FREE online courses on Quality Improvement - Quality Improvement Tools - The Seven QC Tools

 

The seven QC tools are the most popular tools, which are being used by quality conscious companies throughout the world for improvement of quality of products and processes.

 

A brief description of these tools is presented here:

 

1 Check Sheet

 

What is a checksheet: A checksheet is a pre-designed format for collection of data that encourages organized collection and groups data into categories.  Categories are created in advance and may be added as needed.  A check mark is added for each example of a category.  The marks are added to determine subtotals.

 

When to use it:  To keep track of the parameters of an on going process.  It can be used to track events by such factors as timeliness (on time, one day late, two days late, etc.); reason for inspection failure (appearance, performance, etc.); person accomplishing the task (sales calls per representative); when something happens (customer complaints for each day of the month); and many others.

 

How to use it: Look at some preliminary data before developing the check sheet.  This will indicate what categories to use. For example, you might want to track employee mistakes by hour of the day or simply by whether they occur in morning or afternoon.  Include information about who collected the data, the date and the total sample from which it was drawn.

 

Example:  A sugar company wants to check the weight of bags of sugar that are supposed to weigh 100 kg.  The resulting checklist is shown in the following Table.  The checklist indicates that the machine is not filling all the bags within the specification.

Check Sheet

Sugar Bag Weight Check Sheet

          Lot No. : 463 Weigher : RTF

          Specification : 100kg. ± 1 Kg.Shift : 2

Weight (Kg.)  Tally      Total        

98.00            0

98.25  11      2

98.50  1111   4

98.75  11111111      8

99.00  111111111     9       

99.25  11111  5

99.50  11111111      8

99.75  11111111111111111  17

100.00 11111111111111111111111  23

100.25 111111111111          12

100.50 11111111      8

100.75 111111111     9

101.00 1111   4

101.25 111     3

101.50 1111   4

101.75 11      2

102.00 1        1

 

2  Graphs

 

Graphs are used for the purpose of comparison of visual representation of data collected. The most commonly used graphs are in the form of Bar charts, Line charts and Pie charts.

 

3  Histogram

 

What is it:  A display of the distribution of data by category.  It is a more effective way of displaying data than a table of figures because it gives a visual indication of relationships.

 

When to use it:  Use a histogram to observe whether data falls into a specific pattern.  The chart will indicate the extent of variation.  For example, it might show the frequency with which various levels of defects turn up, indicating that most lots have 15-20 defects.  A succession of histograms can be used for comparison.

 

How to use it:  Look at the data to be displayed and find the range (the value of the largest minus the value of the smallest).  Divide the range in between 5 and 10 categories.  Arrange the data range in ascending order on the horizontal scale, the numerical findings on the vertical scale.  Draw a bar for each category representing the value.

 

Example: A company produces a chemical in which a constituent named as Alpha should be present in the chemical more than 30 mgms. in a cubic centimeter of volume. Quality control department is measuring the characteristics of the sample received from the reactor and displays it in a histogram.

 

4  Pareto Chart

 

What is it:  A special form of a bar chart which seeks to determine the most important factors in a situation.  It is based on the idea that a few causes produce a vast majority of variations.  A rule of thumb says 20% of causes account for 80% of variations.

 

When to use it:  When you want to determine which corrective actions would yield the greatest quality payback.  A Pareto chart is a good way to set priorities and to focus your quality efforts.

 

How to use it:  Examine the data indicating the frequency with which each cause of a problem occurs. List them from most to least frequent.  Plot them on a bar graph. The left vertical scale indicates the frequency that each bar represents.  The right vertical scale indicates the percentage of total occurrences that is covered by the sum of the causes.  For each cause, add the per cent of problems that it accounts for to the per cent accounted for by the causes to the left to it and plot points against the right scale that represent this total.  Connect these points with a line.

 

Example: Let us take an example of an airline, which is trying to analyze and prioritise the quality complaints received from its customers. The complaint data are as below:

 

Type of complaint     Number

Baggage delay          23

Missed connections   15

Lost baggage           7

Poor cabin service     3

Ticketing error          2

 

From the pareto chart prepared as per the data, we can find out the relative magnitude of complaints and can identify the most important opportunity for improvement in quality of airline service. As shown in the Chart, 75% of customer complaints are related to baggage delay and missed connections only.

