WF-06 — Student Feedback Collection, Response and Closure

KPI 2.6 is collected in two different ways at the two levels, and IoL controls the one that matters to it. At institution level the Ministry runs its own Student Experience Survey and MBRU supplies contacts and chases responses. At programme level the number comes from MBRU's own course evaluations, run at the end of every course, to a structure the Ministry specifies. That makes the programme-level KPI a direct product of IoL's own operational discipline, and it makes a null Course evaluations score the single most likely way to lose it.

1. Purpose

To ensure that every IoL learner is asked about their learning experience at the end of every course through an instrument matching the Ministry's specified structure, that the responses and their raw per-question detail and dates are retained in a form the Ministry can demand without notice, that the resulting scores reach the programme record complete and on time, and that learners see what changed as a result, so that feedback is a loop rather than a collection exercise.

2. Scope

Scope statement. This process manages student feedback from the design and scheduling of a course evaluation, through issue, response chasing, analysis and score computation, to the publication of the "you said, we did" response and the retention of the raw data.

Applies to. PGDipHPE and MScHPE. Every course taught in every academic period. Also covers IoL's contribution to the Ministry's institution-level Student Experience Survey (SES), which is contact supply and response chasing rather than instrument design.

Does not apply to. Employer and workplace supervisor feedback, which is WF-14. Alumni surveys and graduate outcome capture, which is WF-25. Individual student complaints and appeals, which are WF-11, though the aggregate of those is an input to this workflow's interpretation. The decision to change a course in response to feedback, which is WF-03, or a programme, which is WF-01.

Applicable requirements. OBEF guide v11.5 KPI 2.6; Appendix B evidence provisions for course-evaluation surveys; Appendix C response-rate and small-sample rules; HEDB Data Dictionary 2026 definitions for Courses.xlsx and Students - Enrollment.xlsx; MBRU data protection obligations for student contact and response data [IoL to confirm the retention period MBRU policy permits for identifiable survey data, since section 11 assumes a minimum aligned to the Ministry's request window].

The two instruments, distinguished, because they are constantly conflated.

Question design, which applies to the programme-level instrument. The Ministry sets the structure. Questions carry equal weights and cover teaching quality, curriculum relevance, skills acquisition and the overall learning environment.

3. Trigger, boundary and endpoint

4. SIPOC

5. Accountability

Process owner. IoL Director of Programmes, or the role able to require a course lead to act on feedback and to hold a course score to account.

Escalation.

6. Process steps

1. Confirm coverage before the period begins. List every course in the academic period from the course inventory and confirm an evaluation is scheduled for each. The guide is explicit that an evaluation is run for every course taught, not a sample. [CAPTURE] the coverage list with the period. Serves 2.6.

2. Confirm the instrument matches the Ministry-specified structure. Equal question weights, covering teaching quality, curriculum relevance, skills acquisition and the overall learning environment. [CAPTURE] the instrument version in use, dated. Serves 2.6. Decision point. If the instrument in use differs from the specified structure, the resulting scores may be rejected. Fix the instrument before the period, not the scores afterwards.

3. Decide and document any Appendix C grouping, before any survey runs. Where the population for a programme is too small for the instrument, Appendix C permits merging programmes sharing degree level and graduation year, similar content area, or similar employment fields. PGDipHPE and MScHPE plainly satisfy the content-area and employment-field principles. [CAPTURE] the grouping decision, its rationale, the date and the second signature from Quality and IQA. Serves 2.6. This is the single most timing-sensitive act in the workflow. A grouping decided before the survey is a sanctioned small-sample remedy. The same grouping decided after the results are visible is a manoeuvre, and the charter puts it in the right-hand column by name.

4. Issue the evaluation at course end. Record the issue date. [CAPTURE] survey issue date. Serves 2.6 and the Appendix B metadata requirement.

5. Chase responses. At least two reminders. Response rate, not score, is the binding constraint for a small postgraduate cohort. Techniques that are legitimate: timing the issue when learners are still engaged, in-session completion time, reminders, and explaining what happened to last year's feedback. [CAPTURE] reminder dates and the achieved response rate. Serves 2.6 and Appendix C.

6. Close collection and compute the score. [CAPTURE] Course evaluations score per course, and response-collection date. Serves 2.6.

