Student library · Clinical placement
Clinical Placement
Across a validated placement assessment, the item educators agreed on least was evidence-based practice, at 56%. The item they agreed on most was risk management, at 83%. Knowing how you are actually assessed is worth more than any advice about attitude.
In one line. Placement is assessed with a structured instrument by a clinical educator who is also supervising you. The best-studied of these tools is the Assessment of Physiotherapy Practice, and its measurement properties tell you a great deal about what your grade does and does not mean.
This page is about the assessment and the supervision model, because those are the parts with evidence. It is not a list of tips.
The structured total is more reliable than the gestalt. Inter-rater ICC was 0.92 (95% CI 0.84 to 0.96) for the total score and only 0.72 (0.50 to 0.86) for the Global Rating Scale — the single overall impression. [1] If your educator's overall impression and their itemised marks diverge, the itemised marks are the better measurement. That is also the conversation worth having at mid-placement.
What the evidence shows
| Question | Finding | Source and quality |
|---|---|---|
| Do two educators grade you the same? | Broadly. Total score ICC(2,1) 0.92 (95% CI 0.84 to 0.96); Global Rating Scale 0.72 (0.50 to 0.86). Bland-Altman analysis identified no systematic differences between raters | The authors conclude educators showed a high level of reliability using the instrument [1] |
| Which items do they disagree about? | Agreement per item ranged from 56% for evidence-based practice to 83% for risk management, averaging 70% (SD 7) | Same study [1]. Worth knowing which of your marks rests on the shakiest agreement |
| What difference in score is real? | Standard error of measurement 3.2 points on a 0 to 80 scale; MDC90 of 7.86, about 9% of the scale width | Same study. A four or five point difference between two placements is inside the measurement error [1] |
| Which competencies are hardest? | The item hierarchy was identical across two samples: professional behaviour and communication were easiest to achieve; items related to clinical reasoning were the most difficult | Rasch analysis. Item difficulty was well targeted to student ability [2] |
| Is the assessment biased? | No differential item functioning was identified — the scale performed comparably regardless of the student's age, gender or amount of prior clinical experience, or the educator's age, gender or experience | Same analysis. One item, written communication, misfit the model in both samples but was retained as an important element of competence [2] |
| Does being supervised with a peer harm you? | No. A randomised trial found no significant between-group differences in assessment scores rated by a blinded assessor (p = 0.43), the supervising educator (p = 0.94) or the students themselves (p = 0.99) | Peer-assisted learning versus a traditional model. Educators gained about 6 minutes per day and students received an additional 0.33 written feedback entries per day [3] |
| So why do people dislike it? | Both educator and student satisfaction were higher with the traditional model, despite equivalent performance | Same trial. Preference and outcome diverged, which is a finding worth remembering whenever anyone defends a teaching model by how it feels [3] |
| Does pairing students help anything? | Students reported increased self-directed learning, informal teaching between themselves, and improvements in communication, consultation and task assignment | Small qualitative study, three student pairs, self-reported outcomes [5] |
| Does the instrument travel? | A Chinese adaptation was considered pragmatic for the clinical environment by over 90% of clinical educators, and all students agreed with the rating they received | Single clinical education unit and university; acceptability rather than psychometrics [4] |
Where students get this wrong
1. Optimising for the easy items
Professional behaviour and communication are the easiest items to achieve; clinical reasoning is the hardest. [2] Being punctual, pleasant and well presented gets you the marks almost everyone gets. If you want to move your grade, the evidence says to work on reasoning — being able to explain why you chose this assessment, what you expected, and what you did when the finding surprised you.
2. Comparing scores between placements
MDC90 was 7.86 on an 80-point scale, roughly 9% of its width. [1] A student who scores 52 on one placement and 57 on the next has not demonstrably improved. Educators know this intuitively; students often do not, and read noise as a verdict.
3. Treating the overall impression as the assessment
The Global Rating Scale — the single holistic judgement — had inter-rater ICC 0.72 against 0.92 for the itemised total. [1] The structured marks are the more reliable measurement. If you want useful feedback, ask about specific items rather than "how am I doing", which invites the less reliable answer.
4. Assuming your educator's background is working against you
No differential item functioning was found by student age, gender or prior clinical experience, or by educator age, gender or experience as an educator. [2] The instrument behaved the same way across all of those. That is a genuinely reassuring finding and it is worth knowing before you attribute a mark to who assessed you.
5. Fearing a paired placement
A randomised trial found no difference in performance under peer-assisted learning on any of three independent ratings, including a blinded assessor. [3] Students in paired models also received slightly more written feedback. [3] You may prefer one-to-one — most students and educators in that trial did — but the preference is not evidence that your learning suffers.
6. Not noticing which mark is least reliable
Educators agreed on evidence-based practice only 56% of the time, the lowest of any item. [1] That is partly because it is genuinely hard to observe. If you want it marked accurately, make it visible: say out loud what evidence you are drawing on, and put it in your notes rather than leaving your educator to infer it.
