Mid-Course Feedback Questions: Improve an Online Cohort Before It Ends

End-of-course feedback can help your next cohort. Mid-course feedback can still help the people learning with you now. If the demonstrations are moving too fast, the assignments are unclear or learners cannot find the right resources, waiting until the final lesson leaves the current group with the problem.

A mid-course check-in is a small, deliberate pause. Ask what is helping, what is getting in the way and what you can realistically adjust. It is not a popularity vote on the teacher or proof that learners have mastered the subject.

This guide focuses on live or time-bound online cohorts. It includes a seven-question survey, a three-question version and a worked decision example. The sample feedback is invented, not copied from learners or UnsaidBox submissions.

Choose a moment when a change can still matter

The useful moment is after learners have experienced enough of the course to comment, but before the format is effectively finished. In a short cohort, that may be after the first meaningful exercise. In a longer course, it may be near the midpoint. Choose according to the learning sequence, not a rule that every course must survey in the same week.

Cornell’s mid-semester feedback programme uses the same broad principle: gather feedback while there is still time to adapt. An independent online course needs its own practical process; the university programme is not an endorsement of this template or of UnsaidBox.

Before inviting answers, decide what is flexible. You may be able to slow one demonstration, clarify the next assignment or improve the resource index. You may not be able to change the published timetable, remove an essential skill or add unlimited individual coaching. Knowing the boundary helps you make an honest invitation.

Seven mid-course feedback questions

Keep the survey focused on the current learning experience. Make writing optional and provide non-applicable choices when a learner has not reached the relevant activity.

1. How clear is the main idea in the most recent module you completed? Options: Not clear; Slightly clear; Mostly clear; Very clear; I have not completed a module yet. This is perceived clarity, not a test of understanding.
2. How does the pace feel so far? Options: Much too slow; A little too slow; About right; A little too fast; Much too fast.
3. How clear is what you need to do for the next exercise? Options: Not clear; Slightly clear; Mostly clear; Very clear; I have not reached the exercise yet.
4. What is the main thing making progress difficult right now? Options: Finding time; Understanding explanations; Knowing where to start; Technical access; Finding resources; Nothing significant; Something else. Ask for the main obstacle rather than requiring a detailed personal account.
5. What part of the course is helping you learn? Optional text. Encourage a reference to a teaching approach or resource, without asking for identifying details.
6. Where do you get stuck? Optional text. Add: “A step, instruction or example is enough; please do not include passwords, private project data or personal details.”
7. What is one change that would make the remaining sessions more useful? Optional text. This makes the timeframe explicit and avoids an unlimited request for future modules.

For a three-question version, ask what is helping, where the learner gets stuck and what one change would help now. Use that version when the cohort already has many assignments. A small survey you read and act on is more useful than a comprehensive evaluation you cannot review until after the course ends.

Do not repeat the entire final evaluation at the midpoint. The final guide can assess the completed experience and future improvements. This check-in owns the narrower decision: what can change for the current learners in the remaining time?

Make it safe to say “I do not understand”

A learner may worry that criticism will affect how you see their effort. State that the form is optional and is for improving the learning experience. If you grade or assess people, separate feedback from assessment and do not ask them to attach work that identifies them.

We still have time to improve how the rest of this course runs. Please tell me what is helping and where you get stuck. This feedback is optional and is not an assessment. You do not need to include your name. I will review the answers privately and share the changes I can make, not individual responses. For account access or help with a specific submission, use [support route].

Use that promise only if it matches your process. Keep sharing disabled on the form. In a very small cohort, a distinctive project or personal circumstance may still identify someone, so encourage learners to describe the confusing step rather than their whole situation.

Do not insist on anonymity where a named support request is necessary. A learner who cannot log in may need an account-specific fix. Give them a separate route rather than trying to turn anonymous feedback into a helpdesk ticket with guessed identity.

Give learners a clear way to respond

Share the link in the course channel learners already use, and allow time to complete it away from a live call. You can set aside a quiet moment, but avoid watching individual completion or asking people to announce that they submitted.

Before the next session, please use this optional check-in to tell me what is helping and where the course could be clearer: [link]. I will review responses by [date] and explain any changes before [session]. You can skip questions that do not apply. Critical feedback is welcome.

