CHAPTER 8
8-20. Health care facilities must conform to certain standards in submitting bills to Medicare/Medicaid for processing. The number of bills with errors and the number sampled
are shown in Table 8-17. Construct an appropriate control chart and comment on the
performance of the billing department. Revise the control limits, if necessary, assuming
special causes for out-of-control points. Comment on the capability of the department
= 170
i=1
n = 9900
Centerline
p =
170
= 0.0172
9900
(0.0172)(1 0.0172)
400
= 0.0172 + 0.0195 = 0.0367
r
(0.0172)(1 0.0172)
LCL = 0.0172 3
400
= 0.0172 0.0195 = 0
U CL = 0.0172 + 3
(0.0172)(1 0.0172)
300
= 0.0172 + 0.0195 = 0.0397
r
(0.0172)(1 0.0172)
LCL = 0.0172 3
300
= 0.0172 0.0195 = 0
U CL = 0.0172 + 3
(0.0172)(1 0.0172)
500
= 0.0172 + 0.0195 = 0.0367
r
(0.0172)(1 0.0172)
LCL = 0.0172 3
500
= 0.0172 0.0195 = 0
U CL = 0.0172 + 3
From the above data, we get the following p-chart. It can be observed that the 10th observation is above
the UCL.
Assuming special causes and deleting the 10th observation, the revised centerline is:
p =
155
= 0.0161
9600
Hence, the revised control limits are as follows: For observation 1-6:
r
(0.0161)(1 0.0161)
U CL = 0.0161 + 3
400
= 0.0161 + 0.0189 = 0.0350
r
(0.0161)(1 0.0161)
LCL = 0.0161 3
400
= 0.0161 0.0189 = 0
(0.0161)(1 0.0161)
300
= 0.0161 + 0.0189 = 0.0379
r
(0.0161)(1 0.0161)
LCL = 0.0161 3
300
= 0.0161 0.0189 = 0
U CL = 0.0161 + 3
(0.0161)(1 0.0161)
500
= 0.0161 + 0.0189 = 0.0330
r
(0.0161)(1 0.0161)
LCL = 0.0161 3
500
= 0.0161 0.0189 = 0
U CL = 0.0161 + 3
Hence, it can be seen that, all the values are in between UCL and LCL. Also, it can been seen that
the proportion has been changed from 0.0172(before revision) to 0.0161(after revision).
To achieve target value of 0 error, the feasible limit =
0.01616
= 2.216
0.0073
.
It will be impossible to achieve an error free environment in every sample, but for a few samples error
free environment can be achieved.
8-24. A health care facility is interested in monitoring the primary C-section rate. Monthly
data on the number of primary C-sections collected over the last two and a half years is
shown in Table 8-20.
a. Is the process in control?
Centerline:
p =
1847
= 0.153
12032
With the above data, we get the following p-chart with UCL as 0.2058 and LCL as 0.1004. It can
be seen from the chart that the process is in control.
b. There is pressure to make these data public. Can we conclude that the C-section
rates had shifted to a higher level in the last six months relative to the previous six months?
From the graph, we can see that the C-sections rates had shifted to a higher level in the last six
months.
(Last few observations in the graph)
c. What is your prediction on the C-section rate if no changes are made in current
obstetrics practices?
If no changes are made, the C-section rate is expected to be
0.1531 100 = 15.31%
U CL CL
0.2058 0.1531
=
= 0.0176
3
3
Calculating the standard deviation of the benchmark from the Center Line:
Z =
0.1531 0.10
= 3.017
0.0176
Hence, we can conclude that it is not feasible to achieve the benchmark value of 10 % if no changes are
made to the current system.
8-26. The number of processing errors per 100 purchase orders is monitored by a company with the objective of eliminating such errors totally. Table 8-21 shows samples that
were selected randomly from all purchase orders. The company is in the process of testing the effects of a new purchase order form that it has designed. The last five samples
were made using the new form. Construct a control chart that the company can use for
monitoring the quality characteristic selected. What is the effect of the newly designed
purchase order form? Is the company capable of achieving the desired goal?
It can be observed from the c-chart that sample 9 lies above the UCL.