What the evidence supports — and what it does not
Supported
- Structured itemised assessment over a single global impression. [1]
- Targeting clinical reasoning as the competency that discriminates. [2]
- Peer-assisted placements as non-inferior on performance. [3]
- Interpreting score differences against an MDC of about 8 points. [1]
- Making your evidence use explicit, since it is the least reliably observed item. [1]
Not supported
- Reading small between-placement differences as progress. [1]
- Treating the global rating as the definitive judgement. [1]
- Claims that paired supervision harms learning. [3]
- Attributing your mark to your educator's age, gender or experience. No DIF found. [2]
- Strong conclusions from the paired-placement qualitative work. Three pairs, self-reported. [5]
How certain is this?
Evidence grade: Moderate.
Unusually for a clinical education topic, the core evidence is strong: a formal inter-rater reliability study reporting ICCs, SEM and MDC, [1] a Rasch analysis in two independent samples that agreed on the item hierarchy and found no differential item functioning, [2] and a registered randomised trial with a blinded assessor for the supervision question. [3]
What limits the grade is generalisability. The instrument studied is Australian, and the psychometric work was done in that system; the international evidence here is an acceptability study in one Chinese unit. [4] Your programme may use a different tool whose properties are unpublished. The qualitative supervision study is very small. [5]
A note on the dates. Four of these papers predate 2021. They are the primary validation studies for the instrument and the primary trial of the supervision model, and neither dates the way treatment evidence does.
Common questions
What should I actually work on?
Clinical reasoning. Across two independent samples the item hierarchy was identical, with reasoning items the hardest to achieve and professional behaviour and communication the easiest. [2] Practise articulating your reasoning out loud: what you expected to find, what you actually found, and what you changed as a result.
How do I get better feedback?
Ask item by item rather than in general. The global impression is the less reliable judgement (ICC 0.72 versus 0.92 for the itemised total), [1] so "how am I doing" invites the answer with the most noise in it. Ask instead which specific items your educator would currently mark you down on and what would move them.
My scores dropped between placements. Should I worry?
Not on a small difference. The MDC90 was 7.86 points on an 80-point scale. [1] Different settings, different caseloads and different educators all contribute, and the instrument's own measurement error covers a change of several points. Look at which items moved, not at the total.
Is being placed with another student a bad sign?
No. In a randomised trial, performance under peer-assisted learning did not differ from traditional supervision on a blinded assessor's rating, the educator's rating or the students' own (p = 0.43, 0.94, 0.99). [3] Students in that model received slightly more written feedback. Most participants preferred the traditional model, [3] but preference and performance are different things.
How does this connect to the rest of the library?
Evidence-based practice is the placement item educators agree on least, [1] and it is the one you can most directly prepare for. The evidence skills resources here exist for that: being able to say what a trial showed and how certain it is turns an unobservable competency into something your educator can actually mark.
References
- Dalton M, Davidson M, Keating JL. The assessment of physiotherapy practice (APP) is a reliable measure of professional competence of physiotherapy students: a reliability study. Journal of Physiotherapy. 2012;58(1):49–56. doi:10.1016/S1836-9553(12)70072-3 PMID 22341382 Inter-rater reliability study
- Dalton M, Davidson M, Keating J. The Assessment of Physiotherapy Practice (APP) is a valid measure of professional competence of physiotherapy students: a cross-sectional study with Rasch analysis. Journal of Physiotherapy. 2011;57(4):239–46. doi:10.1016/S1836-9553(11)70054-6 PMID 22093122 Rasch analysis
- Sevenhuysen S, Skinner EH, Farlie MK, et al. Educators and students prefer traditional clinical education to a peer-assisted learning model, despite similar student performance outcomes: a randomised trial. Journal of Physiotherapy. 2014 Dec;60(4):209–16. doi:10.1016/j.jphys.2014.09.004 PMID 25450483 Randomised trial
- Hu J, Jones AYM, Zhou X, et al. Acceptance of the 'Assessment of Physiotherapy Practice (Chinese)' as a standardised evaluation of professional competency in Chinese physiotherapy students: an observational study. BMC Medical Education. 2020 Apr 9;20(1):108. doi:10.1186/s12909-020-02026-3 PMID 32272913 Feasibility and acceptability study
- Jelley W, Larocque N, Patterson S. Intradisciplinary clinical education for physiotherapists and physiotherapist assistants: a pilot study. Physiotherapy Canada. 2010 Winter;62(1):75–80. doi:10.3138/physio.62.1.75 PMID 21197181 Qualitative study
About this resource
How to use this
Written to be learned from, not memorised.
This page describes how placement assessment actually behaves as a measurement, because that is more useful to a student than advice about attitude. Faculty may use this page in teaching with attribution. It carries its review date and its next review date, so you can see at a glance whether it is current before you put it in front of a cohort.