The Psychology of Quality and More

A Toolbook for Quality Improvement and Problem Solving (contents)

Bar Chart: How to do it

The Quality Toolbook > Bar Chart > How to do it

When to use it | How to understand it | Example | How to do it | Practical variations

<-- Previous | Next -->

How to do it

1. Identify the purpose of the Bar Chart, including the questions it may answer and decisions that may be made from it. Common uses include:

• Analyzing the change in measurements across time.
  
• Analyzing the change in measurements for each of a set of related subjects.
  
• Showing the above with additional breakdown within each measurement point.
  
• Comparing several separate sets of measures.

2. Decide on the data that must be displayed in the chart in order to meet the purpose of step 1.

A typical measurement is of all events of a specific type within a defined time period. for example, 'Value of all cosmetics sales within one financial year'. There may be several data ranges, for example, 'Sales figures for this year and last year' (which are to be compared in the chart).
 

3. Identify how individual bars are to be made up. Common methods include:

• Subdivision of total data. For example, all cosmetics sales by month.
• Subdivision by item. For example, 12 months' sales by each cosmetic type.
• Combination of the above. For example, each bar is one month's sales, subdivided again by cosmetic type.

4. Select the number of bars in the chart. If multiple groups of bars are to be shown, avoid over-complication by limiting the number of these groups. As a guideline, it should be possible to clearly see the color or shading of each bar on the final chart. Typically, this will result in a limit of around 10 to 15 bars.

For example, a year's sales figures are better shown as 12 monthly bars, rather than 52 weekly bars. However, if this year's sales are being compared with those of the past 2 years (using groups of 3 bars), then it may be better that each bar represents 3 months, which will again result in 12 bars on the chart.
 

5. Draw a sample chart, and check if it can meet its purpose, both in readability and in the decisions that may be made from it. If necessary, revise the above decisions.

 
6. Collect the data, for example with a Check Sheet. The results may be displayed in a table, such as below.

Ensure that the data collection process gives reliable data, for example by using trained people.

7. Plot the chart, using appropriate scales. Label it to identify it uniquely and help with any subsequent decision-making.
8. Review the chart and act on any identified decision points, such as:

• Bars above or below target value.
• Bars of unexpected height, particularly relative to other bars.
• Trends across sets of bars.

<-- Previous | Next -->