APS 4742 - Statistical

Applications in Industry
and Project Presentation
Quality Control Tools for Six Sigma

Dr. Malima Atapattu

Department of Statistics and Computer Science

1
• In this section, we discuss seven tools that have proven extremely
effective in helping organizations control processes and implement
process improvement projects.

• Seven of the tools were nicknamed the “magnificent seven”.

• Flow chart
• Check sheet
• Control chart
• Histogram
• Pareto chart
• Scatterplot
• Cause and effect diagram.

2
1. Flow chart
• Flow charts are used to describe a process being studied or to
describe a desired sequence of a new, improved process.
• Often this is the first step taken by a team looking for ways to improve
a process.
• The differences between how a process could work and how it
actually does work expose redundancies, misunderstandings, and
general inefficiencies.

3
Flow
Chart of
Filling an
Order

4
2. Check Sheet

• Check Sheets are basic manual data collection mechanisms.

• They consist of forms designed to tally the total number of
occurrences of certain events by category.
• They are usually the starting point of data collection efforts.
• In setting up a check sheet, one needs to agree on the category
definitions, the data collection time frame, and the actual data
collection method.

5
A Typical Check sheet

6
3. Control Chart
• The control chart is a graph used to study how a process changes over
time.
• Data are plotted in time order.
• A control chart always has a central line for the average, an upper line
for the upper control limit, and a lower line for the lower control limit.
• These lines are determined from historical data.

7
Control Chart (continued…)
• By comparing current data to these lines, one can make conclusions
about whether the process variation is consistent (in control) or is
unpredictable (out of control, affected by special causes of variation).
• There are many types of control charts. Each is designed for a specific
kind of process or data.
• Deciding which control chart should be used depends on the data
type.
• The two broadest groupings are for variable data and attribute data.

8
When to use
the basic
control
charts.

9
Basic Procedure
1. Choose the appropriate control chart for your data.
2. Determine the appropriate period for collecting and plotting the
data.
3. Collect data, construct your control chart, and analyze the data.
4. Look for out-of-control signals. When one is identified, flag it on the
chart and investigate the cause. Document how you investigated,
what you learned, the cause, and how it was corrected.

10
Types of Out-of-Control Signals • A single point outside the
control limits. (16)
• Two out of three successive
points are on the same side of
the centerline and farther than
2σ from it. (4)
• Four out of five successive
points are on the same side of
the centerline and farther than
1σ from it. (11)
• A run of eight in a row are on
the same side of the
centerline. Or 10 out of 11, 12
out of 14, or 16 out of 20. (21)
• Obvious consistent or
persistent patterns that
suggest something unusual
about your data and your
process.
11
Basic Procedure (continued…)
5. Continue to plot data as they are generated. As each new data point
is plotted, check for new out-of-control signals.
6. When you start a new control chart, the process may be out of
control. If so, the control limits calculated from the first 20 points
are conditional limits. When you have at least 20 sequential points
from a period when the process is operating in control, recalculate
control limits.

12
Control Limits
• Control limits are not specification limits. In fact, specification limits
should never be placed on a control chart. Specifications reflect
customer requirements, whereas control limits reflect the historical
performance of the process.
• Control limits should be recalculated only when the process has
experienced a permanent change from a known cause and at least 20
plot points have been generated by the changed process. You should
not recalculate limits after each 20 points or subgroups or after each
page of the chart.
• Choose numerical scales that include the values for the control limits
plus a little extra.
13
𝑋 and 𝑅 Chart (Averages and Range Chart)
• The 𝑋 and R chart is a pair of control charts used to study variable
data.
• It is especially useful for data that do not form a normal distribution,
although it can be used with normal data as well.

Construction
1. Determine the appropriate period for collecting and plotting data.
Determine the number of data points per subgroup (𝑛). Collect at
least 20𝑛 data points to start the chart.
For example, with a subgroup size of three, 60 data points will be
needed to create 20 subgroups.
14
𝑋 and 𝑅 Chart Construction (continued)
2. If the raw data do not form a normal distribution, check whether
the averages of the subgroups form a normal distribution. (A
normal probability plot can be used.) If not, increase the subgroup
size.
3. Calculate 𝑋 (X-double-bar), 𝑅 (R-bar), and the control limits for the
X and R chart.

