Ask a clinical supervisor who has been doing workplace-based assessment for five years to compare the narrative comments they wrote in their first year with the ones they write now. Most will not want to. The first-year comments are longer, more specific and more useful. The current ones are shorter, safer and faster.
The usual explanation is time. Clinical schedules have not got lighter while the assessment load has gone up, and fitting the required observations into a working clinic is a recognised problem in the literature rather than a complaint. That is real and it matters.
But it is not the whole explanation, and pretending it is lets the actual problem go unexamined. The deeper reason supervisors write less over time is that they wrote something careful, and then never saw it used.
You wrote two hundred words about a student's approach to a difficult conversation. You do not know whether anyone read them. You suspect not. You are probably right.
The disappearance problem
Here is the shape of it. You observe a student. You notice something worth saying that is not on the form's rating scale: a tendency to talk over anxious patients, or an unusually good instinct for when to stop and check understanding. You write it down properly, because it is the kind of thing that changes a career if someone hears it early.
Then the form is submitted, and from your point of view it goes into a filing cabinet.
You do not see whether the student read it. You do not see whether the next supervisor, six weeks later, knew to look for the same thing. You do not see whether it reached the committee. When you meet the student again, you have no record of what you said the first time, so you start over. And if you are honest, the fourth time you write a careful paragraph into a system that appears to swallow it, you write a shorter one.
That is not laziness. It is an accurate response to a feedback loop that has been cut. You are behaving exactly as a rational person behaves when their effort produces no observable effect.
Three things that have to be true for narrative feedback to survive
1. It has to be a first-class field, not a comment box
In a great many assessment tools, the numeric rating is the data and the narrative is an optional afterthought sitting underneath it in a smaller font. Everything downstream then treats it that way: reports aggregate the numbers, dashboards chart the numbers, the committee reads the numbers.
This is the wrong way round for workplace-based assessment. Gingerich, Regehr and Eva have argued persuasively that rater variation is not simply error to be averaged away; different assessors notice genuinely different things, and the narrative is where that difference lives. Strip it out and you have averaged away the signal along with the noise.
Narrative has to be captured alongside every score as a structured part of the record, and it has to be retrievable later as text, not as an attachment nobody opens.
2. The next supervisor has to be able to see it
The single change that most improves the quality of supervision across a programme is letting the next person see what the previous person said. Not the numbers. The words.
It turns a series of disconnected encounters into something continuous. You arrive at a student already knowing that two colleagues have flagged the same thing, which means you can spend your twenty minutes on that rather than rediscovering it. And it means your own comment lands on someone who can act on it, rather than into a void.
3. The student has to be able to see it, and see it more than once
Feedback delivered verbally at the end of a shift is heard by a person who is tired and thinking about something else. The same feedback, readable again three weeks later when they are calm and preparing for the next placement, is worth several times more.
This is not a small effect. It is arguably the largest single return available from putting assessment into a longitudinal record at all.
What a longitudinal record actually gives a mentor
Practically, and without overstating it:
- The set, not the sample. Every mini-CEX, DOPS, case discussion, team feedback round and reflection about your student, in sequence, with the words intact. Your twenty-minute observation stops being the only thing you know about them.
- Someone else's observation as a starting point. "Two supervisors have mentioned your consultation timing. Let's watch for that today." This is a different conversation from "let's see how you get on".
- Your own history with a student. What you said in November, in front of you in March, so the second conversation builds on the first rather than repeating it.
- Visible destination. Knowing that what you write is read by the student, by the next supervisor and by the committee is the thing that makes it worth writing properly. It is also, over a year, the thing that keeps the quality up.
The honest limits
This does not give you more time. A better record does not shorten a clinic. If your programme's assessment volume is already unsustainable for its supervisors, a platform that makes the assessments easier to complete will not fix an underlying design problem, and it may make it easier for the programme to ask for more of them. Assessment fatigue is a documented phenomenon with an uncomfortably recognisable shape: resistance to completing evaluations, and then rushed and poorly completed ones. Software is not the remedy for that. Programme design is.
Seeing prior feedback introduces a bias risk. If you read that two colleagues thought a student was weak on time management, you are more likely to see time management problems. This is a genuine trade-off, well described in the rater cognition literature, and anyone who sells you continuity without mentioning it is not being straight with you. Our view is that the benefit outweighs the risk in most programmes, because the cost of every supervisor starting from zero is very high and falls hardest on the students who most need consistent attention. But it is a judgement, not a fact, and a programme should make it deliberately rather than by default.
A platform cannot make a colleague write well. It can make it more likely, by showing them their words are used, and it can make it easier, with prompts that ask for something specific rather than a blank box. It cannot substitute for faculty development, and any vendor implying otherwise is selling you the easy half of the problem.
The narrow version of the claim
We think the reason good supervisors stop writing good feedback is not that they stopped caring. It is that they stopped seeing it matter. That is a fixable problem, and fixing it is mostly a matter of making the record continuous and readable by the three people who need it: the student, the next supervisor, and the committee.
Everything else in a platform is secondary to that.
References
- Norcini, J. and Burch, V. (2007) "Workplace-based assessment as an educational tool: AMEE Guide No. 31", Medical Teacher 29(9), 855–871.
- Gingerich, A., Regehr, G. and Eva, K.W. (2011) "Rater-based assessments as social judgments: rethinking the etiology of rater errors", Academic Medicine 86(10 Suppl), S1–S7.
- Schuwirth, L.W.T. and Van der Vleuten, C.P.M. (2011) "Programmatic assessment: from assessment of learning to assessment for learning", Medical Teacher 33(6), 478–485.
- Ott, M. et al. (2022) on learner experience and assessment burden in competency-based medical education, Journal of Graduate Medical Education.
- Schut, S., Maggio, L.A., Heeneman, S. et al. (2021) "Where the rubber meets the road — an integrative review of programmatic assessment in health care professions education", Perspectives on Medical Education 10(1), 6–13.