Lecture Notes Unit IV EHU-601
Objective of quality control is to develop a scheme for sampling a process,
making a quality measurement of interest on sample items, and then making a
decision as to whether or not the process is in the stable state, or in control.
If the sample data suggests that the process is out of control, a cause is for
the abnormality is sought.
A common method for making these decisions involves the use of control
charts.
These are very important and widely used techniques in industry, and
everyone in the industry, even if not directly related to quality control, should
be aware of these.
Statistical process control
Methodology for monitoring a process to identify special causes of variation
and signaling the need to take corrective action.
When special causes are present, the system said to be statistically out of
control.
If the variations are due to common causes alone, the process is said to be in
statistical control.
SPC relies heavily on control charts.
Quality control measurements
Attributes A performance characteristics that is either present or absent in
the product or service under consideration.
Examples: Order is either complete or incomplete; an invoice can have one,
two, or more errors.
Attributes data are discrete and tell whether the characteristics conforms to
specifications.
Attributes measurements typically represented as proportions or rates. e.g.
rate of errors per opportunity.
Typically measured by Go-No Go gauges.
Variable Continuous data that is concerned with degree of conformance to
specifications.
Lecture Notes Unit IV EHU-601
Generally expressed with statistical measures such as averages and standard
deviations.
Sophisticated instruments (caliper) used.
In statistical sense, attributes inspection less efficient than variable inspection.
Attribute data requires larger sample than variable inspection to obtain same
amount of statistical information.
Most quality characteristics in service industry are attributes
Where to Use Control Charts
Process has a tendency to go out of control
Process is particularly harmful and costly if it goes out of control
Examples
at the beginning of a process because it is a waste of time and money to
begin production process with bad supplies
before a costly or irreversible point, after which product is difficult to
rework or correct
before and after assembly or painting operations that might cover
defects
before the outgoing final product or service is delivered
WHAT IS STATISTICAL QUALITY CONTROL?
Statistical quality control (SQC) is the term used to describe the set of
statistical tools used by quality professionals. Statistical quality control can be
divided into three broad categories:
1. Descriptive statistics are used to describe quality characteristics and
relationships. Included are statistics such as the mean, standard deviation, the
range, and a measure of the distribution of data.
2. Statistical process control (SPC) involves inspecting a random sample of
the output from a process and deciding whether the process is producing
Lecture Notes Unit IV EHU-601
products with characteristics that fall within a predetermined range. SPC
answers the question of whether the process is functioning properly or not.
3. Acceptance sampling is the process of randomly inspecting a sample of
goods and deciding whether to accept the entire lot based on the results.
Acceptance sampling determines whether a batch of goods should be accepted
or rejected.
SOURCES OF VARIATION: COMMON AND ASSIGNABLE CAUSES
No two products are exactly alike
because of slight differences in materials, workers, machines, tools, and other
factors. These are called common, or random, causes of variation. Common
causes of variation are based on random causes that we cannot identify. These types
of variation are unavoidable and are due to slight differences in processing.
The second type of variation that can be observed involves variations where the
causes can be precisely identified and eliminated. These are called assignable
causes of variation. Examples of this type of variation are poor quality in raw
materials,
Lecture Notes Unit IV EHU-601
Lecture Notes Unit IV EHU-601
Lecture Notes Unit IV EHU-601
Lecture Notes Unit IV EHU-601
Lecture Notes Unit IV EHU-601
Lecture Notes Unit IV EHU-601
DESCRIPTIVE STATISTICS
The Mean
where
the mean
= observation i, i= 1, . . . , n
n = number of observations
The Range and Standard Deviation
range, which is the difference between the largest and smallest observations.
Standard deviation A statistic that measures the amount of data dispersion
around the mean.
standard deviation of a sample
the mean
= observation i, i= 1, . . . , n
n = number of observations in the sample
Normal distributions with varying standard deviations Differences
between symmetric and
Lecture Notes Unit IV EHU-601
skewed distributions
STATISTICAL PROCESS CONTROL METHODS
Statistical process control methods extend the use of descriptive statistics to monitor
the quality of the product and process.
Using statistical process control we want to determine the amount of variation that
is common or normal. That is, we want to make sure the process is in a state of
control.
Control chart
A graph that shows whether a sample of data falls within the common or normal
range of variation.
Out of control
The situation in which a plot of data falls outside preset control limits.
Developing Control Charts
The x axis represents samples (#1, #2, #3, etc.) taken from the process over time.
The y axis represents the quality characteristic that is being monitored (ounces of
liquid). The center line (CL) of the control chart is the mean, or average, of the
quality characteristic that is being measured.
