What is knowledge retention?
Knowledge retention is what a learner can still demonstrate some weeks after completing a module, as distinct from what they scored on the day. It is measured by assessing the same competencies again after a deliberate interval. Retention is the property an organisation actually cares about, and it is the one least often measured.
Also called: Learning retention
Why the score on the day is the least useful number available
Completion and retention are treated as the same achievement in almost every training report, and the gap between them is where the risk lives. A learner who passed an assessment immediately after reading the material has demonstrated short term recall of a document still open in front of them. Whether they can apply the procedure during an incident two months later is a separate question, and answering it requires a second assessment that most programmes never schedule. The forgetting curve is well established, so the absence of retention data is a choice rather than an unknown.
- A score obtained immediately after reading measures recall rather than retention.
- The interval between completion and application is where knowledge decays.
- Retention data identifies which content needs reinforcement rather than repetition.
- Annual recertification assumes a retention period nobody has measured.
How knowledge retention works
- Competencies are selected where decay would carry operational consequence.
- A delayed assessment is scheduled at a stated interval after completion.
- Items are drawn from a different part of the pool than the original attempt.
- Results are compared against the same learner's score on completion.
- Content covering the weakest competencies is reinforced and reassessed.
How knowledge retention is evaluated
Retention is established by assessing the same person twice at a known interval. Measure the following for each competency assessed:
- score difference between completion and the delayed assessment
- retention by interval length, where several intervals are used
- variation in decay between competencies in the same module
- proportion of learners completing the delayed assessment at all
- effect of reinforcement on the subsequent delayed score
Limitations, risks and common failure modes
- Delayed assessments have poor participation, which biases the figure upwards.
- Learners who remember the original items score on familiarity rather than knowledge.
- Retention is measured on volunteers rather than on the full cohort.
- A single interval is chosen and reported as though decay were linear.