Gibbs vs Rolfe: Which Reflective Model to Use

The NMC’s reflective account requirement doesn’t mandate any particular reflective model. You can use Gibbs, Rolfe, Driscoll, Schön, Borton’s developmental framework, or no formal model at all, as long as the four Form 6 fields are answered.

That said, most UK nurses use either Gibbs or Rolfe. The choice between them is mostly about depth and pace.

The Gibbs Reflective Cycle (1988)

Six stages, designed for complex events with emotional or interpersonal weight:

1. Description: what happened.
2. Feelings: what you thought and felt at the time.
3. Evaluation: what was good and bad about the experience.
4. Analysis: what sense you can make of the experience.
5. Conclusion: what else you could have done.
6. Action plan: what you would do differently next time.

Gibbs takes 30–60 minutes to work through properly. The model deliberately surfaces emotional content as part of the analysis, which is its strength for events that had emotional impact and its weakness for events that were primarily cognitive (a calculation error, a guideline change).

The Rolfe What/So What/Now What model (2001)

Three stages, designed for faster reflective practice:

1. What? What happened.
2. So what? What does it mean, why does it matter.
3. Now what? What changes as a result.

Rolfe takes 10–20 minutes per reflection. Cleaner, faster, less emotional excavation. Better for high-volume reflective practice and for events where the learning is primarily clinical or operational.

Mapping to Form 6

Both models produce content that fits the four Form 6 fields, just with different intermediate structure:

Gibbs to Form 6:

• Description, Feelings map to Form 6 Field 1 (what happened).
• Evaluation, Analysis map to Form 6 Field 2 (what you learned).
• Conclusion, Action plan map to Form 6 Field 3 (how you changed practice).
• Code link is added in Field 4.

Rolfe to Form 6:

• What? maps to Form 6 Field 1.
• So what? maps to Form 6 Field 2.
• Now what? maps to Form 6 Field 3.
• Code link is added in Field 4.

The Rolfe mapping is more direct. Gibbs gives you more raw material to draw from but requires editing down to fit the Form 6 word counts.

When Gibbs fits better

Events with significant emotional, ethical or interpersonal weight:

• A patient death that affected you.
• A complaint that prompted self-doubt.
• An ethical dilemma you worked through.
• A team conflict you resolved.
• A whistleblowing decision.
• An end-of-life care experience.

Gibbs’ “feelings” stage matters here because the emotional content is part of the learning. Skipping it would produce a thinner reflection.

When Rolfe fits better

Events where the learning is primarily clinical, procedural or operational:

• A near-miss caught by checking.
• A new piece of evidence that changed your practice.
• A study day that produced specific behavioural change.
• A piece of feedback on technique or process.
• A guideline update you adapted to.
• A documentation issue you addressed.

Rolfe’s economy works because the emotional layer isn’t load-bearing for the learning.

A worked comparison

The same event, processed through both models:

Event: Near-miss on a drug calculation, caught by rechecking the BNF before administration.

Gibbs version: Description (what happened), then Feelings (how the realisation felt, the moment of recognising the dose was wrong), then Evaluation (what was good: the recheck habit worked; what was bad: the original calculation was wrong), then Analysis (why I made the calculation error, what conditions led to it), then Conclusion (what I would do differently), then Action plan (specific changes to take forward).

Rolfe version: What (the calculation error and recheck catch), then So what (the parameter I was working from was outdated, my assumption that the prescription was current was wrong), then Now what (the change to per-administration parameter checking).

Both end with the same Form 6 content. The Gibbs path took 40 minutes; the Rolfe path 15.

Practical recommendation

For most working nurses doing five reflections across a three-year cycle, the practical pattern is:

• Use Rolfe by default for routine clinical learning events. Faster, gets the Form 6 content out, fits a busy professional’s time.
• Use Gibbs for one or two events in the cycle that had significant emotional or interpersonal weight. The deeper structure earns its time on these.

The five accounts together can mix models. The audit doesn’t see which model you used; it sees the Form 6 output.

This is the end of Part 4 of the guide. The next part (chapters 62 to 69) covers CPD strategy: what counts, what doesn’t, free sources, and the audit reality.