Based on this finding, the airline staff can use cause and effect diagram to figure out the root causes of these two major problems.

 

5  Cause and Effect Diagram

 

What is it:  This diagram represents the relationship between a problem and its potential causes.  It's also known as fishbone or Ishikawa diagram.  It deals only with factors, not quantities.

 

When to use it:  This is a good tool for organizing thinking about a quality problem.  It often stimulates ideas during brain storming sessions and prepares the way for an orderly investigation of the causes of a problem.  You can use it both to investigate current problems and to anticipate factors that may contribute to quality problems before they develop. This tool can also be used to list all the factors that contribute to a desired outcome.

 

How to use it:  First, define the effect or problem that you are going to analyze.  Write it in a box on the far right.  Next, list potential causes on a separate sheet of paper.  List all possible causes without regard to relationships.  Don't overlook causes because they seem improbable at first.  Classify these causes by themes.  Each theme represents a diagonal attached to the “spine” of the diagram.  The individual causes are listed along the diagonal.  Sub-branches can be created to break down factors in the causes.

 

Once you've constructed the diagram, you can go on to investigate the causes.  Common sense should point to a few prominent causes to look at first.  You can compare them by setting up a Pareto diagram.

 

Example:  A company wants to look at the causes for word processing errors.  The causes are organized according to the cause and effect diagram in the following Figure.  Three main causes of the error were identified as, client, time and typist. Various reasons to cause defect at these three classes are identified by cause and effect diagram, which are to be corrected from source level to eliminate the word processing errors.

 

6  Scatter Diagram

 

What is it:  A means for showing a relationship between two variables.  The diagram creates a coordinate for each variable, then plots the occurrences where the values intersect.

 

When to use it:  To find out whether there is a correlation between the variables.  It is often used to find the causes of problems.  For example, if you plot employee errors against the amount of continuous time on the job and find a correlation, then fatigue might be a factor in the errors.  If the correlation doesn't exist, other factors need to be investigated.

 

How to use it:  Establish vertical and horizontal axes with appropriate scales.  Usually, the horizontal axis is the one over which you have control.  Plot each data point.  Look at the pattern.  The more closely the dots group along an axis, the stronger the correlation.  The more scattered they are, the weaker the correlation.  If you determine a correlation, statistical analysis can give a more accurate indication of the relationship.

 

Example: A company wants to investigate whether the operators' errors are related to volume of work.  The number of errors per month are tracked for operators with different level of work volume.  The values are entered on a scatter diagram, as in the Figure.  The relationship of the data points indicates a strong positive correlation.  The more the volume of work, higher the errors.

 

7  Control Chart

 

What is it:  A means of monitoring a process according to tolerance limits.  The chart allows you to track the normal variations that indicate a process is in control and to determine when it goes out of control.  A control chart is closely related to Statistical Process Control.  It's a visual means of representing whether a process is within statistical limits.

 

When to use it:  For any process with frequent and measurable outcomes. A control chart enables you to ignore changes in a process that are the result of random variations and to react immediately to changes that indicate a problem.

 

How to use it:  Take a random sample of outcomes and use statistical techniques to determine upper and lower control limits.  These are not the same as specification tolerances.  Rather, they are the values which, if the outcomes exceed them, indicate that the outcomes are not the result of random variation, but of some specific cause.

 

Once the control limits have been determined, plot the outcomes over time or occurrence.  If a value is obtained that is outside the control limits,  it's necessary to investigate the cause.  If all the values are within the limits, then the process is under control.

 

Example: As shown in the diagram, an engineering company producing hardware components wants to control its shipping in order to have minimum inventory at its works using control chart.

 

Time is measured on the horizontal axis, which usually corresponds to the average value of the quality characteristic being measured. Two other horizontal lines represent the upper and lower control limits. These are chosen so that there is a high probability that sample values will fall within these limits if the process is under control or affected only by common causes of variation. If points fall outside the control limits or if unusual patterns, such as, shifts up or down, trends up or down, cycles and so forth exist, then there is reason to believe that special causes might be present.