Personalizing education for special needs

One of the most compelling arguments for personalized learning is the importance of providing an appropriate education to students with special needs. Such students challenge the system, with unexpected strengths and weaknesses that are out of scale with the norm. Simply slowing down (or speeding up) the pace of instruction won’t serve their needs, particularly as they may be exceptional on more than one dimension and in more than one direction. For them, personalized learning that decouples different skills is imperative, a way to serve their needs and extend their abilities at the same time.

While special-education laws are limited in scope due to their approach of simply setting minimum requirements, they do provide critical safeguards for supporting students at the K-12 level. As they graduate to adulthood, these students are expected to assume more responsibility to advocate for and seek accommodations for their needs if they pursue advanced study at institutions of higher education (IHEs). Even so, these minimum accommodations only grant access, and sometimes may not do enough to constitute effective instruction that enables success. Simply fulfilling minimum requirements may allow IHEs to avoid litigation, but failing to adequately serve their students is a failure to invest their resources wisely.

Challenging though it may be to (re)design instructional materials with different constraints, IHEs may find that special-needs students can provide a valuable test case, instantiating extremes on the spectrum of students they serve. These adjustments will also help them support English language learners, disadvantaged-but-capable students with gaps in their backgrounds, returning students who remember some lessons but forgot others, and career changers in search of very specific skills to flesh out their resume—deserving students whom the traditional system fails, all too often. Not all students fit the same mold, nor should they. Adapting instruction around their needs develops their potential and gives them the opportunity to give back.

Messy personalized learning

Phil Nichols describes his youthful adventures reappropriating the humble graphing calculator to program games:

For me, it began with “Mario” — a TI-BASIC game based loosely on its Nintendo-trademarked namesake. In the program, users guided an “M” around obstacles to collect asterisks (coins, presumably) across three levels. Though engaging, the game could be completed in a matter of minutes. I decided to remedy this by programming an extended version. I studied the game’s code, copying every line into a notebook then writing an explanation beside each command. I sought counsel from online tutorials, message boards, and chat rooms. I sketched new levels on graph paper, strategically placing asterisks in a way that would present a challenge to experienced players. Finally, after a grueling process of trial and error, I transformed my designs into code for three additional stages.

As he summarizes, his non-school-sanctioned explorations of an otherwise school-based tool led to sophisticated discoveries and creations:

[W]ith the aid of my calculator, I’d crafted narratives, drawn storyboards, visualized foreign and familiar environments and coded them into existence. I’d learned two programming languages and developed an online network of support from experienced programmers. I’d honed heuristics for research and discovered workarounds when I ran into obstacles. I’d found outlets to share my creations and used feedback from others to revise and refine my work. The TI-83 Plus had helped me cultivate many of the overt and discrete habits of mind necessary for autonomous, self-directed learning. And even more, it did this without resorting to grades, rewards, or other extrinsic motivators that schools often use to coerce student engagement.

While he positions calculator programming as a balance between the complementary educational goals of “convention” and “subversion,” this also echoes tradeoffs between routine expertise and adaptive expertise, between efficiency and creativity, or between convergent and divergent thinking. It remains an ongoing risk in overly restrictive learning environments. Standards that dictate the time and sequence of each stage of students’ progression fail to allow for the different paths which personalization accommodates. Yet even adaptive learning systems that seek to anticipate every next step a student might take must be careful not to add so many constraints that crowd out productive paths the student might otherwise have pursued. Personalized learning needs to leave room for error and open-ended discovery, because some things just aren’t known yet.

Is adaptivity a qualitative or quantitative problem?

One common criticism of adaptive learning is that by tailoring instruction so closely to students’ needs, it doesn’t challenge them enough. As embodied by James Paul Gee’s critique:

People who never confront challenge and frustration, who never acquire new styles of learning, and who never face failure squarely may in the end become impoverished humans. They may become forever stuck with who they are now, never growing and transforming, because they never face new experiences that have not been customized to their current needs and desires.

While I agree with the dangers of what he describes, I question the causal attribution.

