Assessing Understanding

We believe that a single kind of assessment tells a single kind of story — and it is almost never the whole story.

Here is something that does not get said often enough in conversations about education: how knowledge is assessed shapes what knowledge can be demonstrated. This is not a philosophical position. It is a finding from decades of cognitive science research on how memory is stored and retrieved. When a learner acquires knowledge through hands-on experimentation, collaborative problem-solving, or physical exploration, the brain encodes that knowledge together with the sensory, emotional, and contextual details of the experience — the sights, sounds, social dynamics, and physical actions that surrounded the learning. The brain encodes environmental details alongside the actual information being learned, creating multiple retrieval pathways, and when students encounter similar environmental cues later, these activate the associated memories through a process called cue-dependent retrieval. The implication is significant: knowledge learned through building, doing, discussing, and experimenting may simply not be retrievable in response to a written test — not because the understanding is absent, but because the retrieval cues are mismatched.

This principle is known in cognitive science as encoding specificity, and its classroom implications have been studied and confirmed across many contexts. Memory performance is not simply a function of how deeply information is encoded, but how well the encoding processes match the conditions of retrieval (Conti, 2025). This is sometimes called Transfer-Appropriate Processing — and it means that a student who learned something by doing it may struggle to retrieve that learning when asked to recall it in writing, not because they failed to learn, but because the retrieval context does not match the encoding context. Students can learn in one context and fail to transfer knowledge to another (Smith, 2021) — and when that happens in a traditional testing environment, educators and students alike too often interpret the failure as evidence of insufficient learning rather than what it actually is: a mismatch between how knowledge was built and how it was asked to show itself.

The National Research Council’s landmark synthesis of learning research underscores this directly. How tightly learning is tied to contexts depends on how the knowledge is acquired. Research has indicated that transfer across contexts is especially difficult when a subject is taught only in a single context rather than in multiple contexts. When a subject is taught in multiple contexts, and includes examples that demonstrate wide application of what is being taught, people are more likely to abstract the relevant features of concepts and develop a flexible representation (National Academies., 2000). This applies symmetrically to assessment. When learners are only asked to demonstrate understanding in one context — a written test, a multiple-choice exam — the range of what can be revealed is narrow by design. Knowledge resources that were built through movement, conversation, investigation, and creation remain inaccessible to a format that was never designed to reach them.

This narrowing has real consequences for what we think we know about a child. A major difference between performance-based assessments and other forms of assessment is that performance-based assessments require students to demonstrate their knowledge in a particular context (Myers, 2021) — and choosing that context thoughtfully changes what becomes visible. Research consistently finds that students who appear to struggle on traditional tests often demonstrate sophisticated understanding when given an opportunity to show what they know through building, explaining, designing, or applying. Expecting all students to demonstrate knowledge or skills in the same manner can become a barrier to learning — and fairness results when all students can show that they have mastered the learning goal without encountering barriers from the assessment design (CAST, 2018). A child who learned something by doing it should have the opportunity to show it by doing it. The assessment should reach back toward the original experience, not demand a translation into a format that was never part of the learning.

At Wayfinders, this is exactly why mastery is demonstrated through multiple kinds of evidence, in multiple kinds of contexts, over time. Missions do not end with a test. They end when a learner has shown — in at least two distinct ways — that the understanding has taken root and can be expressed. A student might build a model, write a reflection, teach a peer, solve a novel problem, or defend a decision in discussion. Each of these draws on the same underlying understanding through different retrieval pathways and different contextual cues. Approaching mastery of any given competency requires progression along a continuum of understanding and demonstration of learning across new contexts and situations (Willis et al., 2022). The badge a student earns at Wayfinders is not a record of one performance on one day. It is a body of evidence that understanding is real, transferable, and available — not just in the context where it was learned, but across the many different contexts where it will eventually be needed.