7. Retain the raw data. Per-participant, per-question responses, participant contact data, the survey issue date and the response-collection date. [CAPTURE] the retained export with a checksum or an equivalent integrity marker. Serves 2.6 and Appendix B. Appendix B explicitly permits the Ministry to demand exactly this, which makes retention mandatory rather than good practice. A score submitted without retained raw data is a claim IoL cannot evidence.

8. Analyse themes across courses and across the programme. Separate what a course lead can fix from what requires a course specification change (WF-03) or a programme change (WF-01).

9. Agree actions with named owners. One owner per action, with a date. Feedback that generates no action generates no future responses either, and response rate is the constraint.

10. Publish "you said, we did". Per course where the volume supports it, and per programme in aggregate. Publish before the next evaluation is issued, because the response rate for the next survey is decided by whether learners believe the last one mattered. [CAPTURE] the published statement with its date.

11. Populate the CHEDS record. [CAPTURE] Course evaluations score, Course code, Course Ref Number, Academic Period, Student Numbers. A null score in any course row is the failure mode that produces an unreported KPI. Serves 2.6.

12. Support the institution-level SES. Supply Enroll_Student_Email_Address and the enrolment fields that let the Ministry apply the two-semester eligibility filter, and chase responses within the collection window. [CAPTURE] the contact file supplied and its date. The two-semester rule matters for a short postgraduate programme. Learners who have not completed two semesters are not eligible to participate, which for a one-year programme can exclude a large part of the cohort. Confirm eligibility from Enroll_1st_Academic_Period rather than assuming it.

13. Carry themes and actions to WF-04. The annual monitoring cycle owns the aggregate loop and the classification of any shortfall.

7. OBEF data generated

Capture rule. Four things cannot be reconstructed after the collection window closes, and all four are named in Appendix B: the raw per-participant per-question responses, the participant contact data, the survey issue date and the response-collection date. A mean score can be recomputed from raw data; raw data cannot be recomputed from a mean. If the survey platform is configured to discard responses after aggregation, that configuration is an evidence failure waiting for a request, and it should be changed before the next survey rather than after the first request.

Reproducibility test. A second analyst can reproduce the programme-level KPI 2.6 value by averaging Course evaluations score across the programme's courses for the most recent academic year, provided every course row is populated. They can verify it against the raw data only if the raw data was retained with its dates. On the assumption stated below, they could probably do the first for some courses and the second for none. That is the capture defect.

Assumption, stated as an assumption. This document assumes that course evaluations run at IoL but that coverage is not verified as complete across every course in every period, and that raw per-question data with issue and collection dates is not systematically retained in an exportable form. [IoL to confirm both, and to confirm which survey platform is in use and whether it retains per-respondent data.]

8. Max-score design

The structural point about this KPI. The score itself is unlikely to be IoL's problem. Coverage and retention of evidence are. A programme averaging 4.5 across every course scores 75; the same programme averaging 4.7 across two thirds of its courses, with the rest null, risks producing a number the Ministry can reject. The marginal point is in the null cells, not in the mean.

What this workflow must do to reach the target.

1. Run an evaluation for every course, every period. The guide is explicit and there is no sampling route.
2. Populate Course evaluations score for every course row. A null is not a low score; it is an unreportable KPI.
3. Retain raw per-question data with issue and collection dates, because Appendix B names it.
4. Decide any Appendix C grouping before the survey runs, documented and dual-signed.
5. Publish "you said, we did" before the next survey, because it is the only reliable lever on response rate.
6. Confirm the instrument matches the Ministry-specified structure with equal question weights.

[REDESIGN] actions.

A caution about R6 and about incentives generally. Improving response rates through reminders, timing and legitimate incentives is on the legitimate side of the charter's line. Selecting who is surveyed is not. In-session completion time is a timing intervention offered to everyone on the course, which is why it is legitimate; issuing the survey only to a subset of learners would not be, whatever the reason.

9. Indirect strategy where data cannot be collected

Two categories apply here, and they are different problems with different remedies.

