Run Charts
0 views
Skip to first unread message
Steven Bonacorsi
Aug 17, 2011, 12:26:39 AM8/17/11
to Lean Six Sigma
Adding the element of time will help clarify your understanding of the
causes of variation in your processes. A run chart is a line graph of
data points organized in time sequence and centered on the median data
value. The patterns in the run chart can help you find where to look
for assignable causes of variation.
What can it do for you?
Histograms or frequency plots can show you the general distribution or
variation among a collection of data points representing a process,
but one histogram or one frequency plot can not show you trends or
help pinpoint unusual events. Sometimes, a normal-looking distribution
will hide trends or other unusual data. To spot those trends, the data
must be considered in time order. Plotting data on a run chart can
help you identify trends and relate them to the time they occurred.
This will help you in your search for the special causes that may be
adding variation to your process.
Run charts are especially valuable in the measure and analyze phases
of Lean Six Sigma methodology.
How do you do it?
1.Select a characteristic from one of your processes. This
characteristic could be presenting a problem because excessive
variation often drives it outside of specification limits, or it could
be a cause of customer complaints.
2.Measure the characteristic over time intervals and record the data.
Note the time or the time period that is associated with each data
point.
3.Find the median data value. To do this, list the data values in
numeric order. Include each data point, even if it is a repeat value.
If the number of data points is odd, the median is the middle value.
If the number of data points is even, the median is halfway between
the two values nearest the middle. For example, if the collected data
points were: 5, 1, 18, 8, 12, 9, the ordered values would be: 1, 5, 8,
9, 12, and 18. The middlemost values are 8 and 9. The median is the
average of those values, or 8.5. (Remember, the numerically-ordered
data is only for determining the median. The data must be plotted in
time order on the run chart to be of any value.)
4.Set up the scales for your run chart. The vertical scale will be
the data values, and the horizontal scale will be the time. Make the
horizontal scale about two to three times the distance of the vertical
scale.
5.Label the vertical scale so that the values will be centered
approximately on the median and so the scale is about 1 ½ to 2 times
the range of the collected data.
6.Draw a horizontal line representing the median value.
7.Plot the data points in sequence. Connect each point to the next
point in the sequence with a line.
Some special cause variation reveals itself in unusual run-chart
patterns. These clues can direct you in your search for causes. Count
the number of runs. Runs are sequences of points that stay on one side
of, either above or below, the median line. One way of counting the
runs is to circle these sequences and tally them. Another way of doing
this is to count the number of times the run-line crosses the median,
and then add one. Compare the number of runs you count to the
accompanying chart.
•Numbers of runs outside the range shown for the number of data points
are statistically unusual.
•Too few runs (below the lower limit) generally indicate that
something cyclic is systematically shifting the process average.
•Too many runs could point to a problem of consecutive, over-
compensating process adjustments or indicate that the data points
actually came from two sources with different process averages.
•Look for sequences of ascending or descending values. Seven or more
continuously increasing or continuously decreasing points indicates a
trend that is shifting the process average. When counting points,
ignore any points that repeat the previous value. Repeated values
neither add to the length of the run nor break it.
•Search for seven or more consecutive points on the same side of the
median line or 10 of 11 points or 12 of 14 or 16 of 20. (Ignore any
points that are exactly on the median.) Such a sequence indicates that
something has occurred to shift the process average in that direction.
•A sequence of 14 or more data points alternating up and down
suggests a variation related to sampling (such as one reading early in
the day and one reading toward the end) or that the data is coming
from two sources with different process averages (such as from two
machines making the same part.) In looking for up-and-down
alternation, ignore any points that are exactly the same as the
preceding point.
•A sequence of seven or more points with exactly the same value
usually should signal you to look for a special cause. While it is
possible that your process has improved to the extent that the
existing measurement technique is no longer sensitive enough to
measure variation, it is usually more probable that a gauge is stuck
or broken or that someone is making up the data.
Now what?
Run charts can be very valuable in helping your search for sources of
variation. They are easy to plot and easy to interpret. The sampling
is uncomplicated, and there are no statistical computations to make.
They can also be applied to almost any process or any data.
On the other hand, they are not an instant indicator. They are best
used for spotting trends; short shifts in the process cannot always be
detected with run charts. In addition, special causes that produce
general piece-to-piece variation will not be readily detected on run
charts.
Finally, a simple run chart cannot establish the natural capabilities
of a process, so it isn't possible to use one to predict what
specifications a process can actually meet. To do that, you need to
create a control chart, a run chart with statistical control limits.
Steven Bonacorsi is a Certified Lean Six Sigma Master Black Belt
instructor and coach. Steven Bonacorsi has trained hundreds of Master
Black Belts, Black Belts, Green Belts, and Project Sponsors and
Executive Leaders in Lean Six Sigma DMAIC and Design for Lean Six
Sigma process improvement methodologies. Bought to you by the Process
Excellence Network the world leader in Business Process Management
(BPM)
Author for the Process Excellence Network (PEX Network / IQPC)
Process Excellence Network
Steven Bonacorsi, President of International Standard for Lean Six
Sigma(ISLSS)
Certified Lean Six Sigma Master Black Belt
47 Seasons Lane
Londonderry, NH 03053
Phone: +(1) (603) 401-7047
E-mail: sbona...@islss.com
Process Excellence Network: http://bit.ly/n4hBwu
ISLSS: http://www.islss.com
Article Source: http://EzineArticles.com/?expert=Steven_Bonacorsi
causes of variation in your processes. A run chart is a line graph of
data points organized in time sequence and centered on the median data
value. The patterns in the run chart can help you find where to look
for assignable causes of variation.
