Evaluate training by matching each intended outcome to a suitable measure. Keep satisfaction, knowledge, confidence, demonstrated skill and workplace application separate, and report how many people were assessed, when and by whom.
Choose the question before the measure
If the objective is clearer coaching questions, a satisfaction survey is insufficient. If the objective is familiarity with a policy, an observed role-play may not be the most efficient first check. Start with the decision the evidence needs to support.
Agree the evaluation plan before delivery so the trainer can allocate time, prepare materials and explain the process to participants.
Keep five different signals separate
| Signal | Possible method | Limit |
|---|---|---|
| Satisfaction | Participant feedback | Does not establish learning. |
| Knowledge | Scenario questions before and after | Does not establish workplace behaviour. |
| Confidence | Self-rating of a specific task | Self-report, not competence. |
| Demonstrated skill | Observed practice with a rubric | A sample in a practice setting. |
| Application | Follow-up examples or manager observation | Other influences may affect results. |
Check whether the assessment fits the timetable
Thirty participants receiving fifteen minutes of individual observation each require 450 assessor-minutes. One assessor cannot complete that in a three-hour block. Two parallel assessors would still need 225 minutes each, before allowing for transitions and breaks.
Peer observation can support learning, but label it as peer observation. Sampling can be useful, but state who was observed and avoid implying full individual coverage.
Use criteria people can see
A short rubric should describe observable actions. For a feedback conversation, these might include naming a specific example, explaining the effect, asking for the other person’s perspective and agreeing a next step.
Provide the criteria before practice. If more than one assessor is involved, discuss how the criteria will be interpreted. Record task behaviours rather than private details shared by a participant.
Report the denominator and the limits
“Confidence improved” is incomplete without the question, scale, timing and response count. Where possible, compare matched responses from the same participants and explain missing data. Do not turn a small voluntary follow-up sample into a claim about the entire workforce.
For the sponsor, provide cohort-level patterns and practical recommendations. Agree separately whether individual results are needed, who may see them and why.
Connect follow-up to an actual opportunity
Ask the manager when participants will use the skill and what support is available. A follow-up check is more meaningful when people have had a chance to apply the learning.
Use the results to decide what comes next: more practice, clearer work instructions, manager support or a change to the programme. Do not promise return on investment without a credible measurement design and relevant business data.
Published by iKON Insight Editorial. Examples are illustrative unless expressly identified as delivered work. This guide supports programme planning; confirm scope, availability and any country-specific requirements for the actual engagement.
How we approach our articles ↗Turn the question into a clear brief.
Bring the audience, dates and intended workplace change to a trainer-fit conversation.
Prepare your brief ↗
