Problems and operational improvement
Turn Recurring Errors into an Improvement Experiment
Organise the conditions in which errors occur in observable terms and design a small prevention experiment.
- People
- 4–16
- Session time
- 150 minutes
- Format
- In person
English session plan
Original program status: draft. Source safety category: Medium; this is not a guarantee of safety.
Preparation: Standard preparation. Facilitation team: Four roles, including someone with facilitation experience.
What this session can decide
Decide hypotheses and experiments for improving work. Do not determine incident responsibility, discipline, or legal causation.
Outside its scope
- Do not substitute for an incident investigation.
- Do not make individual attentiveness the sole cause.
- Do not guarantee experiment results in advance.
Primary caution: Do not substitute for an incident investigation.
Before starting
- Question and scope
- Organise the conditions in which errors occur in observable terms and design a small prevention experiment. Before starting, state the subject and time period in one sentence.
- What this session decides
- Decide hypotheses and experiments for improving work. Do not determine incident responsibility, discipline, or legal causation.
- Output recipient
- The Recorder stores the final goal/experiment/measurement table, managing changes, observation indicators, owners, and ending conditions.
- Participation and information boundaries
- Explain the choices to pass, take a break, or leave. Keep personal information, confidential information, and attribute-based evaluation off the record surface. Assign Participant support separately.
- Operating roles
- Assign Lead facilitator, Recorder, Timekeeper, and Participant support before starting, and confirm who announces switching and stopping.
- Next review
- Set a review date, viewers, and the person responsible for the next decision on the goal/experiment/measurement table.
Who takes responsibility
- Lead facilitator
Keep Turn Recurring Errors into an Improvement Experiment within its purpose and exclusions, and announce the start and end of each stage.
From before the start through the end
- Recorder
Record observation items and recording scope, facts/meaning/next questions, and the goal/experiment/measurement table at each stage, then hand them to the recipient at the end.
From preparing record surfaces before the start through storage after the end
- Timekeeper
Notify the Lead facilitator 3 minutes before each stage ends and when switching to breaks or buffer time.
At the start of every stage and 3 minutes before its end
- Participant support
Receive requests to pass, take a break, leave, or check individually, and decide with the Lead facilitator whether resumption is possible.
From before the start through the end
Session flow
Minutes are measured from the session start. Method time, transition time, breaks and buffer are shown separately; they are not additional overlapping activities.
0–33 minutes
Align the facts to observe
Method: Observation Contract · 0–30 minutes. Transition: 30–33 minutes (3 minutes).
Decide what is observed, when, by whom, and within what scope. Exclude names and assumptions, recording only actions and conditions that can actually be seen or heard.
Responsible: Lead facilitator; Recorder; Participant support
Output: Observation items and recording scope
Handoff: Pass observed facts to the stage for separating meaning and next questions from facts.
This stage may be shortened only in accordance with the stated conditions.
33–61 minutes
Separate meaning and next questions from facts
Method: What, So What, Now What · 33–58 minutes. Transition: 58–61 minutes (3 minutes).
First share only the facts of what happened. Next offer multiple interpretations of their meaning, and finally choose questions you will check yourselves.
Responsible: Lead facilitator; Recorder; Participant support
Output: Facts, meaning, and next questions
Handoff: Pass selected questions and hypotheses to experiment design.
This stage is not marked as shortenable.
66–129 minutes
Design a prevention experiment and measurement
Method: Goal Experiment Measure Loop · 66–126 minutes. Transition: 126–129 minutes (3 minutes).
State the desired change in one sentence and design one small change. Decide indicators to check before and after the experiment, an owner, period, and stop conditions.
Responsible: Lead facilitator; Recorder; Participant support
Output: Goal/experiment/measurement table
Handoff: Pass the experiment table to the work owner and the person responsible for measurement.
This stage is not marked as shortenable.
What is retained and who receives it
Observation items and recording scope
Owner: Recorder
Avoid evaluative language and collect facts that can be compared.
Facts, meaning, and next questions
Owner: Recorder
Develop improvement hypotheses without mixing observation and interpretation.
Goal/experiment/measurement table
Owner: Recorder
Track changes, observation indicators, owners, and ending conditions.
Materials and setup
Material names, quantities, groups and handling instructions retain the fixed draft's original helper-derived guidance. Where a format is marked “Source material specification”, the entire bound English source material entry takes precedence over the old helper-derived format guidance. This format-only correction does not resolve other instruction or safety concerns. English names are not substituted into a Japanese-keyed lookup.
Everyday tools
- Action sticky notes
Quantity: 3 per person
Format: Sticky notes about 75 × 75 mm, or A4 copy paper cut into 4 pieces
Do not buy special-purpose materials. Write “Action” at the top of each paper piece. If participants will write on it, use one item per piece and handwrite the required headings before starting.