By assuming special causes and removing the 9t h sample, we have:
90
c
=
= 3.75
g
24
3.75 = 1.936
The revised Center Line, that is, 3.75 is removed from the goal value if 0 in the standard deviation
units as:
(3.750)
Z =
= 1.936
1.936
Therefore, it is quite possible for the company to achieve 0 errors on a consistent basis, but not
probable. However, if more samples are taken into consideration, we would be able to estimate about
the companys chances of meeting the goal.
domly selects 300 m2 and counts the number of blemishes. The total number of blemishes
for 30 samples is 80. Construct the centerline and control limits for an appropriate chart.
Is it reasonable for the contractor to set a goal of an average of 0.5 blemish per 100 m2 ?
Given:
Number of blemishes (n) = 30
Sample size (s) = 80
80
s
=
= 2.667
center line =
n
30
(0.889)2
1.8778
=
= 0.9389
2!
2
Hence, From the Appendix, we see that about 6.11 % of the time, the process could go out of the
control, although a small value, the process cannot be deemed totally capable.
8-29. The number of imperfections in bond paper produced by a paper mill is observed
over a period of several days. Table 8-23 shows the area inspected and the number of
imperfections for 25 samples. Construct a control chart for the number of imperfections
per square meter. Revise the limits if necessary, assuming special causes for the out-ofcontrol points.
Given in the question,
Number of samples = 25.
From the table, sum of all imperfections, ci = 189
Sum of the area inspected,ni = 4850
With the given data, the center line of the u-chart, u
=
189
4850
= 0.039
r
u
0.039
= 0.039 3
n
n
It can be seen from the u-chart that the samples 6 and 19 are out of control and are above the UCL.
Assuming special causes and revising the control limits by removing the 6th and the 19th observation,
we have:
The revised center line of the u-chart, u
=
164
4600
= 0.0356
0.0356
baru 3
= 0.0356 3
= (0, 0.0922)
n
100
The revised control limits for sub-group size 150,
r
r
u
0.0356
baru 3
= 0.0356 3
= (0, 0.0818)
n
150
8
0.0356
baru 3
= 0.0356 3
= (0, 0.0756)
n
200
0.0356
baru 3
= 0.0356 3
= (0, 0.0714)
n
250
0.0356
baru 3
= 0.0356 3
= (0, 0.0683)
n
300
8-37 Refer to Exercise 8-36. Set up 2s control limits. What is the probability of detecting a change in the process average number of dietary errors per 100 trays to 8 on the first
sample drawn after the change? Explain under what conditions you would prefer to have
these 2s control limits over the traditional 3s limits.
From the data in the questions 36 and 27, we have
Sample size = 100
The center line is calculated from the data in the Table 8-22 as
181
CL =
= 7.24
25
It can be observed from the c-chart constructed that sample 7 lies above the UCL.
Assuming special causes are identified and removing the 7th sample.
The Center Line,
CL = c =
164
= 6.875
24
Probability of detecting a change in the process average number of dietary errors per
100 trays to 8 on the first sample drawn after the change:
The probability of a Type II error when a process average changes is
= P [X < 12.119 | c] P [X = 1.63 | c] = P [X < 12 | c] P [X < 1 | c]
Given, c = 8
= P [X < 12 | c = 8] P [X < 1 | c = 8] = 0.936 0.03(F romAppendixA 2)
hence, = 0.933 Therefore, the probability of detecting a change in the process average number of
dietary errors per 100 trays to 8 on the first sample drawn after the change is 1 0.933 = 0.067.
Conditions under which the 2 control limits are preferred over the traditional 3 control limits:
We would prefer the 2 control limits for detecting small changes in the process average as quickly
as possible neglecting the increase in Type I error it would cause.
8-43. Consider Exercise 8-38. However, now assume that the interval between complications follows an exponential distribution. Construct an appropriate control chart and
comment on the process assuming a type I error rate of 0.005.
From the data of this question and the exercise 8-38, we have,
Sample size = 25
Complication rate = 0.1%
Interval between complications follows an exponential distribution Type I error rate of 0.005
Using an exponential distribution with parameter = 0.001 and = 0.005, we get
10
Center line,
CL =
0.6931
0.6931
=
= 693.1
0.001
1
0.005
ln(1 ) =
ln(1
) = 1000 ln(0.9975) = 2.50
2
0.001
2
U CL =
1
0.005
ln( ) =
ln(
) = 1000 ln(0.0025) = 5991.46
2
0.001
2
Hence, with the above data, we get the following with UCL as 5991.46 and LCL as 2.50.
