One of our customers got a question they did not expect. A shareholder in another country wanted to see for themselves that the people doing sensitive work understood the rules they were working under.
They had the reports. Almost every employee had completed every mandatory module. The reports included dates and names, and everything could be exported. The answer that came back was that this was not what had been asked for.
Completion tells you that the training happened. It does not tell you whether it worked. That difference is worth understanding before somebody asks you to demonstrate it.
Completion and comprehension are different records
A completion record shows that a person opened a training module and reached the end. A comprehension record contains evidence of what the person understood in a form that somebody else can review.
They answer different questions. One cannot be used to infer the other.
Why most training reports cannot answer the question
A common compliance module contains a video followed by multiple-choice questions. The learner watches the video, answers four questions, reaches a pass score and receives a certificate.
That format scales. It also has a limit that reporting cannot remove: recognising an answer is not the same as understanding it.
When four options are visible, the answer is visible too. People can use test-taking patterns without understanding the subject. The longest answer looks plausible, absolute wording looks suspicious, and two options can often be dismissed immediately.
The stored record says that a user selected option B and that B was correct. It records a choice. It gives a third party very little evidence from which to form an independent judgement.
The coffee problem
We started calling this the coffee problem because that is what it looks like in practice.
The video starts. The learner goes to the kitchen, returns when the progress bar is full and answers questions they could have answered before the video began. Nobody is being dishonest. The system cannot tell the difference between somebody who paid attention and somebody who did not.
A completion rate of 95 percent remains a completion rate of 95 percent even when half the group was making coffee. The number can be accurate and still be useless for judging comprehension.
What stronger evidence looks like
An open question asks the learner to remember the idea, decide what the situation calls for and explain it in their own words. The useful record is the answer itself, not only the score attached to it.
- 01
Another person can review it
A compliance lead, auditor or shareholder can read the answer and form their own view instead of trusting a score without context.
- 02
The reasoning becomes visible
Two people can reach the same conclusion for different reasons. One of those reasons may expose a gap that the final score hides.
- 03
Group and individual gaps separate
A distribution can show that 30 percent chose option C. Open answers can show why they chose it and what they misunderstood.
- 04
Timing and revision add context
Time spent, revisions and the path to the answer can add useful context. None of those signals is conclusive on its own.
The honest tradeoff
Open answers have to be evaluated. Some scoring can be automated. The part that cannot be automated requires somebody with subject knowledge to read what people wrote.
That can be manageable for a small group of new joiners. Across a large workforce, it is a real cost that needs an owner and a plan.
Not every training item needs this treatment. A recurring refresher on a low-risk policy can be a video and a short quiz when the goal is to inform people.
Use stronger evidence where being wrong is expensive and the correct behaviour depends on judgement in a situation, not only on remembering a rule.
What to put in place
If you need to demonstrate that somebody understood an obligation, start with four practical decisions.
- 01
Choose where stronger evidence is justified
Start with the training that carries real risk. Make it a deliberate list instead of applying one format to everything.
- 02
Write scenario questions
A definition question tests recall. A realistic situation asks the learner to recognise the situation, decide what to do and explain what they would record.
- 03
Keep the answers
If the system stores only pass or fail and discards the written response, it throws away the part that another person can independently assess.
- 04
Make the export readable without explanation
A useful record should include the person, date, question, answer and relevant scoring context so an authorised reviewer can form their own conclusion.
The question worth asking
You probably already know your completion rate.
If somebody asked you tomorrow to demonstrate that a specific person understood an obligation, what would you be able to show them? A date and a checkmark may not answer the question being asked.
Where CapoFine fits
Match the learning activity to the evidence you need.
Grow supports company-approved learning, open-text questions, interactive scenarios, formal assessments, progress and reporting. Keep practice, feedback and formal assessment clearly labelled so learners and reviewers know what each record means.