- Writing implements
Quantity: 1 per person
Format: Dark broad-tip pens or whiteboard markers
Check that ink flows and place them within participants' reach. Avoid pale-coloured pens.
- Writing implements
Quantity: 1 per person
Format: Dark broad-tip pens or whiteboard markers
Check that ink flows and place them within participants' reach. Avoid pale-coloured pens.
- Broad-tip pens
Quantity: 1 per person
Format: Dark broad-tip pens or whiteboard markers
Check that ink flows and place them within participants' reach. Avoid pale-coloured pens.
Sheets to prepare by hand
- Table for agreeing actions and observation methods
Quantity: 1 per group plus 1 for whole-group consolidation
Format: Source material specification: One A3 observation-contract sheet per group, plus one for whole-group integration
Source-bound format display correction; the other fixed helper-derived guidance is unchanged.
Divide the work surface into 7 fields: “Action to try”, “Facts to observe”, “Observer”, “How results are returned to the person”, “Record format”, “Observation deadline” and “Where results will be reviewed”. Do not rely on colour alone; write each field heading in words.
- Example-record cards
Quantity: 1 per group
Format: Sticky notes about 75 × 75 mm, or A4 copy paper cut into 4 pieces
Before starting, write just one example close to the content used in this method. Show it as an example of how to write, not a correct answer for participants to copy.
- Three-field worksheet
Quantity: 1 card per person
Format: A4 copy paper
Create fields headed “What happened?”, “So what can we learn?” and “What will we do next?” on one sheet. Write each heading in words rather than relying on colour alone.
- Large paper or whiteboard for the whole-group record
Quantity: 1 surface
Format: Large paper or a whiteboard
Divide the recording surface into three fields: “What happened?”, “So what can we learn?” and “What will we do next?”. Write the headings in thick lettering and leave room to attach each group’s records.
- Homemade 3-column goal, experiment, and measurement sheet
Quantity: 1 per group
Format: Two A4 sheets placed side by side; join with masking tape if more space is needed
Create Goal, Experiment, and Measurement columns on one sheet. Write a heading for each column rather than relying on colour alone.
- Table for comparing experiment priorities
Quantity: 1 per proposed experiment
Format: A4 copy paper
Divide the work surface into 3 columns: Expected impact, Effort required, and Confidence in the result. Write a heading for each column rather than relying on colour alone.
- Measurement sheet for baseline, collection method, purpose, storage, and deletion
Quantity: 1 per proposed experiment
Format: A4 copy paper
Create Baseline, Collection method, Purpose of use, Storage, and Deletion fields on one sheet. Write a heading for each field rather than relying on colour alone.
- Learning sheet for stop conditions, responsible person, and short-cycle update date
Quantity: 1 per proposed experiment
Format: A4 copy paper
Create Stop conditions, Responsible person, and Short-cycle update date fields on one sheet. Write a heading for each field rather than relying on colour alone.
Information to gather before the session
- Information sheet for checking participation and handling of records
Quantity: 1 copy per affected person
Format: A4 copy paper or viewing data with the same content
Use language a first-time reader can understand to explain What participation involves, Consent is voluntary, How to withdraw consent, and Whether existing records will be retained or deleted after withdrawal. Add the creation date, the person who checked it, and a contact for questions. If distributing it, also prepare a large-print version or viewing file with the same content.
Setup for each method
Observation Contract
- One island table per group of no more than 5, with observers and people carrying out actions in separate seats
- At least 10 cm of right-hand margin on each group's contract-sheet work surface for adding a check date at the end of each row
- One wall for posting together rows using the same record format
What, So What, Now What
- One island table per evenly sized group of 4–7, and one whole-group recording surface visible to everyone
Goal Experiment Measure Loop
- Set goals as desired states participants can influence and display that numerical achievement will not be used for penalties, guarantees, or personnel evaluation
- Before each experiment, identify affected perspectives, an implementation lead, specialists to consult, stopping conditions, and how to reverse it
- As a rule, collect no personal data for measurement. Use existing values or non-identifying aggregates and set purpose, collector, storage period, deletion date, and update date
- Explain the experiment and measurement plainly to affected people and first show consent, how to withdraw, and choices after withdrawal to delete, anonymise, or stop use of existing records
When to switch or stop
Fallbacks
- There are more than 10 observation items.
Narrow them to one item each before, during, and after the occurrence.
Method guide: Observation Contract
Stop conditions
- A serious incident, legal breach, or suspected misconduct emerges.
- Stop the improvement dialogue, preserve the records, and refer the matter for formal investigation.
- Ability appraisal using individuals' names begins.
- Anonymise the information and return to working conditions and mechanisms.
Sources, rights and limits
This site has composed this programme's purpose, connections between methods, facilitation, roles, outputs, and switching conditions. See each method's individual page for its sources and licence.