Include people who missed the live session if the form concerns resources they can access asynchronously. Keep non-applicable options so they do not have to invent ratings. If the invitation reaches only the most active learners, record that limitation instead of assuming silence means everyone else is doing well.

Distinguish a learning obstacle from a preference

Read answers twice. First, understand what is being described. Second, separate the obstacle, the request and the action you might take. A learner’s proposed solution can be useful, but it is not always the only way to address the problem.

“Make all sessions shorter” may describe cognitive overload. “Add more examples” may mean the current examples skip an important intermediate step. “Give us more time” may concern a deadline, the pace of a demonstration or the amount of independent practice. These are different problems.

Use a private review note with five fields: theme, evidence summary, proposed change, owner and review date. Keep raw personal details out of that working summary. Count responses mentioning a theme, not the number of times a word appears.

If a sender opts into private follow-up, ask a focused clarification: “At which step does the example stop being clear?” Do not ask who they are or which private project belongs to them. They return using their secret link; this is not an automatic email exchange or live support channel.

A worked example: the course is not necessarily too difficult

Invented scenario: a six-session cohort receives 16 feedback submissions after session three. Several describe the practical exercise as confusing. Some ask for slower teaching. Two say they cannot find the files used in the demonstration. The same submission can contain more than one of these points.

Before removing a topic, inspect the exercise instructions. You discover that the instructions refer to an earlier file without linking to it, and the demonstration moves from setup to the finished result without showing an intermediate step.

A proportionate change is to add the resource link, show the missing step and begin the next session with one complete example. Keep the learning goal. You have addressed a plausible access and explanation problem without assuming that learners cannot handle the subject.

Afterwards, ask one short question about whether the revised instructions made the exercise clearer. Do not claim that a higher clarity rating proves learning improved. If you need to know whether learners can perform the skill, review an appropriate practice task separately from the feedback survey.

Decide what to change, explain and leave alone

• Fix immediately: broken links, missing files, conflicting instructions and other clear obstacles.
• Adjust thoughtfully: pacing, worked examples, practice time and the order in which resources appear.
• Explain: essential but challenging activities, fixed timetable constraints or why a requested shortcut would remove the skill being taught.
• Save for the next cohort: structural changes that would create confusion or unfairness if introduced halfway through.

Do not promise to implement every suggestion. Learners may want opposite things. Where possible, separate optional depth from the main path: keep the core explanation clear and offer extra practice for those who need it. Describe the trade-off without singling out a respondent.

Run the check-in with UnsaidBox

1. Create your account and open Forms and Feedback.
2. Create a form for this specific cohort check-in. Add the selected choice and text questions and an honest introduction.
3. Keep sharing disabled for private learner feedback. Make non-applicable questions optional and test the form on a phone.
4. Finalise the question structure before publishing. Share the form link through your existing course communications.
5. Review responses before the date you promised. Use Pro CSV exports if helpful for private organisation, but do not imply that export automatically evaluates learning.

The current form builder supports text, single-choice and 1–5 ratings. Use a choice field when you need labelled options or “Not yet.” A separate short form can support a later review round; check the current plan allowances rather than assuming that every old link should remain active indefinitely.

Tell the cohort what happens next

Thank you for the check-in. I’m adding clearer resource links and a slower walkthrough of the exercise setup. We are keeping the main learning goal and the remaining session dates. The next session will begin with the missing intermediate step. I’ll ask whether the new instructions helped after you have tried them.

This closes the loop without publishing private answers. It also makes the feedback visible as a service improvement rather than a request that disappeared into a dashboard. If no change is possible, explain why and what help remains available.

Frequently asked questions

Does a midpoint survey replace the final evaluation?

No. The midpoint check-in supports immediate adjustments. A final evaluation reviews the completed experience. Keep each short and focused on its own decision.

What if only a few learners respond?

Use specific feedback to identify issues worth checking, but do not call it the view of the whole cohort. Confirm practical problems yourself where possible and avoid percentages that make a small response set sound conclusive.

Should I ask learners to rate the instructor?

For immediate improvement, questions about explanations, exercises and access often lead to clearer actions. Broad personal ratings can be difficult to interpret and do not establish teaching effectiveness on their own.

For the end of the programme, use our course evaluation question guide. For now, create your free UnsaidBox account and ask the current cohort what would make the next session more useful.