Comment: It can be seen from the chart that the process is in control and all the observations are within control limits.
11
CHAPTER 9
9-13. A pharmaceutical company producing vitamin capsules desires a proportion of calcium content between 40 and 55 ppm. A random sample of 20 capsules chosen from the
output yields a sample mean calcium content of 44 ppm with a standard deviation of 3
ppm. Find the natural tolerance limits of the process. If the process is in control at the
present values of its parameters, what proportion of the output will be nonconforming,
assuming a normal distribution of the characteristic?
Given,
Proportion of Calcium content = 40 to 55 ppm
Random sample size = 20
Mean, = 44 ppm
Standard deviation, = 3 ppm
Tolerance limits:
U pper N atural T olerance Limit = + 3 = 44 + 3(3) = 53
Lower N atural T olerance Limit = + 3 = 44 3(3) = 35
The standard normal values at UNTL and LNTL are:
(4044)
= 1.33
3
(5544)
=
= 3.67
3
Z1 =
Z2
Using the standard normal tables, the proportion below the lower specification limits is 0.0918, while
the specification above the upper specification limit is 0.0000.
9-15. The emergency service unit in a hospital has a goal of 3.5 minutes for the waiting
time of patients before being treated. A random sample of 20 patients is chosen and the
sample average waiting time is found to be 2.3 minutes with a sample standard deviation
of 0.5 minutes. Find an appropriate process capability index. Comment on the ability of
the emergency service unit to meet the desirable goal, assuming normality. What are some
possible actions to consider?
Given,
Upper specification limit = 3.5 mins
Average waiting time = = 2.3 mins
Standard deviation = = 0.5 mins
12
(U SL )
(3.5 2.3)
=
= 0.8
(3 )
(3 0.5)
U ppercapabilityindex =
As CPU is less than 1, the emergency unit will not fully meet the desirable goal.
Assuming a normal distribution of waiting time, we can determine the proportion of patients who
will have to wait longer than the specified goal as,
Z =
(3.5 2.3)
= 2.4
0.5
Using the standard normal tables, the proportion of patients who will have to wait longer than 3.5
minutes is 0.0082.
To prevent this, the following steps can be taken:
Expanding the facilities provision at the unit.
Increasing the quantity of experienced staff.
Decreasing the time between each treatment.
9-17. The diameter of a forged part has specifications of 120 5mm. A sample of 25 parts
chosen from the process gives a sample mean of 122 mm with a sample standard deviation
of 2 mm. (a) Find the Cpk index for the process, and comment on its value. What is the
proportion of nonconforming parts assuming normality? If the target value is 120 mm, find
the Cpm and Cpmk indices and comment on their values. If the process mean is to be set
at the target value, how much of a reduction would occur in the proportion nonconforming?
Given, Specifications of the diameter of a forged part = 120 5 mm
Sample size = 25
Sample mean, = 122 mm
Sample standard deviation, s = 2 mm
We know,
Cpk = min
Cpk = min
U SL LSL
,
3
3s
As Cpk is less than 1, an undesirable solution exists. The standard normal values at the specification
limits are:
Z1 =
Z2 =
115112
= 3.5
2
125122
= 1.5
2
13
From the standard normal tables, the proportion below the LSL is 0.0000, while the proportion above
the USL is 0.0668.
Since the target value is 120, we have
2 = 2 + ( )2 = 22 + (122 120)2 = 8
Cpm =
8 = 2.8284
U SL LSL
125 115
=
6(2.8284) = 0.589
6
/
Comment: The process mean is off from the target value by 1 standard deviation.
Cpmk =
Comment: Cpmk, being the smallest indicates that the process mean is close to one of the
specification limits in standardized units and also is deviating from the target value.
(b) Parts with a diameter below the lower specification limit cost $ 1.00 per part to
be used in another assembly; those with a diameter above the upper specification limit
cost $ 0.50 per part for rework. If the daily production rate is 30,000 parts, what is the
daily total cost of non conformance if the process is maintained at its current setting? If
the process mean is set at the target value, what is the daily total cost of non conformance?