The upper control limit (UCL) is the maximum acceptable variation from the mean
for a process that is in a
state of control.
Similarly, the lower control limit (LCL) is the minimum acceptable variation from
the mean for a process that is in a state of control.
Usually Set : Mean 3 standard deviations from the mean
Control limits will capture 99.74 percent of the normal variation
Lecture Notes Unit IV EHU-601
Quality control chart for Cocoa Fizz
Types of Control Charts
The different characteristics that can be measured by control charts can be divided
into
two groups: variables and attributes.
A control chart for variables is used to monitor characteristics that can be
measured and have a continuum of values, such as height,weight, or volume.
A soft drink bottling operation is an example of a variable measure,
since the amount of liquid in the bottles is measured and can take on a number
of different values.
Other examples are the weight of a bag of sugar, the temperature of a baking oven,
or the diameter of plastic tubing. 3 2
Variable
A product characteristic that can be measured and has a continuum of values (e.g.,
height, weight, or volume).
Attribute
A product characteristic that has a discrete value and can be counted.
Lecture Notes Unit IV EHU-601
A control chart for attributes, on the other hand, is used to monitor characteristics
that have discrete values and can be counted.
An attribute
requires only a single decision, such as yes or no, good or bad, acceptable or
unacceptable e.g., the apple is good or rotten
CONTROL CHARTS FOR VARIABLES
Control charts for variables monitor characteristics that can be measured and have a
continuous scale, such as height, weight, volume, or width.
When observed values go outside the control limits, the process is assumed not to be
in control.
Mean (x-Bar) Charts
Lecture Notes Unit IV EHU-601
x-bar chart
A control chart used to monitor changes in the mean value of a process.
A mean control chart is often referred to as an x-bar chart. It is used to monitor
changes in the mean of a process. To construct a mean chart we first need to
construct
the center line of the chart. To do this we take multiple samples and compute their
means. Usually these samples are small, with about four or five observations. Each
sample has its own mean, . The center line of the chart is then computed as the mean
of all _ sample means, where _ is the number of samples:
_
_
To construct the upper and lower control limits of the chart, we use the following
formulas:
Upper control limit (UCL) _
Lower control limit (LCL) _
where _ the average of the sample means
z _ standard normal variable (2 for 95.44% confidence, 3 for 99.74%
confidence)
_ standard deviation of the distribution of sample means, computed as
_ _ population (process) standard deviation
n _ sample size (number of observations per sample)
Example 6.1 shows the construction of a mean (x-bar) chart.
Lecture Notes Unit IV EHU-601
Lecture Notes Unit IV EHU-601
Lecture Notes Unit IV EHU-601
Another way to construct the control limits is to use the sample range as an estimate
of the variability of the process.
Constructing a Mean (x-Bar) Chart from the Sample Range
Lecture Notes Unit IV EHU-601
Lecture Notes Unit IV EHU-601
Lecture Notes Unit IV EHU-601
Range (R) Charts
Range (R) charts are another type of control chart for variables. Whereas x-bar
charts measure shift in the central tendency of the process, range charts monitor the
dispersion or variability of the process. The method for developing and using R-
charts is the same as that for x-bar charts. The center line of the control chart is the
average range, and the upper and lower control limits are computed as follows:
Lecture Notes Unit IV EHU-601
Using Mean and Range Charts Together
Lecture Notes Unit IV EHU-601
You can see that mean and range charts are used to monitor different variables.
The mean or x-bar chart measures the central tendency of the process, whereas the
range chart measures the dispersion or variance of the process. Since both variables
are important, it makes sense to monitor a process using both mean and
Lecture Notes Unit IV EHU-601
The p-Chart
The control charts we looked at so far were applicable to quality
measurements that possessed continuous probability distribution.
This sampling is commonly referred to as sampling by variables.
In many cases we merely want to assess whether or not a certain
item is defective.
We can observe a number of defective items from a particular
sample of series of samples. This is referred as sampling by
attributes.
Suppose a series of k independent samples, each of size n, is
selected for a particular process.
Let p denote the proportion of defective items in the population
(total production for a certain time period) for a process in control.
Let Xi denote the number of defective items in the ith sample.
Then Xi is a binomial variable, assuming random sampling from a
large lot, with mean E(Xi) = np, and Var(Xi) = np(1-p).
We typically consider the fraction of defective items in a sample
(Xi/n) rather than the observed number of defectives.
