Technical Challenges on the Path to Personalization

Personalized learning will boost engagement, persistence, and achievement. That seems to be one of the most prevalent memes of our time (and the basis for the 2011 book Getting Smart).

15 years ago the meme was aligned instruction. Rather than alignment around an age cohort, the new bet is alignment around every student, individual learning progressions that leverage student interests and address student needs.

As summarized last year, we see evidence for the personalized learning thesis in four tech-enhanced categories:

  1. Results from next generation school models (Aspire, KIPP, Summit, Mooresville , etc.)
  2. Technology-enabled math products have boosted achievement (DreamBox, ST Math)
  3. Technology-enabled literacy products have boosted achievement (i-Ready, Read 180, Waterford)
  4. Studies of online and blended learning show efficacy.

Fundamental to both achievement and completion rates is student engagement. Ace Parsi from NASBE notes that, “Student engagement represents the capacity and inclination for students to take ownership of their past, present, and future educational experiences by enlisting their cognitive, behavioral, and emotional investment in learning.”

Successful school networks focused on engagement, including Big Picture, Edvisions, Envisions, International Studies, and New Tech Network, focus on authentic work and demonstrations of learning. With advice from experts at the Buck Institute for Education, these networks engage students in projects that promote deeper learning. We profiled 20 of these schools and saw teachers working hard, often in relatively low-tech environments, to personalize learning and help students build portfolios of personal bests.

The combination of these best practices and new tools gives us optimism that personalized learning will be adopted at scale and will lead to the anticipated gains. But first, there are a few technical and implementation challenges that need to be solved.

Definitions. Before we dive into the challenges, a few definitions. We believe a comprehensive learner profile is key to personalization. It includes a:

  • Data Backpack: An expanded common electronic student record; an official transcript that follows students through every transition, grade to grade and school to school. The Backpack includes traditional transcript data such as demographic information, state testing data, and supplementary student supports. However, it would also include additional information in order to represent a more holistic picture of a student’s work and achievement, such as a grade book of standards-based performance data and a portfolio of personal best work samples, and better capture the student’s progression at any moment in time.
  • Learner Profile: An expanded learner profile builds on the data backpack to provide additional clues to unlock learner needs, preferences, and potential. While each student’s Data Backpack would be populated by a set of common elements for all students at a new minimum level, the components of each student’s Learner Profile could be customized based on student needs, platform data requirements, and family decisions. It would track college and career readiness and could include a narrative discussion of learner assets and challenges. The postsecondary version could feature financial data: cost to complete, aid opportunities, and payment schedule.
  • Student Portfolio: A learner curated collection of personal best work products as well as a record of work experience and community service. It could include, as some colleges offer, a domain and a blog that the student can leave with.

Technical Challenges. There are five challenges to overcome to unlock the power of personalized learning:

  • Student record. Each state will need to define a common student record for a portable data backpack. More broadly, a learner profile will be rather fluid with lots of opportunity to customize.
  • Interoperability. A common data standard, like IMS’s Learning Tools Interoperability (LTI), will guide how information is shared between systems.
  • Combining formative: Many U.S. schools benefit from more information from many sources of formative assessment, often embedded in digital learning experiences, but have no way to combine the information from multiple sources in ways that are useful for driving instructional improvement or managing student progress. The solution probably involves tagging content and assessment data (the way Houston requires partners to use Thin Common Cartridge). However, “Tagging has to be considered carefully,” says Dan Ingvarson who built the tagging scheme for Australia but has also seen several generations of tagging go down the drain with the introduction of new standards.
  • Estimating growth. As more students progress on personalize pathways, it will become necessary to develop comparable growth rates to ensure that all students are making adequate progress. Current scales (like Lexile/Quantile) are frequently used to compare growth. New more subtle measures aligned with new standards would be even better.
  • Correlation. Better use of formative assessment (and less reliance on big year end summative tests) will require comparable achievement and growth rates so student learning can be compared from school to school. In addition to tagging, post hoc data barrage can correlate data sequences from different environments after the fact by analyzing thousands of data points.IMS’s Caliper Analytics standards support both post hoc and real-time data feeds for millions of students daily.

Storage won’t be a big problem because learner profiles wouldn’t actually carry all the data around, they would used linked data to see information from many different systems.

These are imminently solvable problems–they are more politically than technically challenging. Solutions will take a little leadership from industry groups, EdTech leaders, and foundations.

For more see:

This post is a part of a Student Data Backpack blog series in the upcoming “Getting Smart on Personalization and Privacy” Smart Bundle produced in partnership with the Foundation for Excellence in Education’s Digital Learning Now initiative (@DigLearningNow) and the Data Quality Campaign (@EdDataCampaign). Join the conversation on Twitter using #EdData.

Tom Vander Ark

Tom Vander Ark is the CEO of Getting Smart. He has written or co-authored more than 50 books and papers including Getting Smart, Smart Cities, Smart Parents, Better Together, The Power of Place and Difference Making. He served as a public school superintendent and the first Executive Director of Education for the Bill & Melinda Gates Foundation.

Discover the latest in learning innovations

Sign up for our weekly newsletter.

1 Comment

Benjamin L. Stewart

When I hear the term "personalized learning" I am instantly drawn to others terms like the dilemma of choice ( and the paradox of choice ( In my view, it's more about differentiated instruction and negotiation among individuals (all actors who have a role in the educative experience). For me, a word association around "negotiation" generates terms like "consensus", "compromise", "higher-order of thinking", "rationale", decision-making with a purpose", "leadership", "sharing", etc.

Leave a Comment

Your email address will not be published. All fields are required.