Given,
Cost of parts with a diameter lower than the specification limits = $ 1.00
Cost of parts with a diameter higher than the specification limits = $ 0.50
Number of parts produced daily = 30,000
Process mean is set at a target value of 120 which means the standard normal value would be 2.5
at the specification limits. Therefore, total proportion non-confirming would be 2(0.0062) = 0.0124.
As a result, reduction in the proportion non-confirming = 0.0668 0.0124 = 0.0544
With the process mean at 122, the daily costs are calculated as follows:
Costs of parts below LSL = 0.0000 30,000 1.00 = $ 0.00
Costs of parts above USL = 0.0668 30,000 0.50 = $ 1002.00
14
9-19. A major automobile company is interested in reducing the time that customers
have to wait while having their car serviced with one of the dealers. They select four
customers randomly each day and find the total time that each of those customers has to
wait (in minutes) while having his or her car serviced. Next, from these four observations,
the sample average and range are found. This process is repeated for 25 days.
Given,
25
X
Xi = 1000
i=1
25
X
Ri = 250
i=1
X
Center Line on the X-chart,
=
Center Line on the R-chart, R
1000
25
250
25
= 40
= 10
+ A2R
= 40 + (0.729 10) = 47.29
U pper control limit = U CL = X
A2R
= 40 (0.729 10) = 32.71
Limit control limit = LCL = X
For R-chart,
b.Assuming that the process is in control and that a desirable value on the upper bound
of the waiting time is 50 minutes, calculate a process capability index, and comment on its
15
value.
Upper bound on the waiting time = 50 minutes
Lower specification limit = = centerline = 40 minutes
10
An estimate of the process standard deviation = = 2.059
= 4.857
The upper capability can be calculated as:
CP U =
50 40
(U SL )
=
= 0.686
3
3(4.857
(U SL )
50 40
=
= 2.058
4.857
Using the standard normal tables, the proportion of customers who will have to wait more than 50
minutes is 0.0197.
To reduce the waiting time of customers, the service manager hires some additional
mechanics, which reduces the average waiting time to 35 minutes. What proportion of the
customers will still have to wait more than 50 minutes if the variability in service times is
the same as before?
Given, new average waiting time = 35 minutes
Therefore, new standard normal value,
Z =
(U SL )
(50 35)
=
= 3.088
4.857
From standard normal tables, the proportion of customers who will have to wait more than 50 minutes after the hiring is 0.0010.
9-23 In Exercise 9-22, suppose that the specifications for the gap are 1.05 0.15 cm. An
assembly with a gap exceeding the upper specification limit is scrapped, whereas that with
a gap less than the lower specification limit can be reworked to increase the gap dimension.
The unit cost of rework is $ 0.15 and that for scrap is $ 0.40. If the daily production rate
is 2000, calculate the daily total cost of scrap and rework. How can this cost be reduced?
16
Given question,
Specifications for the gap = 1.05 0.15 cm
Unit cost of rework = $ 0.15
Unit cost of scrap = $ 0.40
Daily production rate = 2000
From the previous question,
The standard deviation of the components are calculated as follows:
F orComponentA, 1 =
(U SL LSL)
10.5 9.5
=
= 0.167
6
6
ComponentB : 2 =
4.2 3.8
(U SL LSL)
=
= 0.067
6
6
ComponentC : 3 =
(U SL LSL)
5.1 4.9
=
= 0.03
6
6
0.03347 = 0.183
Z1 =
Z2 =
(LSL )
0.9 1
=
= 0.55
0.183
(U SL )
1.2 1
=
= 1.09
0.183
From the standard normal tables, the proportion below the LSL is 0.2912, while the proportion above
the USL is 0.1379.
Therefore, daily cost of rework = 2000 0.15 0.2912 = $ 87.36
Daily cost of scrap = 2000 0.40 0.1379 = $110.32
Daily total cost of rework and scrap = $ 87.36 + $ 110.32 = $ 197.68
To reduce the total cost:
17
The mean dimensions of B and C can be decreased so as to bring the mean value of gap to the
target value of 1.05.
The variability of the operations can be reduced and the measures for the same should be thought
over.
18