A Problem-Solving Evidence Plan defines what evidence the team needs next to reduce uncertainty in an investigation.
The purpose is to move from “we think this is the cause” to “here is the evidence that supports or rejects the hypothesis.”
Start with the question
Every evidence activity should answer a specific question.
Examples include whether the defect occurs only on one machine, whether temperature changes before the failure, whether the suspected cause reproduces the problem, or whether the countermeasure removes the effect.
Hypothesis-Driven Problem Solving helps frame suspected relationships as testable statements.
Identify the evidence type
Useful evidence may include direct observation, measurement, comparison, controlled trial, physical inspection, and historical data.
The strongest method depends on the question.
Do not collect data simply because it is available.
Define the measurement method
Data Collection Plan helps define what will be measured, how, where, when, and by whom.
Consistency matters when several people collect evidence.
Define the comparison
Evidence becomes stronger when the team can compare good versus bad, before versus after, one machine versus another, or suspected condition present versus absent.
Is-Is Not Analysis can help identify meaningful contrasts that narrow the investigation.
Define expected results before testing
Before running the test, state what should be observed if the hypothesis is true.
This reduces the risk of interpreting any result as support.
A test should be capable of rejecting the hypothesis.
Prioritize high-information tests
The next test should reduce the most important uncertainty with reasonable time, cost, and risk.
A fast discriminating test can be more valuable than a large uncontrolled data collection exercise.
Record the decision
Cause Hypothesis Log provides a useful place to record hypothesis, evidence, test, result, and decision.
The evidence plan defines what comes next; the log preserves what was learned.
Move only verified causes forward
Cause Verification helps confirm that the suspected cause actually changes the problem.
Permanent countermeasures should be based on verified cause, not confidence alone.
Common mistakes
Collecting data without a question, measuring everything, failing to define expected results, changing the hypothesis after seeing the evidence, accepting correlation as proof, using inconsistent measurement methods, and moving to countermeasures before important uncertainty is resolved are common mistakes.
Practical sequence
- define the current uncertainty.
- state the hypothesis or question.
- select the evidence type.
- define the measurement method.
- define the comparison.
- state the expected result.
- run the test.
- record the evidence.
- accept, reject, or refine the hypothesis.
- choose the next evidence step.
The practical lesson
A Problem-Solving Evidence Plan makes investigation intentional. Good problem solving is not about collecting more information; it is about collecting the information that changes the decision.