Category 2, the population is too small for the instrument. IoL's cohorts are small. For KPI 2.6 at programme level the evaluation base is a course cohort, which for a postgraduate health professions education programme may be a handful of learners, and for the institution-level SES the eligible population is further reduced by the two-semester participation rule, which for a one-year programme may exclude most of the cohort outright. Appendix C provides the sanctioned remedy: merge programmes sharing degree level and graduation year, similar content area, or similar employment fields. PGDipHPE and MScHPE plainly satisfy the content-area and employment-field principles, being postgraduate health professions education awards taken by practising health professionals working in the same field.

The condition on that remedy is timing, and it is absolute. The grouping must be decided and documented before the survey runs. The charter's legitimacy table draws the line in one sentence: grouping small cohorts per Appendix C, decided and documented before the survey, is legitimate; revising a grouping after seeing which grouping scores better is not. Because the two look identical in the final submission, the only thing that distinguishes them is a dated record made in advance. That is action R3, and it should be taken now rather than when the small-sample problem becomes visible.

A limit on what can be claimed here, stated honestly. Appendix C is truncated in the source PDF at "1 of 4", so further small-sample rules exist that IoL has not seen. IoL's cohorts are small enough that those rules matter. Do not build a strategy that depends on a reading of Appendix C that IoL cannot see in full. Obtain the complete appendix before relying on any threshold figure.

Category 4, the activity exists but is invisible. This is the more likely failure at IoL, and it is mundane. Course evaluations happen; they are not consistently recorded. A null Course evaluations score makes the KPI unreportable regardless of how satisfied the students were. The register names this directly: nulls are the failure mode that produces an unreported KPI, and partial coverage produces a programme-level number the Ministry can reject. The remedy is not a tactic. It is R1: make the evaluation a condition of course closure so that the record exists as a by-product of finishing the course.

A note on the substitute-data route, and why it is weaker here than elsewhere. The guide explicitly invites institutions to submit their own data directly to MoHESR for KPIs 1.1, 1.2, 2.4, 2.6, 3.1 and 4.1 where they consider it more accurate than the standard survey, and 2.6 is on that list. But the programme-level 2.6 instrument already is MBRU's own data, so the invitation adds little at programme level; its value is at institution level, where the SES sample for a small postgraduate department may be tiny and IoL's own near-census of its learners may genuinely be more accurate. The test the guide sets is more accurate, not more favourable, and the price of using it is being able to produce the raw per-respondent data with dates on demand. That price is exactly what R2 pays.

A defect in the submission template, flagged because it can lose this KPI outright. The HEDB Data Dictionary labels the KPI 2.6 result row as 2.6.N rather than 2.6.R, and gives the database field name as OBF2.6.N twice. The same duplication appears in the header of OBF Self Report.xlsx, where columns 33 and 34 are both 2.6.N. Confirm with CHEDS which physical field carries the 2.6 result before the first submission, or the student satisfaction result will be written into the numerator field and lost. This is not IoL's error to fix, but it is IoL's KPI to lose, and WF-26 should carry the check.

Boundary check. This workflow must never:

• decide or revise an Appendix C grouping after any results are visible;
• select which learners are surveyed, or exclude a course or a cohort from evaluation because its score is expected to be low;
• submit a Course evaluations score for a course whose evaluation was not actually run;
• discard raw per-question data, participant contact data, or the issue and collection dates, which Appendix B names explicitly;
• alter, exclude or reweight individual responses after collection closes;
• present a partial-coverage average as a full-coverage programme score without declaring the coverage in the Mechanism of collecting the data field;
• offer an incentive conditional on the content of a response rather than on responding;
• submit substitute data at institution level because it is more favourable rather than because the methodology is genuinely more accurate.

10. Service standards

11. Records and evidence

Appendix B readiness. This is the workflow where Appendix B exposure is most explicit and most specific. The Ministry may demand raw data and metadata from course-evaluation surveys, including participant contact data, per-participant per-question raw data, the survey issue date and the response-collection date. Retention is therefore mandatory, not optional, and it is a configuration decision in the survey platform rather than a policy aspiration. On the stated assumption, IoL could produce scores within 15 working days and could not produce raw per-question data with dates for past cycles, because it was not retained. Nothing can be done about past cycles. R2 stops the next one joining them.

12. Risks and controls

13. Performance measures

The first three measures decide this KPI and the last outcome measure decides whether it stays decided. A programme can score 4.7 and lose the KPI to null cells, and it can score 4.7 this year and 3.9 in three years if nothing visibly changes in response to what learners say.

14. Change control