Reflective Models Compared: Kolb, Gibbs, Driscoll and Schön

Sooner or later, every CIPD learner writing reflectively meets the same question: which model should I use? Briefs sometimes name one, tutors mention Kolb, Gibbs, Driscoll and Schön almost interchangeably, and the textbook summaries all sound equally plausible. This guide is a working reflective models comparison: what each of the four best-known models actually is, when it suits, where it tends to go wrong, and a simple method for choosing the right one for your situation, your level and the way you think.

Two ground rules before the detail. First, a model is scaffolding for thinking, not a subject to write about, and no model can rescue an account with nothing genuine in it; the pillar guide What Is a CIPD Reflective Account and How to Write One covers those fundamentals, and this guide builds on them. Second, always check your current assessment brief and your study centre's guidance before you choose, because some units expect a named model, some leave the choice open, and requirements vary by centre and over time.

Driscoll: What, So what, Now what

Driscoll's model asks three questions in sequence. What happened: a brief, factual account of the situation and your part in it. So what: what the experience meant, why you acted as you did, and what it revealed about how you work. Now what: what you will do differently as a result, stated concretely enough for a reader to check later.

It suits almost everything, which is precisely its value. First reflective accounts, tight word counts, busy practitioners, and any unit where the brief does not name a model: Driscoll fits them all, because the three questions map directly onto what assessors mark. It is the hardest model to get wrong, since there are no stages to trudge through and nowhere for the structure to become the subject.

Its risk is thinness rather than error. Because the model gives so little prompting, a hurried writer can answer so what in a single sentence and believe the job done, and there is no explicit push towards feelings, theory or other perspectives. Driscoll rewards writers who interrogate themselves without being told to; if you need more scaffolding to reach depth, a fuller model may serve you better.

Here is a clearly fictional micro-example. Dana, an invented HR assistant at a made-up logistics firm, used Driscoll on a payroll error she had failed to query, giving three sentences to what happened and most of her words to why she had trusted the spreadsheet over her own doubts. The account worked because each question got an honest answer and the middle one got the bulk of the space.

Gibbs: the reflective cycle

Gibbs' cycle works through six stages: description, feelings, evaluation, analysis, conclusion and action plan. Its defining features are the explicit feelings stage, which makes your emotional reaction a legitimate object of study rather than something to edit out, and the evaluation stage, which forces a balanced judgement about what worked and what did not before deeper analysis begins.

It suits emotionally charged situations best: conflict, criticism, feedback that stung, a decision that kept you awake. When your reaction was part of the story, Gibbs stops you skipping past it, and the staged structure carries writers who would otherwise freeze at a blank page.

The risk is mechanical stage-trudging. Six headings with two dutiful sentences under each produces writing that visits every stage and reflects at none, and the feelings stage can slide into venting if the emotion is reported but never analysed. Gibbs also gives description a stage of its own, which tempts the narrative to sprawl. If you use it, spend your words on evaluation and analysis and keep the other stages lean.

Another invented example makes the fit clear. Marcus, a fictional learning coordinator at an invented insurance company, chose Gibbs for a training session in which a delegate openly challenged his material, because his flash of defensiveness was the real story. The feelings stage let him name the embarrassment honestly, and the analysis stage examined what that defensiveness had done to his handling of the room.

Kolb: the experiential learning cycle

Kolb frames learning as a repeating loop: concrete experience, reflective observation, abstract conceptualisation and active experimentation. You do something, you notice what happened, you draw a working conclusion, and you test that conclusion in later practice, which generates the next experience and starts the loop again.

It suits repeated practice: recurring meetings, a skill built over months, any situation where you genuinely tested a change and saw the result. Because the experimentation stage demands evidence that thinking altered behaviour, Kolb builds proof of change into the structure itself, which is exactly what stronger briefs ask for.

The risk is forcing a loop that never happened. If you have a single event and no later opportunity to test what you concluded, Kolb turns speculative, and the conceptualisation stage tempts writers into abstract generalisation about how people learn instead of concrete conclusions about themselves. Choose it when you can show the full cycle honestly, not because it sounds the most theoretical.

A fictional example shows the loop working. Ines, an invented people partner at a made-up retail chain, reflected on her monthly absence reviews across a quarter rather than on one meeting. Each cycle she drew one working conclusion, changed one thing the following month and recorded what actually happened, which gave her account built-in evidence of change.

