# The problem solving toolkit - An easy guide

## Problem identification and solution - the right tool for the job!

#### 1. The problem solving action cycle

• The action cycle provides a method for controlling the problem solving process.
• Problem solving without a method is random and doomed to failure. The only exercise you will get is jumping to conclusions!
• The cycle is repeated as often as is required.

### Always follow the action cycle!

#### 2. Pareto principle (80:20 Rule)

• Separate the vital few from the trivial many.
• Group the data into logical categories with largest category first and "Other" last.
• Plot the cumulative result.

### Start work on the vital few first!

#### 3. Cause and effect

• Identify potential causes and areas to start work.
• Use with operators not managers. They know the real process.
• Get all the causes and then rank using Pareto.

### Solve the cause, don't just treat the effect!

#### 4. Flow charts

• Map the actual process.
• Find the decision points and areas of confusion.
• Improve the process to meet the ideal.

### If a process cannot be made into a flow chart it will not work!

#### 5. Mind maps

• Get all of your ideas down, small notes only.
• Organise by Main Ideas first.
• Do not try to prioritise or number when mapping. Get it down on paper first.
• Look for relationships. Let one note spark another.

### Get the ideas down on to paper.

#### 6. Histograms

• Get lots of numbers in an easily understood visual form.
• Group the numbers into cells or ranges.
• Plot to give a histogram, look for the average and the shape (distribution).

### Easy to understand and decide on action.

#### 7. Scatter charts

• Quickly see if two variables are related.
• Go for quick and dirty numbers.
• Correlation is not causality. A relationship does not indicate a direct connection only that Variable 1 influences Variable 2 - there may be a deeper connection.

### Find the relationship quickly!

#### 8. Run Charts and Control Charts

• Good way to show data over time.
• Collect data over time. Look for the average.
• Consider the moving average for rapidly changing results.
• Set control limits for upper and lower boundaries.
• Look for "out of control" patterns. Points outside control limits, runs up or down, points always under or over the average value.