Using the relationship between the physical and abstract worlds gives us the following framework of questions to help solve problems:
State the problem
Identify the involved objects and their properties and extent
Detail how the ojects are related
Describe the observed actions of the objects and actions to those objects
List ideas of possible causes of the problem
Explore what the ideas predict
Test the ideas against the observed evidence and eliminate ideas that the evidence disproves
Design and complete additional tests if required
Defect problem solving
The approach to problem solving outline in the introduction can been seen in more detail when you apply it to defect problem solving. Imaging a system has been working and there is now a problem. The tables below detail how you can identify the problem and know how to fix it.
The problem
State the problem
Write a description of the problem.
Is
Is Not
What
Detail what objects are involved in the problem
Detail what objects are not involved but might have reasonably been expected
Where
State where the problem is located
State where the problem is not found but might have reasonably been expected to be
When
Specify when the problem(s) occurred
Specify when the problem(s) did not occur but might reasonably have been expected
Extent
Quantify the what, where and when for the problem. For example:
How many objects are affected?
How many at a particular location?
How many objects were affected over a period of time?
Quantify the what, where and when for everything that might reasonable have been expected to be part of the problem but is not. For example:
How many objects that might have been affected but were not?
How many objects not affect at a particular location
How many objects were not affected over a period of time?
Change
Detail changes made that preceded the problem :
What
Where
When
Extent
Detail what was not changed:
What
Where
When
Extent
Cause and affect
List the cause and effect actions between the objects involved
Proposed Causes
Proposed Causeof the problem
What does the proposed cause predict should happen?
What does the proposed cause predict should not happen?
Cause 1
Cause 2
etc
Elimination
Check for evidence that challenges the possible cause.
If there is evidence provisionally eliminate the possible cause.
Evaluate Possible Causes
Cause
Does not explain
Explains only if
If “X” is the true cause, how does it explain both the IS and IS NOT information?
What assumptions have to be made to explain the evidence?
State most probable cause
Of the causes tested against the observed behavior, which best explains the Is and IS NOT data?
Describe a test to support or disprove the cause
The possible cause(s) that are not provisionally eliminated are the mostly to be the actual cause.
Design and conduct tests for the remaining possible causes. By continuing the process you should arrive at the cause.