Murphy Diagram
Murphy diagrams are based on the premise that "if something can go wrong, it will go wrong." They are similar to other analysis methods such as fault trees as they analyze errors based on the potential causes of those errors. The diagrams utilize eight categories of behavior:
1. Activation or detection
2. Observation and data collection
3. Identification of system state
4. Interpretation of situation
5. Task definition and selection of goal state
6. Evaluation of alternative strategies
7. Procedure selection
8. Procedure execution
When analyzing failures retrospectively.
1. DEFINE THE TASK OR SCENARIO UNDER ANALYSIS. This is typically a retrospective look, although it is possible to use this method for predicting future events.
2. DATA COLLECTION. If conducting a retrospective analysis, collect the data available for the incident. If the goal of the analysis is to predict potential failures, conduct a walkthrough of the events in order to collect data.
3. DEFINE ERROR EVENTS. Begin by defining the first error as clearly as possible.
4. CLASSIFY ERROR ACTIVITY INTO DECISIONMAKING CATEGORY. After defining and describing the error, classify it using one of the eight behavior categories.
5. DETERMINE ERROR CONSEQUENCE AND CAUSES. After classifying the error, identify what the consequences were and what other potential consequences of the error are.
6. CONSTRUCT A MURPHY DIAGRAM. After identifying all existing and potential consequences, construct the Murphy diagram.
7. PROPOSE DESIGN REMEDIES. It is a good idea to include a column in the Murphy diagram for proposed error or design remedies.
Allows multiple potential causes for an error to be identified.
Offers useful application for tasks that involve teams, as it can portray teamwork and failures alongside team-based causes.
Any proposed solutions will be entirely subjective.
Analyst is provided little guidance.
Intricate, sizeable tasks can result in large diagrams that may not be particularly useful.
Stanton N, Salmon P, Walker G, et al. Process charting methods. Human factors methods: a practical guide for engineering and design. Great Britain: Ashgate; 2005. p. 109-37.