The Quality Improvement Model

1
TheQualityImprovementModel
Define Process
Select Measures
Collect Interpret Data
Collect Interpret Data Displaying Measures
IsProcessStable?
No
Investigate Fix Special Causes
Purpose Begin collecting and analyzing data from
the process.
Yes
IsProcessCapable?
No
Improve Process Capability
Yes
Use SPC to Maintain Current Process
2
Graphical Tools for Displaying Measures from
Processes

• Run Charts
• Histograms
• Pareto Charts

3
Run Charts

• A plot of the data in time order.
• Time is on the horizontal axis and the data
  values are plotted on the vertical axis.
• Run charts show the process variation over time.

4
Histograms

• A bar chart showing frequency of occurrence is
  shown on the vertical axis.
• Histograms show the pattern of variation.

5
Pareto Charts

• A bar chart showing the relative importance of
  some observed characteristic.
• The frequency, percent or cost is shown on the
  vertical axis.
• The characteristic (type of defect, cause, etc.)
  is shown on the horizontal axis.
• The characteristic is usually plotted in order of
  decreasing magnitude.

6
Pump Maintenance
For each week (time period) record the number of
pump failures.
One possible run chart would be to plot the
number of pump failures for each week (time
period). The opportunity for failures should
remain constant from week to week.
Week Failures Failure Type 1 6 Seal,
Align... 2 1 Fitting, Seal... 3 2 Align,
Gear... 4 4 Seal, Fitting... . . . . . .
20 7 Align, Seal...
Collect information about causes for each
failure for use in a Pareto Chart. Pareto Charts
could also be based on pump location, pump
environment, etc.
7
Pump Maintenance Data
Run Chart
Number of Failures
Pareto Chart
Failures
60
50
40
Week
30
Histogram
Frequency
20
6
5
10
4
0
3
Seal
Alignment
Fitting
Gear
Other
2
Type Failure
1
0
0-1
2-3
4-5
6-7
8-9
10-11
12-13
Failures
8
Shipping Process
For a specified time period n Shipments
Made x Late Shipments p x/n
A good run chart would be to plot p for each time
period. A time period could be a week or month.
Week n Late p Reason 1 75 10 0.13
A,C,F... 2 84 6 0.07 B,F,A...
3 78 12 0.15 F,B,B... . . . . .
. . . . . 30 70 10 0.14 B,F,I...
It would also be good to collect other
information about the late shipments for use in a
Pareto Chart.
9
Shipping Data
Run Chart
Proportion Late
Week
Histogram
Frequency
Proportion Late
10
Purchase Order Process
Completed Purchase Orders
Purchase OrderProcess
2,7,5,4,5 3,10,2,5,3 5,7,3,12,1
4,7,8,3,5 3,3,9,2,4
Week 1 Week 2 Week 3 Week 4
Week 20
5 Purchase Orders are selected each week. The
time (in days) it took to process each of the 5
POs is recorded, and the average of the 5
calculated. The average is the measure tracked.
A possibility would be to subgroup the data( i.e.
combine 5 purchase orders and plot their average.)
Week A B C D E Average 1 2 7 5 4 5 4.6
2 3 10 2 5 3 4.6 3 5 7 3
12 1 5.6 4 4 7 8 3 5 5.4 . . . . . .
. . . . . . . . 20 3 3 9 2 4 4.2
It might also be informative to plot a histogram
of all the times to see the pattern of variation.
11
Purchase Order Data
Histogram of 100 total observations
Frequency
Run Chart of 20 Averages (of size 5)
Time (Days)
Time (Days)
Week Sample Taken
12
Polymer Manufacturing Process
Material Produced (lots)
ProductionProcess
Samples
A quality characteristic is measured on each
sample.
One possibility would be to collect a sample of
the product every 4 hours, and measure the
characteristic of interest on that sample. A run
chart could then be constructed of this data.
Sample b 1 1.51 2 1.89 3 1.42 . .
. . 134 1.63
It would also be informative to plot a histogram
of all the times to see the pattern of variation.
b is a measure of yellowness
13
Polymer Manufacturing Data
Note b is a measure of yellowness
14
Minitab Open the Dataset

• Real data from Organic Chemicals Division.
• Found in the Hq file in the Minitab Datasets
  folder.

15
Minitab Making a Run Chart
Make a run chart of the HQ Water data by
following the instructions below. The final
output should look like this
16
Minitab Creating a Histogram
Make a histogram of the HQ Water data by
following the instructions below. The final
output should look like this
17
Minitab Copy and Paste Output into PowerPoint
If you paste the graph into PowerPoint, you can
double-click the graph to edit it using
MINITAB. To make the graph static and reduce
its file size, select Edit ? Paste Special
and select a Picture format.
18
The Catapult Process
Landing Point
Distance (inches)

• 30 Consecutive deliveries were made.
• One set of conditions was used.

19
Catapult Data
Run Chart
Distance (inches)
Histogram
Frequency
7
6
5
4
3
2
1
0
100
105
110
101
102
103
104
106
107
108
109
111
112
113
Distance (inches)
2
4
6
8
10
12
14
16
18
20
22
24
26
28
30
Run Order
20
Exercises

• 1.) Your Catapult Team should complete pages 6
  and 7 of the blue Catapult Process handout.
• 2) Be ready to make a PowerPoint presentation
  of your results.
• Limit yourselves to 15 minutes for this
  exercise.