Original source records and their required notices are retained below. Their presence does not mean a new legal or cultural-safety clearance. Original citation titles are kept in their source language.
Sources and required notices for the 3 methods used
Observation Contract
- 対話の譜面編集部: 観察の契約 — Original description and diagrams
- Source: Taiwa no Fumen Editorial Team: Observation Contract | Original description and diagrams
- Attribution: Taiwa no Fumen Editorial Team | This site's text and diagrams
- Cultural caution: open-method screening complete: No closed, ritual, or sacred cultural practice is used as a source; declining participation, sensitive information, and power differences are addressed in each method's limits and branches.
- Limits of this record: The editorial team independently created the text, procedure, and diagram. No exclusive rights are claimed over general dialogue operations or ideas themselves.
What, So What, Now What
- Liberating Structures Fieldbook (2026): What, So What, Now What — Henri Lipmanowicz / Keith McCandless / Nancy White — Adaptation under CC BY-SA 4.0; changes made
- CC BY-SA 4.0
- Source: Liberating Structures Fieldbook (2026): What, So What, Now What — Henri Lipmanowicz / Keith McCandless / Nancy White | https://www.liberatingstructures.com/what-so-what-now-what | CC BY-SA 4.0 adaptation; changes made
- Source: CC BY-SA 4.0 | https://creativecommons.org/licenses/by-sa/4.0/
- Attribution: Henri Lipmanowicz and Keith McCandless; Keith McCandless and Nancy White | Original-source author and contributor credit
- Attribution: Taiwa no Fumen Editorial Team | Japanese adaptation and this site's own diagrams
- Changes and adaptation: Adapted as a Japanese facilitation score; changes made
- ShareAlike: Provided under CC-BY-SA-4.0
- Cultural caution: open-method screening complete: No closed, ritual, or sacred cultural practice is used as a source; declining participation, sensitive information, and power differences are addressed in each method's limits and branches.
- Limits of this record: Adapted with changes as a Japanese facilitation score under CC BY-SA 4.0 and provided under the same licence. Only the current page is covered; material from June 2026 or earlier, official icons, and third-party material were excluded.
Goal Experiment Measure Loop
- Open Practice Library: How to do GEMs — Adaptation under CC-BY-4.0; changes made
- CC-BY-4.0の適用条件 — CC-BY-4.0
- Source: Open Practice Library: How to do GEMs | https://openpracticelibrary.com/practice/gems/ | Adaptation under CC-BY-4.0; changes made
- Source: CC-BY-4.0 terms | https://creativecommons.org/licenses/by/4.0/ | CC-BY-4.0
- Attribution: Taiwa no Fumen Editorial Team | This site's text and diagrams
- Cultural caution: This manifest contains no structured cultural-safety assessment. A person checks context, the option to decline participation, and power differences.
- Limits of this record: Only the notices displayed in the public YAML were transcribed. No claim is made that individual rights do not exist.
Japanese resources and historical source records
Original Japanese plan and its print view · Japanese operation pack or starter materials
An English print edition is not provided here. The linked plans, print views and operation materials are Japanese resources.
Historical recorded publication holds
The following original hold statements are retained verbatim with their fixed source versions. They are not a statement of current acceptance, nor does this historical label resolve a source-specific restriction.
- English translation and rendered-page independent acceptance pending
- Method dependency source/text SHA binding and current source-specific rights verification pending
- Publication not cleared
Historical fixed-source and English-draft provenance
Japanese program SHA-256: 22388e98c4690354775a7b61097d7384a01a32b99706b8e1c6be5ac87aed5e1e
Fixed source context SHA-256: fc3980a6b5192777291c7031585e7795507a3c174a5454ca2d37c5d9e76d5a7d
English draft SHA-256: 1bd1651c78fbc9027e2d3163778688487db62c1708d1496f407b78dbe9a392f3
- observation-contract: source
70e5f59bb84942843d00c690048ab396fd5caa3fdca8dc9fa26b5fb4060166ad; English drafta700e2f5b974c571ba15dcb5c1156205d527285ea15ed08ccd71dd1ae74abaed - opl-goal-experiment-measure-loop: source
fe257cfadd32f02277e486ba8acff192334e4e3309764403eb8c35b14bbda9d3; English draft423f81ccb6fff2f9a3bffc17785a681e8e7d5f75f4c1d15e29cb9524e254b7ca - what-so-what-now-what: source
f0ca8ccf31289f712e5ca11086c04e4dd8e3c46c6e50067c2989378e202c2aa2; English draftf48f84412d26c71ba0b392fe00736dbf726180accd3095c47ea5a9a56d6a3da4
These hashes identify the checked inputs; they do not grant independent acceptance or publication clearance.