First, adaptive learning systems that indulge in too much customization may instead be guilty of relying on a too-narrow prescription for the student’s “zone of proximal development (ZPD)”. Individualized learning does not require giving only incremental steps; it can (and should) include more ambitious steps to occasionally challenge students, perhaps just beyond their conventional ZPD (or at the limits of their ZPD when defined by “lots of help”). Students need to struggle—manageably—as part of their learning. Adapting to students’ needs can include optimizing the nature and amount of that struggle based on past experiences and future expectations.

Second, this can also be overcome by building a certain amount of variability into the system, for the sake of both the students and the system. Occasionally presenting students with problems that may or may not lie within their ZPD can help them learn “what to do when you don’t know what to do” (in the words of a dear colleague of mine, Joe Wise). Whether framed as desirable difficulties, germane cognitive load, preparation for future learning, or the development of adaptive expertise rather than just routine expertise, unexpected challenges can offer invaluable learning opportunities. Further, adaptive learning systems need to reach beyond what is already known in order to improve themselves. A truly intelligent system should be discovering new knowledge about its particular learners and even about learning in general. The possible paths a student might take are infinite, and the system’s designers don’t know what’s best—only what tends to be better compared to other paths that have already been examined. That is, adaptive learning must itself be an adaptive learner.

Both of these issues point to a quantitative problem due to adapting too narrowly or too often. The deeper question is whether adaptivity is a fundamental, qualitative problem: Does having any adaptivity at all invite complacency among students accustomed to having their learning experiences at least partly tailored to their needs? Given the well-established importance of scaffolding instruction according to students’ needs, I would argue that adaptive learning is a valuable tool not simply for accelerating but also for enriching instruction.

MOOCsperiments: How should we assign credit for their success and failure?

That San Jose State University’s Udacity project is on “pause” due to comparatively low completion rates is understandably big news for a big venture.

We ourselves should take pause to ponder what this means, not just regarding MOOCs in particular, but regarding how to enable effective learning more broadly. The key questions we need to consider are whether the low completion rates come from the massive scale, the online-only modality, the open enrollment, some combination thereof, or extraneous factors in how the courses were implemented. That is, are MOOCs fundamentally problematic? How can we apply these lessons to future educational innovation?

Both SJSU and Udacity have pointed to the difficulties of hasty deployment and starting with at-risk students. In an interview with MIT Review, Thrun credits certificates and student services with helping to boost completion rates in recent pilots, while the inflexible course length can impede some students’ completion. None of these are inherent to the MOOC model, however; face-to-face and hybrid settings experience the same challenges.

As Thrun also points out, online courses offer some access advantages for students who face geographic hurdles in attending traditional institutions. Yet in their present form, they only partly take advantage of the temporal freedom they can potentially provide. While deadlines and time limits may help to forestall indefinite procrastination and to maintain a sense of shared experience, they also interfere with realizing the “anytime, anywhere” vision of education that is so often promoted.

But the second half of “easy come, easy go” online access makes persistence harder. Especially in combination with massive-scale participation that exacerbates student anonymity, no one notices if you’re absent or falling behind. While improved student services may help, there remain undeveloped opportunities for changing the model of student interaction to ramp up the role of the person, requiring more meaningful contributions and individual feedback. In drawing from a larger pool of students who can interact across space and time, massive online education has great untapped potential for pioneering novel models for cohorting and socially-situated learning.

Online learning also can harness the benefits of AI in rapidly aggregating and analyzing student data, where such data are digitally available, and adapting instruction accordingly. This comes at the cost of either providing learning experiences in digital format, or converting the data to digital format. This is a fundamental tension which all computer-delivered education must continually revisit, as technologies and analytical methods change, as access to equipment and network infrastructure changes, and as interaction patterns change.

The challenges of open enrollment, particularly at massive scale, replay the recurring debates about homogeneous tracking and ability-grouping. This is another area ripe for development, since students’ different prior knowledge, backgrounds, preferences, abilities, and goals all influence their learning, yet they benefit from some heterogeneity. Here, the great variability in what can happen exaggerates its importance: compare the consequences of throwing together random collections of people without much support, vs. constraining group formation by certain limits on homogeneity and heterogeneity and instituting productive interaction norms.

