Learn · Training

How to measure whether training was understood

Completion only proves that a module played to the end. To measure understanding, put knowledge checks inside the training, at the steps that matter. Then track first-attempt accuracy, the number of attempts, what learners rewatched, and where they stopped. Together, those signals show which part of the training needs work.

By 4K Labs · Updated October 6, 2026

Why completion rates mislead

A learner can finish a video with the sound off in another tab. A quiz at the end measures what they recall a few minutes later, and it rarely tells you which step confused them. Most learning systems report exactly those two things: finished, and passed. Neither one tells you what to fix.

Five signals that show understanding

1. First-attempt accuracy. The share of learners who get each check right the first time. A low score flags the step before it.
2. Attempts to correct. How many tries people need. Three attempts on one check is a sign the explanation was unclear.
3. Rewatches before a check. When many people scrub back to the same moment, that moment is doing too much work.
4. Drop-off points. The exact second people leave. A cluster at one step usually means that step is too long or too hard.
5. Paths chosen. In scenario training, which option people pick before feedback tells you what they would really do on the job.

Put the checks where the learning happens

A check works best inside the moment it tests. In a safety module, that means asking the learner to tap the hazards in the warehouse scene itself, not to pick "spill on the floor" from a list afterwards. In software training, it means trying the step in a working copy of the app.

The research supports doing over watching. A review of 225 studies found that students in lecture-only classes were 1.5 times more likely to fail than students in classes with active learning, and that active learning raised exam scores by about 6% (Freeman et al., PNAS, 2014). In a controlled study of e-learning, interactive video improved learning and satisfaction, while the same video without interaction did not (Zhang et al., 2006).

Getting the data into your LMS

SCORM, the older e-learning standard, reports completion, pass or fail, score and time to a learning management system. xAPI is the newer standard for recording any learning activity as statements such as "Priya answered the lockout question correctly on the second attempt". It can carry each interaction, not only the final score. 4K Labs sends each learner's interactions as xAPI statements. For the differences in detail, see interactive video vs. SCORM.

A report to start with

For each module, track four numbers: completion, first-attempt accuracy for each check, the three most-rewatched moments, and the biggest drop-off point. Review them after the first cohort, fix the weakest step, and compare the next cohort.

Questions