Interaction frequency measures how often learners perform an active action — answering a quiz question, clicking a hotspot, submitting an annotation, raising a hand in a live session, or posting a discussion reply — within a defined time window or per unit of content length. It distinguishes between a video that is merely watched and one that generates genuine cognitive processing, and it underpins the concept of active learning in asynchronous formats. Each interaction is captured as a distinct xAPI statement — for example, "learner answered question" with a result extension, or "learner interacted with hotspot" — and stored in an LRS (Learning Record Store), making it straightforward to compute frequency per learner, per video, or across a course cohort. The metric correlates positively with learning outcomes in controlled studies: learners who answer more questions during video viewing score higher on delayed post-tests than passive viewers of the same content. The instructional design implication is to embed check questions or interactive elements at least every five to seven minutes in longer videos. The frequency data should feed back into a redesign cycle if interactions per minute fall below the target for a given content type. One trade-off to watch is that artificially inflated interaction frequency — adding trivial click-throughs to boost the metric — inflates numbers without improving learning.