15
Control Chart Constants for 𝑋 and 𝑅 Chart

16
𝑋 and 𝑅 Chart Construction (continued)
4. On the “Average” part of the chart, mark the numerical scale, plot
the subgroup averages, and draw lines for the average of the
averages and for the control limits for X. On the “Range” part of the
chart, mark the numerical scale, plot the ranges, and draw lines for
the average range and for the control limits for R.
5. Continue to follow steps 4, 5, and 6 of the basic procedure.

17
Example - 𝑋 and 𝑅 Chart
• A team collected a set of 40 values, arranged in time sequence, for
product purity.

18
Example - 𝑋 and 𝑅 Chart (continued…)

1.880

3.267

1990.8

1990.8 20 99.54
3.6

3.6 20 0.18
19
Example - 𝑋 and 𝑅 Chart (continued…)

1.880 0.18 0.34

99.54 0.34 99.88

99.54 0.34 9.20

3.267 0.18 0.59

20
Example - 𝑋 and 𝑅 Chart (continued…)

21
𝑋 and 𝑠 Chart
An 𝑋 and 𝑠 chart is very similar to an 𝑋 and 𝑅 chart, with one
exception: the standard deviation of the sample, 𝑠, is used to estimate
subgroup variation. Computer software is desirable for performing the
calculations.

Procedure
Follow the procedure for constructing an 𝑋 and 𝑅 chart. However,
instead of an 𝑅 chart, you will construct an 𝑠 chart.

22
𝑋 and 𝑠 Chart

23
Chart of Individuals (𝑋 − 𝑅 Chart)

• The chart of individuals is a pair of control charts used to study

variable data that are not generated frequently enough for an 𝑋 and
𝑅 chart and that form a normal distribution.
• Data are not sub-grouped.
• Individual data points are plotted on one chart, and the differences
between successive points, the moving ranges, are plotted on the
second.

24
Chart of Individuals - When to Use?
• When you cannot use an 𝑋 and 𝑅 chart because frequent data are
costly or not available, such as with destructive testing or a slowly
changing process, or …
• When you cannot use an 𝑋 and 𝑅 chart because the measurement
remains constant for a relatively long period before the process
changes, such as in batch operations where conditions within each
batch are constant.

25
Procedure of 𝑋 − 𝑅 Chart
1. Determine the appropriate period for collecting and plotting data.
Collect at least 20 data points, arranged in time order, from the
process to be studied.
2. Determine if the distribution of the data is normal. See the
histogram for a discussion of normal distribution and see normal
probability plot for ways to determine whether it is normal.
3. Calculate the average, called 𝑋 and the control limits.

26
Procedure of 𝑋 − 𝑅 Chart (Continued…)

27
Procedure of 𝑋 − 𝑅 Chart (Continued…)

28
Procedure of 𝑋 − 𝑅 Chart (Continued…)
4. On the “value” part of the chart, mark the numerical scale, plot the
individual values, and draw lines for the average and the control
limits for 𝑋.
5. On the “moving range” part of the chart, mark the numerical scale,
plot the moving ranges, and draw lines for the average moving
range and the 𝑈𝐶𝐿 . (𝐿𝐶𝐿 is zero.)
6. Continue to follow steps 4, 5, and 6 of the basic procedure.

29
𝑋 − 𝑅 Chart – Analysis and Considerations
Analysis
1. Check the “moving range” part of the chart for out-of-control signals.
2. If the chart is in control, check the “value” part of the chart for out-of-
control signals.
Considerations
• The chart of individuals does not pick up process changes as quickly as the
𝑋 and 𝑅 chart. If you have a choice between them, use the 𝑋 and 𝑅 chart.
• When data come from a nonnormal distribution (such as a skewed
distribution), the control limits of a chart of individuals will not represent
the process accurately. Depending on the situation, those limits will result
in many more or many less out-of-control signals than are actually
occurring. Use an 𝑋 and 𝑅 chart or moving average–moving range chart
instead.
30
𝑋 − 𝑅 Chart –Considerations
• Although the same word, range, is used in this chart and the 𝑋 and
𝑅 chart, they are calculated and interpreted differently.
• The chart of individuals does not separate variation in the process
distribution’s width from variation in the average, as the 𝑋 and
𝑅 chart does.

31
Chart of Individuals - Example.
• There is an out-of-control
signal around points 70 to 76:
seven points in a row below
the average.
• Notice the period between
reports 95 and 117 when the
process went wildly out of
control. The team identified a
special cause: new people in
the group.
• Immediately following that
period, the team implemented
its first improvements. Notice
the drop in both the average
and the UCL.

32