Problem Analysis
Describe the Problem
State the Problem
- What object (or group of objects) has the deviation?
- What deviation does it have?
- What tells us a deviation exists? (senses, measures)
Specify the Problem
- Describe as completely, factually as possible
| IS | IS NOT | |
| WHAT | What specific object(s) has the deviation? | What similar object(s) could have the deviation, but does not? |
| What is the specified deviation? | What other deviations could be observed, but are not? | |
| WHERE | Where is the object when the deviation is observed? (geographically) | Where else could the object be when the deviation is observed, but is not? |
| Where is the deviation on the object? | Where else could the deviation be located on the object, but is not? | |
| WHEN |
When was the deviation observed first? (in clock and calendar time) |
When else could the deviation have been observed first, but was not? |
|
When since that time has the deviation been observed? What pattern?* |
When since that time could the deviation have been observed, but was not? What could be the pattern?* |
|
| When, in the object’s history or life cycle, was the deviation observed first? |
When else, in the object’s history or life cycle, could the deviation have been observed first, but was not? |
|
| EXTENT | How many objects have the deviation? |
How many objects could have the deviation, but do not? |
| What is the trend* in the number of objects with the deviation? |
What could be the trend* in the number of objects, but is not? |
|
| What is the size of a single deviation? | What could be the size, but is not? | |
| What is the trend* in the size? | What could be the trend,* but is not? | |
| How many instances of the deviation are on each object? |
How many instances could be on each object, but are not? |
|
| What is the trend* in the number of instances? |
What could be the trend* in the number of instances, but is not? |
- Pattern* = continuous, periodic, sporadic, single occurrence
- Trend* = increasing, decreasing, stable
Identify Possible Causes
1. Use knowledge or experience, or...
- Explain how the cause creates the deviation
- What object + deviation format
- What could have caused this deviation?
2. Use distinctions and changes
- Record data, new and true, only of the IS
- What is different, odd, special, or unique about each IS compared to IS NOT?
- What changed in, on, around, or about, each distinction? When?
- How could the distinctions and changes cause the deviation?
Evaluate Possible Causes
Test possible causes against the IS / IS NOT
- Eliminate causes that cannot be explained
- If (possible cause) is the cause of (problem), then how does it explain both IS and IS NOT?
- What assumptions must be made?
Determine most probable cause
- Pick the one with the fewest assumptions
- Which possible cause best explains the IS and IS NOT?
Confirm True Cause
Verify assumptions, observe, experiment, or try a fix and monitor
- Pick the safest, easiest, quickest, cheapest, surest way to prove true cause
- What can be done to confirm true cause?