One critical element common among competency based education, mastery based learning and data-driven instructional design is identifying and aligning learning outcomes – or student-centered measurable statements of what is being learned and assessed to instructional content, activities and assessments. This alignment – which we call the course design triangle – helps ensure that no skill is “orphaned” and students have ample practice opportunity to master learning objectives and demonstrate proficiency against component skills.
Developing this type of effective outcomes model starts with a Skill Graph or map of what students should know or be able to do by the end of a module, lesson or instructional session. Starting with high-level competencies and then moving down in granularity to learning objectives and then to skills, or discrete knowledge components contained within each learning objective, the Skill Graph becomes the course blueprint.
By thinking about and connecting skills to learning objectives in this way, the process of developing lesson content and formative practice against each skill, and then by extension, summative assessments is greatly simplified – and codified. Once the course is running, and data is collected against each specific learning objective, it becomes possible to quickly see where in a course there might misalignment between content, activities and assessments and target improvements.
How does the skill graph enable adaptive learning?
Because competency-based education and mastery-based learning involve not just mastering skills, but also having students demonstrate that they can apply the skills in practical ways, it’s essential to have both clarity about the skills being measured, and a way to actually measure the skills. The skill graph is the link between the what the student is doing, and what the learning model – or predictive inference engine embedded in the course – is measuring.
As students “learn by doing” the platform is capturing learning data and making a predictive estimate for each student against every learning outcomes. If a student demonstrates mastery against a specific learning objective, for example, then there’s no need for further practice of those skills. If the student hasn’t mastered the LO, the platform generates more practice – individualized for them, based on their demonstrated skill gaps.
What does a skill graph look like?
How might a typical skill graph look? Consider this example from a biology course. To pass the course, students must be able to analyze the inter dependencies of organisms and their environments. That’s the highest-level competency students must master and be able to demonstrate.
For this competency, students have two learning objectives: (1) analyze how organisms function within their environments and (2) compare and contrast types of interactions and relationships between species within a community.
Students won’t be able to compare and contrast interactions between species without a range of more granular skills. One of these component skills is recognizing and understanding predatory behavior.
In this skill graph, this skill is “tagged” to formative activities about predators and prey; say, owls and mice. As the student completes each question about the predatory relationship between owls and mice, the Acrobatiq inference engine is making a dynamic and predictive estimate about whether students are successfully learning the objective on their way to being able to compare and contrast relationships between species. If not, the platform will generate more practice about predatory behavior – which might also include robust hints and targeted feedback to help students get unstuck.
And if they are still not mastering the skill? Then the instructor has a signal that they may need to spend more time on this concept in class or assign additional learning resources. If this too is insufficient at helping students master the concept, they can be connected to “live” help such as office hours, a TA, or course mentor.
Developing an effective skill graph
Every Acrobatiq adaptive courseware in the content library includes a robust and student-tested skill graph. But more importantly, with Acrobatiq Smart Author, faculty and instructional designers have the ability to build and modify skill graphs for individual courses.
To generate a well-developed skill graph, it helps to make sure an instructional designer, learning engineer, or some other expert with a background in learning science is involved in the course design process. While identifying a learning objective is straightforward (maybe an instructor wants students to learn to add fractions), breaking that objective down into its component skills (such as finding a common denominator or simplifying the added fraction) is more difficult.
Most skill graphs need to be tested and fine-tuned before they can produce the best results in a course. An instructional designer may decide, for example, that one skill — such as simplifying fractions — needs to be broken down into two skills (such as determining if a fraction can be simplified and simplifying the fraction).
A skill graph is an essential but often overlooked component of a well designed and developed course that can save educators considerable time once the course begins enrolling students. If course creators take time to build and refine a quality skill graph, the result will be an adaptive learning course that correctly evaluates and supports students’ progress toward the course goals and enables students to reach their best possible outcomes.