Personalized learning is not used on a large scale because it is hard - for all the stakeholders in the educational enterprise.
Teachers. Currently, the challenge of personalized learning falls squarely on the shoulders of classroom teachers. It requires a heroic effort, as well as extensive knowledge. Customizing instruction is the pinnacle of professional skill: teachers must have the pedagogical content knowledge to accurately diagnose current student understandings; they must have the content knowledge to related these diagnoses to ideas and concepts in the domain that students are struggling with; they must have the time and instructional design skills to create appropriate curriculum; they must have exemplary classroom management skills to be able to handle individual students or groups of students working on different problems and activities simultaneously. Furthermore, historically, teachers have not been provided with the material basis to support adapting curriculum or creating new curriculum. Within K-12 classrooms and many undergraduate classrooms, instruction is dominated by the textbook; the ultimate one-size-fits-all material. Apparently the US leads the world in the percentage of instruction in K-12 classrooms which is textbook-led. Within undergraduate settings, faculty are increasingly relying on slide presentations developed by the publisher and distributed with the textbook, thus perpetuating the lecture model of instruction in higher ed. For instance, suppose a high school Earth science teacher determined that several students do not have the quantitative skills to complete a lab assignment while others do not have the reading skills to understand the student text: what is she to do? Have the students re-read the text louder and slower until they get it?!?!? Have the students continue to mindlessly crunch numbers using a formula they do not understand? Teachers need a selection of materials readily available for them to customize their instruction to meet these learners' specific needs, and they need powerful tools to help scaffold some of the knowledge and skills that they need to customize instruction to meet diverse learner needs.
Learners. Personalized learning often requires significant self-actualized learning skills on the part of students. Cognitive science research has repeatedly shown that the best predictors of success in learning are content knowledge and meta-cognitive skills. Yet, these are the very skills and knowledge which many learners lack. It is the ultimate "rich-get-richer" scenario: the more you already know and the more skilled you are at learning, the easier it is to learn new stuff. Metacognitive skills influence ones' ability to "learn how to learn" and include monitoring one's state of understanding, identifying knowledge gaps, and determining when more information is needed; whereas, domain content knowledge is often essential for determining the relevance, applicability and accuracy of information sources. These skills are particularly critical in a content-rich world offering a broad range of choices for learning materials and learning experiences. In my work with colleague Kirsten Butcher from the University of Utah, we often talk about the sensemaking paradox faced by 21st century students: we routinely expect them to employ information literacy skills to learn complex scientific topics, yet they need significant science domain knowledge in order to apply their information literacy skills effectively. This paradox is particularly troubling when contemplating the use of OER in K-12 education, where students and many teachers lack basic science content knowledge to make effective choices about materials to support their own teaching and learning. Further, Margolis has recently highlighted how learners from underrepresented groups in STEM education are additionally challenged by digital learning, lacking the "preparatory advantage" that accrues from growing up in technology-rich households.
School Districts and Researchers. It might seem strange to lump these two stakeholder groups together, however, I want to discuss a shared mindset which hinders both groups making progress on personalized learning: the desire for implementation fidelity. For the past twenty years, one of the gold standards for measuring the success of a new curriculum, technology, or intervention is the consistency with which different teachers are able to implement the intervention across different settings and students. The idea being that consistent implementation leads to consistent learning outcomes. In the business world, this is often referred to as replicating best practices, which is a generally a good thing to do. The challenge for personalized learning is the inherent tension between supporting customization while supporting standardization or fidelity. We want teachers to benefit from best practices and proven innovations, yet personalizing instruction requires that they adapt and modify these practices and innovations to better serve their local needs and on-the-ground conditions. As researchers, we are often asked, whether by NSF panels or school district curriculum coordinators, how much can teachers and learners customize or personalize yet still maintain fidelity to the program and coherence in the curriculum? This is a very difficult question for which we often do not have a satisfying answer. The old adage about your grandfather's ax comes to mind: you've changed the handle five times and the blade four times but it is still your grandfather's ax. As such, in some of our applied research, we have been trying to shift the conversation from supporting teachers to implement curriculum with fidelity to supporting teachers to "customize with integrity."
This tension between supporting customization while supporting standards is further complicated by the challenges of scale: in particular, large urban school districts need technology infrastructure, tools, and professional development processes to support teachers district-wide to tailor and deliver curriculum, while still ensuring fidelity to learning goals. In our project with Denver Public Schools, we are developing and evaluating a "mass customization" software infrastructure for teachers. The Curriculum Customization Service enables teachers to mix-and-match digital learning resources to augment, modify, and/or replace their district's core (adopted) curriculum. The Service uses learning goals as the central organizing concept of the interface. Teachers can select a specific learning goal and immediately see the various curricular components they have to work with that support the selected goal. To date, these curricular components include digital versions of the student textbook, digitized components of the associated teachers' guide (formative assessments, teaching tips, instructional resources, and background knowledge readings), and web-based digital library resources. Digital library resources are further broken down by Top Picks (recommended), Interactives/Visuals, Animations, and Inquiry with Scientific Data. Web 2.0-style tagging, rating, commenting, and sharing features enable teachers to share and learn best practices for customizing curriculum from each other.
Educational Technologists. I won't belabor the point, but those of us interested in using technology to support personalized learning continue to be plagued with "last mile" challenges, particularly in K-12 settings. While it is possible to pilot test innovations, conducting large field trials and scaling up innovations remains a huge challenge due to the poor technical infrastructure available in many classrooms. I am not talking necessarily about one-to-one laptops, but even more basic issues. For instance, in some of the classrooms we work in there is a single electrical outlet and the teacher is not allowed to use extension cords or power strips because of fire codes. Any technology requiring power becomes problematic. Similarly, the firewalls employed in many districts to block potentially problematic content are themselves turning into the problem. In addition to complicating research, the lack of infrastructure makes it hard to market new, commercial software-based approaches, and thus hinders the development of a mature and broad market for widespread diffusion of educational technology.
Scale may be on the horizon. There is some feeling amongst those in the business of personalized learning that we are heading for a tipping point, where uptake of personalized learning will rapidly increase and begin to pervade the mainstream. For instance, in "Disrupting Class", Clayton Christensen and colleagues boldly suggest an inflection point around 2012. Certainly the past year has seen an enormous amount of change towards the use of digital content, which I regard as a harbinger for personalized learning approaches. Within K-12 settings, the dual desires of controlling materials costs coupled with the possibility of adapting open content to meet local needs is leading states and school districts to formally advocate the use of Oer. For instance, in a first in the nation initiative, the state of California selected 10 open access textbooks for use in high school science and mathematics this year. In 2008, the Indiana State Board of Education redefined "textbook" as a systematically organized collection of materials, paving the way for districts and schools to use internet resources and interactive materials. A school district in Arizona has eschewed the concept of textbook altogether, and is relying entirely on teachers to find educational materials on the web and integrate these materials into teacher-produced content. Perhaps the time is ripe for personalized learning and the interests of Microsoft in organizing a book and workshop dedicated to the topic further underscore the sea change that is underway.