environments and adaptive learning technologies make it possible to support a learner’s individual learning path.”
Hmmm. Now, the text is moving toward an implication that support for personalized learning may or should mostly come from digital technologies. I thought letting the student out of the classroom in itself was supporting “a learner’s individual learning path.” The report’s abstract goes on to say:
“The biggest barrier to personalized learning, however, is that scientific, data-driven approaches to effectively facilitate personalization have only recently begun to emerge; learning analytics, for example, is still evolving and gaining traction within higher education.”
And, there we have it – if artificial intelligence is not the driver, personalization cannot be fully supported. I vote for human intelligence as the driver; we don’t want to replace an overbearing teacher with an overbearing machine.
And, of course, nowhere in this section are eportfolios mentioned. That is a big error in conceptualizing “personalized learning.” That is like saying personal mobility depends on the car while ignoring the highway system. But, I hasten to applaud the report in general for referring to very good examples of colleges and universities who are incorporating personalized learning. That these institutions may be using eportfolios to support their move to personalized learning was somehow missed in this report.
The Report, as a whole, is a magnificent survey of technology in higher education. It gives detail to the general digital environment in which learners use their eportfolios. We can see, by reading through the Report how pervasively digital technologies have infused higher education with new capabilities for all purposes. If anything, the Report explains why eportfolios are necessary: in such a digitized ecology of learning, in which the classroom is no longer “the center,” where is the center? For learners, the new center is their eportfolio.
Let’s add a note that the Report should have included:
In the digitized environment in which learners now work and learn, integration of diverse experiences is newly crucial, and reflection on artifacts from those experiences is the key to learning. Portfolio evidence is the new learning pathway: what is collected, curated, analyzed, and published is how a learner makes sense of learning. The eportfolio process provides personalized coherence and a learning space that replaces the classroom as the center of learning.