Before Big Data, Mastery Learning Showed a Way to Personalized Learning

blog-image-mastery

While many work towards the future of personalized computer-based instruction through big data learning analytics, successive approximations can be made towards this goal now.

A great place to start is taking what works in classroom instruction and translating that into computer-based instruction. Consider the design of an instructional system that fuses three teaching strategies:

  1. Mastery Learning
  2. Big Ideas Curriculum Framework
  3. Bloom’s Taxonomy of Educational Objectives

First, Mastery Learning proposes that every classroom should simultaneously run three curricula: a baseline or normative curriculum, an accelerated or enriched curriculum, and a remediation or corrective curriculum. Enabled by formative assessment, student activities flow between the three as needed.

Second, a truly daunting aspect to providing personalized learning pathways is the need to develop many alternative materials and resources to accommodate differing individual needs, learning preferences, and personal interests. But rather than building out many pathways, we can use the Big Ideas Framework to suss out the major inflection points in a course of study. What are the essential questions, the enduring understandings? What are the most common misconceptions? Now we know where along the path to place extra help and where to gain the most value from enrichment.

And third, while Mastery Learning and Big Ideas Framework provide a macro-perspective to personalization, we still need micro-level response to meet students wherever they are on their own continuum of learning and doing. Hence, the integration of Bloom’s Taxonomy allows a finer focus into student needs in terms of responding to issues of comprehension, or of application, or of synthesis and transfer. In essence, Bloom’s Taxonomy functions as a learning activity processor. (Dr. Yungwei Hao of the National Taiwan Normal University wrote a paper about operationalizing a similar type of “learning ecology” in 2004 when we were worked together at UT-Austin).

While still complex to program, such a system provides a viable blueprint to personalized learning. So, even before we can rely on predictive models enabled by “big data” — that is enough data to make reliable inferences upon — we can transform some of our best practices in adaptive classroom instruction into the design of adaptive computer-based instruction.