What can it do for you?
Histograms or frequency plots can show you the general distribution or
variation among a collection of data points representing a process,
but one histogram or one frequency plot can not show you trends or
help pinpoint unusual events. Sometimes, a normal-looking distribution
will hide trends or other unusual data. To spot those trends, the data
must be considered in time order. Plotting data on a run chart can
help you identify trends and relate them to the time they occurred.
This will help you in your search for the special causes that may be
adding variation to your process.
Run charts are especially valuable in the measure and analyze phases
of Lean Six Sigma methodology.
How do you do it?
1.Select a characteristic from one of your processes. This
characteristic could be presenting a problem because excessive
variation often drives it outside of specification limits, or it could
be a cause of customer complaints.
2.Measure the characteristic over time intervals and record the data.
Note the time or the time period that is associated with each data
point.
3.Find the median data value. To do this, list the data values in
numeric order. Include each data point, even if it is a repeat value.
If the number of data points is odd, the median is the middle value.
If the number of data points is even, the median is halfway between
the two values nearest the middle. For example, if the collected data
points were: 5, 1, 18, 8, 12, 9, the ordered values would be: 1, 5, 8,
9, 12, and 18. The middlemost values are 8 and 9. The median is the
average of those values, or 8.5. (Remember, the numerically-ordered
data is only for determining the median. The data must be plotted in
time order on the run chart to be of any value.)
4.Set up the scales for your run chart. The vertical scale will be
the data values, and the horizontal scale will be the time. Make the
horizontal scale about two to three times the distance of the vertical
scale.
5.Label the vertical scale so that the values will be centered
approximately on the median and so the scale is about 1 ½ to 2 times
the range of the collected data.
6.Draw a horizontal line representing the median value.
7.Plot the data points in sequence. Connect each point to the next
point in the sequence with a line.
Some special cause variation reveals itself in unusual run-chart
patterns. These clues can direct you in your search for causes. Count
the number of runs. Runs are sequences of points that stay on one side
of, either above or below, the median line. One way of counting the
runs is to circle these sequences and tally them. Another way of doing
this is to count the number of times the run-line crosses the median,
and then add one. Compare the number of runs you count to the
accompanying chart.
•Numbers of runs outside the range shown for the number of data points
are statistically unusual.
•Too few runs (below the lower limit) generally indicate that
something cyclic is systematically shifting the process average.
•Too many runs could point to a problem of consecutive, over-
compensating process adjustments or indicate that the data points
actually came from two sources with different process averages.
•Look for sequences of ascending or descending values. Seven or more
continuously increasing or continuously decreasing points indicates a
trend that is shifting the process average. When counting points,
ignore any points that repeat the previous value. Repeated values
neither add to the length of the run nor break it.
•Search for seven or more consecutive points on the same side of the
median line or 10 of 11 points or 12 of 14 or 16 of 20. (Ignore any
points that are exactly on the median.) Such a sequence indicates that
something has occurred to shift the process average in that direction.
•A sequence of 14 or more data points alternating up and down
suggests a variation related to sampling (such as one reading early in
the day and one reading toward the end) or that the data is coming
from two sources with different process averages (such as from two
machines making the same part.) In looking for up-and-down
alternation, ignore any points that are exactly the same as the
preceding point.
•A sequence of seven or more points with exactly the same value
usually should signal you to look for a special cause. While it is
possible that your process has improved to the extent that the
existing measurement technique is no longer sensitive enough to
measure variation, it is usually more probable that a gauge is stuck
or broken or that someone is making up the data.
Now what?
Run charts can be very valuable in helping your search for sources of
variation. They are easy to plot and easy to interpret. The sampling
is uncomplicated, and there are no statistical computations to make.
They can also be applied to almost any process or any data.
On the other hand, they are not an instant indicator. They are best
used for spotting trends; short shifts in the process cannot always be
detected with run charts. In addition, special causes that produce
general piece-to-piece variation will not be readily detected on run
charts.
Finally, a simple run chart cannot establish the natural capabilities
of a process, so it isn't possible to use one to predict what
specifications a process can actually meet. To do that, you need to
create a control chart, a run chart with statistical control limits.
Steven Bonacorsi is a Certified Lean Six Sigma Master Black Belt
instructor and coach. Steven Bonacorsi has trained hundreds of Master
Black Belts, Black Belts, Green Belts, and Project Sponsors and
Executive Leaders in Lean Six Sigma DMAIC and Design for Lean Six
Sigma process improvement methodologies. Bought to you by the Process
Excellence Network the world leader in Business Process Management
(BPM)
Author for the Process Excellence Network (PEX Network / IQPC)
Process Excellence Network
Steven Bonacorsi, President of International Standard for Lean Six
Sigma(ISLSS)
Certified Lean Six Sigma Master Black Belt
47 Seasons Lane
Londonderry, NH 03053
Phone: +(1) (603) 401-7047
E-mail: sbona...@islss.com
Process Excellence Network: http://bit.ly/n4hBwu
ISLSS: http://www.islss.com
Article Source: http://EzineArticles.com/?expert=Steven_Bonacorsi
Reply all
Reply to author
Forward
0 new messages