The Gibbs reflective cycle is a six-stage framework for structured reflection: description, feelings, evaluation, analysis, conclusion, and action plan. Published by Graham Gibbs in 1988, it is the most widely used model in nursing and allied health education because it gives written reflection a defensible shape.
The marking criteria almost always weight analysis most heavily, and most submissions weight description most heavily. That single mismatch explains the majority of disappointing grades.
A reflective piece that spends 600 words recounting what happened and 100 words analysing it has produced a story, not a reflection. The event is only the raw material. What is being assessed is whether you can explain why it happened, what it reveals about your practice, and what you will do differently on evidence rather than instinct.
A useful discipline is to draft the analysis first and write the description afterwards, cut down to only the detail the analysis actually needs.
The six stages, and what each is really asking
Description asks what happened, factually and without interpretation. Keep it brief and anonymised. No patient identifiers, no colleague names, no institution.
Feelings asks what you were thinking and feeling at the time and afterwards. Honesty scores better than professionalism here. Writing that you felt confident throughout, when the event clearly unsettled you, closes off the analysis you could have drawn from that discomfort.
Evaluation asks what was good and bad about the experience. Both are required. A piece that finds nothing that went well, or nothing that went badly, has not evaluated.
Analysis asks what sense you can make of the situation, and this is where the literature belongs. Bring in clinical guidelines, professional standards, or research to explain the event. A reflection with no sources in the analysis stage rarely scores well at degree level.
Conclusion asks what else you could have done. It should follow from the analysis rather than introducing new material, and it should name specific alternatives.
Action plan asks what you would do if it arose again. This is the second most common place to lose marks.
Weak action plans say the writer will "read more around the topic" or "communicate better in future". Neither is testable, so neither can be assessed.
A strong action plan names a specific action, a mechanism, and a timeframe. For example: to complete the trust's deteriorating patient recognition module before the next placement block, and to use a structured handover tool for every escalation call during that block, reviewing with a mentor at the midpoint. That can be checked. The vague version cannot.
Consider a student who did not escalate a deteriorating patient quickly enough.
Description: during a late shift, observations showed a rising early warning score over two hours; the student rechecked rather than escalating immediately, and the patient was later reviewed urgently.
Feelings: uncertainty about whether the reading warranted a call, and reluctance to trouble a visibly busy senior colleague.
Evaluation: the observations were taken accurately and on time, which was good; the delay in escalation was not.
Analysis: this is where the work happens. The delay was not a knowledge deficit, since the student knew the escalation threshold. It was a hierarchy problem, and the literature on escalation failure consistently identifies perceived seniority barriers rather than knowledge gaps as the driver. Structured communication tools exist precisely to lower that barrier by giving juniors a script.
Conclusion: the student could have escalated at the first threshold breach and used a structured tool to frame the call.
Action plan: use the structured tool for every escalation in the next block and review at the midpoint with a mentor.
Notice that the analysis stage carries the insight, and it required a source. Without the literature, the analysis would have concluded only that the student should have been braver.
Gibbs, Kolb, and Driscoll
Kolb is a learning theory with four stages, describing how experience becomes knowledge. Gibbs turned that into a writing framework with explicit prompts. Driscoll reduces reflection to three questions: what, so what, and now what. Driscoll is quicker and suits short reflective entries; Gibbs suits assessed pieces that need visible structure.
They are not interchangeable, and rubrics are usually written against one named model. Using Gibbs when the rubric specifies Driscoll costs marks regardless of content quality.
Reflective writing describes real clinical events, and professional codes require confidentiality. Remove names, dates, locations, and any detail combination that could identify a patient or colleague. A rare condition plus a named unit plus an approximate date can identify someone even with no name attached.
Reflection often forms one component of a larger assessment. If yours sits within a practice-change project, see evidence-based practice and PICOT projects. If you are writing up a single patient's course in depth, that is a healthcare case study rather than a reflection. For clinical documentation, see SOAP note documentation, and for care planning, nursing care plan writing.
For programme-wide coursework support, see nursing writing.