Engagement in the context of learning video refers to the degree to which a learner is actively interacting with content rather than passively receiving it, and it is measured as a composite of multiple signals rather than a single number. Passive signals include watch-time and completion rate; active signals include answering in-video quizzes, submitting annotations, posting questions, re-watching segments, and clicking on hotspots — each of which can be captured as an xAPI statement and stored in an LRS (Learning Record Store). The distinction between passive and active engagement matters because research consistently shows that active retrieval and interaction improve retention more than equivalent time spent watching. Platforms typically aggregate these signals into an engagement score or index, weighting active events more heavily than passive ones, but the exact weights are an internal design decision with no universal standard. A common gotcha is conflating high engagement with high learning: a learner who frantically re-watches the same confused segment shows high engagement but low comprehension. Engagement metrics therefore work best as early-warning signals to trigger instructor intervention or content revision, not as standalone outcome measures. The most useful engagement analysis combines signal type with timestamp, so designers can identify not just that learners disengaged but exactly where in the video the disengagement occurred.
