AliveSim.ai

Measurement

Can a pre/post test measure behavior change?

How vignette-based pre/post tests, paired with scenario situations that are similar to the vignettes but never identical, can measure changed performance instead of recall.

Yes, when the design has three parts. First, the pre and post tests are vignettes: mini-situations with context, a decision, and plausible options, so the instrument measures the learner's judgment in a situation rather than recall of content. Second, the learning activity between the tests contains situations similar to those vignettes but never identical: the same concept in a messier, more ambiguous, less structured form. Third, the changed post-test response then demonstrates a changed performance, not a larger store of facts.

The similar-but-never-identical requirement is the part most programs miss, and it carries the whole claim. Identical situations train recognition, so the pre/post gain measures rote memory. Similar-but- different situations train the concept itself, so a changed vignette response means the learner can now apply it where they could not before.

Most pre/post tests measure only knowledge because they are built from recall items and the activity between them delivers content rather than situations. The instrument every stakeholder already accepts can carry a much stronger claim. It just has to be built for it.