/Aristotl
Language
Demo

What is knowledge retention?

Updated

Short answer

Knowledge retention is how well people remember and can still apply what they were taught, weeks and months after the session. It is the difference between training that was delivered and training that is working, and the part no completion record registers.

The distance between two records

Most organisations hold one record: someone completed something on a date. What they are asked about is a different thing: whether the rule is being followed now. Retention is what sits between those two, and it is the part nobody registers.

A forgetting curve falling steeply from full recall on the day of the session, then flattening over the twelve weeks after it, while the completion record stays a flat line at the top from a dot at day zero; the space between them is marked as the gap nobody registers.
The record stays at 100 percent from day zero; what is still applied does not.

The law expects the second record, not the first. In the Netherlands the employer must also supervise compliance with the instructions given; in Belgium, training is repeated and adapted as risks evolve, which presumes you know what has faded. The guide on workplace instruction law in Belgium and the Netherlands sets out both duties.

What decay looks like on a floor

It is rarely dramatic. The shortcut that saves ninety seconds and skips a check. The step done in the old order because the old order is muscle memory. The colleague asked instead of the procedure, who answers from the version he learned two revisions ago.

A Dutch warehouse instructs everyone on a new pedestrian route after a near miss. Attendance is complete. Eleven weeks later, the reach truck drivers on nights have gone back to the old crossing point, because it is shorter and nobody has said anything since the session. The instruction happened. The retention did not.

What actually holds it

Three mechanisms do most of the work, and none of them is a longer course. Short knowledge checks that force recall rather than recognition. Repetition spaced over time rather than concentrated in one session. And application: the check is about the decision the person makes at the machine, not the definition of a term.

Measurement can be as modest. A two-question check at four weeks tells you more about a site than a completion percentage, because it separates the sites where the rule stuck from the sites where the session was attended. The guide on tracking completion across locations reads both figures side by side.

Sources

  1. Arbeidsomstandighedenwet, artikel 8
  2. Codex over het welzijn op het werk, art. I.2-21 (FOD Werkgelegenheid)
  3. Ebbinghaus, Memory: A Contribution to Experimental Psychology (1885, English translation)
  4. Roediger and Karpicke, Test-enhanced learning, Psychological Science (2006)

Read next