As we all continue to explore better methods for facilitating learning, we should be alert to the distinction between integral and incidental factors that hinder progress.

Repositioning personalized and adaptive learning

Amidst all the excitement about personalized and adaptive learning is a lot of confusion about what the terms mean. In this post, I will examine some of the commonly-used definitions around personalized learning and explore the relationships between the ideas to clarify how I intend to use the terms here. My next post will explore the differences between personalized and adaptive assessment.

Education Growth Advisors describes personalization as “moving beyond a one-size-fits-all approach to instruction,” in which students might receive additional work as challenge or remediation, depending on their current performance. In their view, adaptivity is a more specialized form of personalization which “takes a more sophisticated, data-driven, and, in some cases, non-linear approach to remediation.” Some examples might use past performance to adjust the focus, timing, or path of content delivered to a learner.

In contrast, EdSurge places personalized learning at the top of its hierarchy with this definition: “When instruction is truly geared to the student: learning objectives, content, presentation methods and pace may all vary depending on the learner.” By their definitions, adaptive learning varies the content, individualized learning is self-paced, and differentiated learning varies the content and presentation methods, all three being lesser versions of personalization.

While both groups classify adaptive learning as a subcategory of personalized learning, they differ in where they place that subcategory along the personalized-learning spectrum, with Education Growth Advisors deeming it more sophisticated and EdSurge considering it less sophisticated compared to other forms of personalized learning. My definition will likewise situate adaptive learning as being narrower than personalized learning, but without making any claims as to which is superior in either sophistication or effectiveness. In being broader than adaptive learning, personalized learning can be both better and worse. The difference between them depends on what is being tailored to the individual, not how well it is executed.

EdSurge restricts adaptivity to selection of content; my definition will also encompass adaptivity of learning objectives, presentation methods, focus, timing, and path through content. (This is akin to Education Growth Advisors’ definition of adaptivity and closer to EdSurge’s definition of personalization.) The reasoning is that the learning environment may be tailored along all of these dimensions according to individual needs.

In my definition, personalized learning includes other features beyond adaptive learning to make the learning experience more personal—quite simply, involving the person. These may include individual preferences, chosen by the user; instructor actions, informed by a faculty dashboard offering analytics and recommendations but with decisions left up to the teacher; and social interaction, facilitated by collaborative tools and engineered to encourage productive learning, but again leaving room for human input. While adaptive learning relies on data-driven decisions from machine intelligence to tailor the experience to the learner, personalized learning also adds an explicit role for judgments made by human intelligence.

A fully personalized learning ecosystem may allocate some decisions to be determined by machine intelligence and delivered through its automated adaptive learning system, designating others to be handled by human intelligence and provided by various actors in the ecosystem. The adaptive component optimizes the environment, to the extent that it can control it, along parameters which its data and algorithms predict will be beneficial. The personalized components make allowances where human intervention is deemed to be more valuable (such as user choice, yet-to-be-modeled human expertise, and social interaction). Still, the unpredictability of human behavior also means that those components cannot claim to be truly adaptive, only personalized.

Thus, personalized learning has the potential to be much more sophisticated and powerful than adaptive learning, if realized effectively. Early iterations of personalization may not take full advantage of these possibilities, and in some cases, human decisionmaking may end up being maladaptive. Our goal is to help clarify those dimensions and parameters to enable better design and use of personalized learning.

Why personalized learning and assessment?

Much of the recent buzz in educational technology and higher education has focused on issues of access, whether through online classes, open educational resources, or both (e.g., massive open online courses, or MOOCs). Yet access is only the beginning; other questions remain about outcomes (what to assess and how) and process (how to provide instruction that enables effective learning). Some anticipate that innovations in personalized learning and assessment will revolutionize both, while others question their effectiveness given broader constraints. The goal of this blog is to explore both the potential promises and pitfalls of personalized and adaptive learning and assessment, to better understand not just what they can do, but what they should do.