Seven QC Tools
Tool #5: Part 1-Run Chart
Lesson Structure
Run Chart
STEP 1: Determine what to measure
STEP 2: Draw the graph
STEP 3: Plot the data.
Follow-up: Decide On The next Steps
Summary
Run Chart
Run Chart is a type of line graph used as an analysis tool to:
Collect and interpret data
Create a picture of what is happening in the situation you are analyzing
Find patterns yielding valuable insights
Compare one period of data to another, checking for changes
The Run Chart is similar to the Scatter Diagram in that two variables are plotted against
each other. However, unlike the Scatter diagram, which is used to investigate cause and
effect relationships between two variables, the Run Chart tracks changes in the variable
being measured over a period of time to identify patterns.
Creating a Run Chart consists of three major steps:
STEP 1: Determine what to measure
STEP 2: Draw the graph
STEP 3: Plot the data.
The example that follows shows how one professional used the Run Chart to verify a
hunch.
Ravinder, an engineering Analyst,…………..
realized it was a part of life that the computer system would go down, but this was
ridiculous. Twice today and four times this week it shut down, and it was only Wednesday!
Ravinder felt a clear pattern was developing, since the machine always seemed to go down
around the same time. He had used the Run Chart in the past to prove similar hunches;
this seemed like the perfect opportunity to use it again,………
Step 1: Determine What to Measure
The first step in constructing a Run Chart is selecting one key measure to track over a
period of time. This measure should be a “quality or productivity” (external customer or
internal process) indicator providing useful information for making decisions.
Note: Measure can also be tracked against other bases, such as production batches, shifts and soon.
Possible measures that can be plotted include:
Volume (i.e. how much over a
specified period of time).
Cycle time (i.e. how long
something takes)
Errors (i.e. how many incorrect
over a period of time)
Waste (i.e. how much is
reworked or rejected).
After you have determined what to measure, decide on the period of time during which you
will collect data. Possible time intervals include: hourly, daily, weekly, monthly, quarterly,
etc.
Computer down time for July 200X
19
18
17
Minutes of computer down
16
15
14
13
12
11
10
time
9
8
7
6
5
4
3
2
1
0
1 3 5 7 9 11 13 15 17
Days in July
Diagram 5.1, Draw the Run Chart
Ravinder determined that his quality measure…………..
would be “minutes of computer down time”. He decided to start with a daily measurement
but realized the measurement may have to be changed to hourly at some point in the future.
Step 2: Draw the Graph
Drawing the Graph consists of three simple steps:
Name the graph. The name should describe what you are measuring and the time
duration you have chosen.
Draw the vertical axis. The measure is always on the vertical axis, and is shown in
number of occurrences or percentage.
Draw the horizontal axis. The time interval or other measurement base is always
shown on the horizontal axis.
With the quality measurement chosen……….
and the time interval determined, the rest was a breeze. Ravinder named graph “Computer
down time for July 200X” and then drew the vertical and horizontal axes. He labeled
vertical axis “minutes of computer down time” and the horizontal axis “Days in July”.
(See Diagram-5.1).
Step 3: Plot the Data
Collect the data for your chosen measurement. Each data point should then be plotted on
the graph in the appropriate location given its level of occurrence (or percentage) and time
interval. If you connect the points it will be easier to interpret the graph. Finally, calculate
the average occurrence or percentage and plot that on graph. The average line will help you
to see abnormalities in the process condition (i.e., something other than random points
around the average line).
Follow these tips for collecting and plotting data:
Plot the data points in the order they occur.
Collect the data on regular basis.
Evaluate the data on a regular basis.
Computer down time for July 200X
19
18
Minutes of computer down time 17
16
15
14
13
12
11
10
9
8
7
6
5
4
3
2
1
0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17
Days in July
Diagram-5.2 Plot the data
As the days in July went by…
Ravinder continued to plot the data on his Run Chart. The other workers in Ravinder’s
department began to take interest in the chart as the month progressed because they
could finally see more down time existed than they had previously thought. Ravinder
was careful to plot down time as it occurred, thus ensuring the accuracy of the
measurement. He had also connected the data points to help him see both the trends and
the extreme levels of variation in the amount of down time (see diagram-5.2).
Follow–up: Decide On Next Steps
Once you have plotted the graph, you have the option to do one or more of the
following:
Search for patterns in the data (e.g.
errors are higher on Tuesday
during the second shift)
Determine the root cause of the
error (e.g. a new employee is
working on Tuesday’s second shift
and has not been trained yet).
Investigate extreme highs or lows
in data points (extreme variations
around the average line indicates
opportunity for improvement).
Continue measuring to track the
effect of changes (i.e. you can
prove that the changes are working
by tracking the data).
Create a control chart to provide
more information about process
Computer down time for July 200X
19
18
17
16
Minutes of computer down time
15
14
13
12
11
10 9.76 min.
9 Average
8
7
6
5
4
3
2
1
0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17
Days in July
Diagram-5.3 Interpret the data
At the end of the month,…………
Ravinder calculated the average down time and drew the line on the graph. His original
hunch seemed correct. A pattern was developing and he decided to continue collecting
data to further verify it. Also present were some extreme highs and lows in the data points
on the graph (see diagram-5.3). This looked unusual to Ravinder, so he decided his next
step would involve converting the Run Chart to the Control Chart. That way he could
discover whether the down time was normal or whether something truly unusual was
taking place.
Summary
In summary, use the Run Chart when:
Trying to show trends in measurements over time or comparisons from one
measurement base to another (such as different shifts, teams, production runs, etc.).
Since the Run Chart is always based on some comparable measurement, such as
time, comparison of data is quick and easy.
You need to check for long-run changes. Often data changes slowly over time. The
Run Chart, if regularly updated, will show overall changes in the process average,
which may indicate underlying instability in process.
You want to show process variations. The Run Chart will help you to spot extremes
in process variations.
You want to evaluate changes made to the process. After improvements have been
made to the process, continue to collect data using the Run Chart to verify the
effectiveness of the change.