Schön: reflection-in-action and reflection-on-action

Schön's contribution is a distinction rather than a cycle: reflection-in-action is the thinking you do while acting, and reflection-on-action is the examination you give the same events afterwards. Writing with Schön means reconstructing how you read a situation as it unfolded, then re-examining that live thinking with the benefit of distance.

It suits experienced practitioners, because so much of senior practice is judgement exercised in the moment: adjusting an argument mid-meeting, reading a room, changing tack when a conversation turns. An account that captures both the in-flight adjustment and the honest afterthought produces naturally analytical material, which is why the distinction features prominently in 7CO03 Reflective Writing at Masters Level.

The risks run the other way from Driscoll's. There is no staged scaffold, so less experienced writers can drift, and reconstructing in-the-moment thinking honestly is genuinely hard, because hindsight quietly rewrites what you noticed at the time. Used carelessly, Schön becomes a licence to claim you were thinking things you were not.

One last invented example. Tomas, a fictional employee relations manager at a made-up utilities firm, wrote about noticing mid-hearing that his questions were leading a witness, and correcting course there and then. His account examined the live adjustment first and then, at more length, why his preparation had pointed him towards a conclusion before the evidence arrived.

How to choose: situation, level and thinking style

Start with the situation. An emotionally charged single incident points to Gibbs. A recurring practice where you tested a change points to Kolb. Live professional judgement, especially at senior level, points to Schön. Everything else, and any situation where you are unsure, points to Driscoll, which is never a wrong answer.

Level matters less than learners fear, but it shapes the sensible default. Early in the qualification, Driscoll's simplicity protects you from the commonest failure, which is writing about the model instead of yourself. At Level 5, any of the four works if it is used analytically. At Level 7, Schön and Kolb tend to fit the material senior practitioners actually have, though a Driscoll account with genuinely critical answers will outscore a mechanical tour of a fuller cycle every time.

Finally, match the model to the way you actually think. If you process events through how they felt, Gibbs' feelings stage will draw out material you would otherwise suppress. If you think in experiments and iterations, Kolb will feel natural. Whatever you choose, one model used deeply beats three used cosmetically, and switching between models mid-account signals confusion rather than range.

Using any model invisibly

However you choose, the model should function as scaffolding, not subject matter. Assessors award marks for the quality of the thinking a model produces, not for your ability to summarise the model, and a paragraph explaining a cycle is a paragraph of reflection you did not write. Name your model once if the brief expects it, then let the structure do its work silently.

The practical technique is to draft with the stages as private headings and then delete them, smoothing the joins so the account reads as continuous thought. The test of success is that a marker could reconstruct your model from the shape of the writing alone: a little description, sustained analysis, honest learning, concrete action. If the shape is visible, the scaffolding has held, whether or not the model is ever mentioned.

A five-step decision method

1. Check your current brief and your centre's guidance; if they name a model, use that one and go straight to step five.
2. Identify what kind of material you have: one emotionally charged incident, a repeated practice with a tested change, live in-the-moment judgement, or none of these.
3. Map the material to the model: Gibbs for the charged incident, Kolb for the tested loop, Schön for live judgement, Driscoll for everything else.
4. Sense-check the choice against your own thinking style, and if the model feels like a costume rather than a scaffold, drop down to Driscoll.
5. Draft with the stages as private headings, delete the headings, and check the finished account mentions the model no more than once.

Common pitfalls

• Describing the model instead of using it, so the account reads as a book report with a story attached
• Forcing every stage, with dutiful sentences under headings the experience does not need
• Mixing models mid-account, which fragments the thinking and suggests none of them was understood
• Choosing a model to look sophisticated rather than because it fits the material
• Letting the feelings stage become venting, with emotion reported but never analysed
• Claiming a learning loop with no tested change, which assessors read as theory without practice

Go deeper

This guide is a companion to the pillar What Is a CIPD Reflective Account and How to Write One, which covers the fundamentals every model exists to serve: honest first person, analysis over narrative, specific learning and checkable action. Its sister guide, Reflective Writing at CIPD Level 3, 5 and 7: What Changes, shows how the same skill deepens across the qualification. For unit-specific depth, see 5CO03 Reflective Writing: How to Evidence Your Examples and 7CO03 Reflective Writing at Masters Level.

Choosing a model takes five minutes; using one well takes practice, and an outside reader helps most with the gap between the two. Our reflective writing support includes help selecting and applying a model, coaching on structure, and honest review of your drafts, while the experience, the thinking and the words stay entirely yours.