Openness to feedback reads like a personality trait and behaves like a skill. Whether someone takes a difficult comment well depends on who runs the conversation, what order the questions come in, and whether reaction gets its own place. Research on feedback models in professional education says that order can be designed.
Quick Overview
What you’ll learn:
- Why openness to feedback behaves like a learned behaviour rather than a fixed trait
- Which named feedback models the literature describes, and how they differ from each other
- What a dedicated reaction phase changes before a conversation moves to content
- Who owns the step that follows a development conversation
Who this article is for:
- Mentors and line managers who run development conversations
- Specialists who want to leave a review with a plan rather than a grievance
- Anyone designing mentoring tracks or development programmes
Reading time: 7 minutes
A Behaviour in a Conversation, Not a Fixed Trait
Appraisal language treats openness to feedback as a personal quality — something a person has or lacks, sitting in a competency grid next to “communication” and “teamwork”. The framing survives because it is cheap to apply and almost never tested. Test it once and it comes apart: the engineer who absorbs a hard critique at a whiteboard shuts down on identical content delivered at a quarterly review.
Nothing about the person changed between those occasions. What changed was whether the purpose of the conversation had been stated, whether the verdict arrived before or after they were asked for their own reading of the work, and whether the ending was a label or an agreed next step. Each of those is a property of the conversation, and whoever runs it owns all of them.
The consequence lands on whoever writes the competency note. “Closed to feedback” is a claim about a conversation that already happened, and it names the wrong participant. Before the phrase goes into a review it is worth asking which of those conditions was in place when the judgement formed — because where none of them was, the note records the quality of the conversation rather than the quality of the person.
Named Models Are What Actually Gets Practised
A development conversation does not have to be improvised. A review published in the Journal of Educational Evaluation for Health Professions and indexed on pubmed.ncbi.nlm.nih.gov sets out the six most common feedback models: the Feedback Sandwich, the Pendleton Rules, the One-Minute Preceptor, the SET-GO model, the R2C2 model and ALOBA (Agenda Led Outcome-based Analysis). For each one the authors describe the structure, strengths and weaknesses, what it demands of the educator and the learner, and the encounters it suits.
The fact that such a comparison is possible at all is the point here. The models differ in who speaks first, how much room self-assessment gets, and whether the conversation ends with a plan. The Sandwich buries the message between two compliments and is often read as ritual. The Pendleton Rules hand the floor to the recipient before the educator comments. ALOBA opens with an agenda raised by the learner, so the conversation starts on whatever is actually bothering them.
Choosing a model is a decision, not a matter of taste. The review is explicit that these are practical frameworks to adopt and adapt, combining or modifying them to suit the context rather than reciting them. The recipient’s side of the skill shifts with that choice: a conversation whose shape is recognisable is easier to stay open in.
The Reaction Phase Is the Part Usually Missing
The R2C2 model described in Academic Medicine and indexed on pubmed.ncbi.nlm.nih.gov runs in four phases: build relationship, explore reactions, explore content, and coach for performance change. Its authors developed it iteratively, drawing on earlier research into what influences giving, receiving, accepting and using feedback.
The second phase is the interesting one, because a standard appraisal does not have it. Reaction — surprise, disagreement, relief — arrives regardless; without a place of its own it leaks into the discussion of content and turns it into an argument. Setting aside an explicit moment to ask what the recipient feels on hearing the result is not a courtesy. It is what keeps the rest of the conversation usable.
The follow-up study in residency education, also indexed on pubmed.ncbi.nlm.nih.gov, reports what happened when supervisors used the model in their regular mid-rotation and end-of-rotation sessions. Supervisors found the structured format useful; both residents and supervisors reported that the coaching phase was novel and helpful and that the model produced collaborative, reflective, goal-oriented discussions. The authors keep their claim narrow — participants found the conversations meaningful and identified goals — which is a different statement from a promised improvement in results.
Who Owns What Happens Next
Everything above concerns a single conversation. What makes a conversation matter beyond its own hour is whether anybody owns the step that follows it, and that is a question about roles rather than about skill.
The line manager assesses the work and owns the agreed ending. The role is formal and carries employment consequences, so this is the person who marks what is a requirement and what is a suggestion.
The mentor works from experience and does not assess formally. That is precisely the advantage: a mentee will raise with a mentor what they would never raise with the person writing their review.
The development sponsor is whoever confirms inside the organisation that the agreed step has time and budget. Without that confirmation, a plan from a development conversation competes with delivery work and loses.
Splitting these functions removes the conflict in which one person assesses, develops and funds at once. For the recipient it does something simpler: it says who to go back to with what they heard.
Read Also
- Openness to feedback — a shorter take, useful before the conversation
- Constructive feedback: a practical framework for mentors — the same skill viewed from the giver’s side
- A coaching question toolkit for mentors: A Mid-Size Company Implementation Case Study — the questions that probe what was heard instead of stopping at accepting it
Build Your Skills
This topic is developed in the Coaching: Stimulating and Supporting Employee Development and Coaching in IT Projects courses. The first works on the craft of the development conversation; the second places it inside a technical team. Contact us to match the scope to your role.
Frequently Asked Questions (FAQ)
What makes feedback constructive?
Feedback is constructive when it describes a behaviour and its effect instead of assigning the recipient a trait, and when it closes on an agreed step. The review of models indexed on pubmed.ncbi.nlm.nih.gov shows that constructiveness is a property of the structure of the conversation — order, room for self-assessment, ending — rather than of polite phrasing.
How do you run a difficult conversation that has to carry feedback?
Open by saying why the conversation is happening and what will be done with its outcome, then make room for reaction before moving to the substance of the assessment. That is exactly the sequence the R2C2 model sets out — relationship, reaction, content, coaching — as described in Academic Medicine. Skipping the reaction phase does not remove it from the conversation; it relocates it into an argument about facts.
How do you receive feedback without going on the defensive?
Separate the moment of reacting from the moment of deciding. Say out loud what you are hearing and what it sets off, ask for a specific situation instead of a generalisation, and only then discuss what to do about it. A recipient can also ask the person running the conversation to start with their own reading of the situation.
Which feedback technique suits an annual review?
Models that open with the recipient’s self-assessment or with an agenda they raised — the Pendleton Rules and ALOBA — fit a conversation summarising a longer period. Short formats such as the One-Minute Preceptor were built for feedback given on the spot next to a task, and will be too narrow for an annual review.