( ) ( )
.
size sample total
observed defective of number total
: of estimator unbiased The
samples. from data the using
limits, control the estimate must we unknown, is Since
. / ) 1 ( / and / since,
) 1 (
3
interval in the or mean, their of
deviations standard ree within th lie should fractions sample All
1
nk
X
p
p
k
p
n p p n X Var p n X E
n
p p
p
k
i
i
i i
=
= =
= =
Thus the estimated control limits are given by:
Lecture Notes Unit IV EHU-601
The c-Chart
In many quality control problems the particular items being subjected to
inspection may have more than one defect.
We may wish to count # of defects instead of merely classifying at item as to
whether or not it is defective.
If Ci denotes the # of defects observed in the ith inspected item, we can safely
assume that Ci has a Poisson distribution.
Let this Poisson distribution have a mean of for a process in control.
For us to be 99.74% sure, almost all the Cis should fall within three standard
deviations of the mean if the process is in control.
( ) ( )
. 3
: become then limits control estimated The
.
1
: is of estimator unbiased then the inspected, are items If
. 3
: are limits control the , and Since
1
C C
C
k
C
k
C Var C E
k
i
i
i i
= =
Acceptance sampling by attributes
Inspections in which a unit of product is classified simply as defective or non-
defective is called inspection by attributes.
Lots of items, either raw materials or finished products, are sold by a producer
to consumer with some guarantee for quality.
We determine quality, here, by the proportion p of defective items in the lot.
.
) 1 (
3
n
p p
p
Lecture Notes Unit IV EHU-601
To check this characteristics of the lot, the consumer will sample some of the
items, test them, and observe the number of defectives.
Generally, some defectives are allowed because defective-free lots may be too
expensive for the consumer to purchase.
But if the number of defectives is too large, the consumer will reject the lot
and return it to the producer.
Sampling is generally used because the cost of inspecting the entire lot may be
too high or the inspection process may be destructive.
Before sampling inspection takes place for a lot, the consumer must have in
mind a proportion p0 of defectives that will be acceptable.
Thus, if the true proportion of defectives p is no greater than p0, the consumer
wants to accept the lot, and reject otherwise.
This maximum proportion of defectives satisfactory to the consumer is called
the Acceptable Quality Level (AQL).
So, we have a hypothesis-testing problem.
We are testing null hypothesis of the true defectives proportions being less
than or equal to the acceptable proportion; against an alternative hypothesis
that the true proportion is actually greater than the acceptable one.
A random sample of n items is selected from the lot, and the number of
defectives Y is observed.
The decision concerning rejection or non-rejection of Ho (and consequently
rejecting or accepting the lot) will be based on the observed value of Y.
The probability of Type I error () in this problem is the probability that the
lot is rejected by the consumer when, in fact, the proportion of defectives is
satisfactory to the consumer.
This is referred to as the producers risk.
The value of calculated at p = p0 is the upper limit to the proportion of good
lots rejected by the sampling plan being considered.
. : against :
0 0
p p H p p H
a o
> s
Lecture Notes Unit IV EHU-601
The probability of Type II error calculated for some proportion of defective
p1, where p1 > p0, represents the probability that an unsatisfactory lot will be
accepted by the consumer.
This is referred to as the consumers risk.
The value of calculated at p1 = p is the upper limit to the proportion of bad
lots accepted by the sampling plan being considered, for all
.
Lecture Notes Unit IV EHU-601
Lecture Notes Unit IV EHU-601
ACCEPTANCE SAMPLING
Acceptance sampling, the third branch of statistical quality control, refers to the
process of randomly inspecting a certain number of items from a lot or batch in
order to decide whether to accept or reject the entire batch. What makes acceptance
sampling different from statistical process control is that acceptance sampling is
performed either before or after the process, rather than during the process.
Acceptance sampling before the process involves sampling materials received from a
supplier, such as randomly inspecting crates of fruit that will be used in a restaurant,
boxes of glass dishes that will be sold in a department store, or metal castings that
will be used in a machine shop. Sampling after the process involves sampling
finished items that are to be shipped either to a customer or to a distribution center.
Examples include randomly testing a certain number of computers from a batch to
make sure they meet operational requirements, and randomly inspecting
snowboards to make sure that they are not defective.
You may be wondering why we would only inspect some items in the lot and not the
entire lot. Acceptance sampling is used when inspecting every item is not physically
possible or would be overly expensive, or when inspecting a large number of items
would lead to errors due to worker fatigue. This last concern is especially important
when a large number of items are processed in a short period of time. Another
example of when acceptance sampling would be used is in destructive testing, such
as testing eggs for salmonella or vehicles for crash testing. Obviously, in these cases
it would not be helpful to test every item! However, 100 percent inspection does
make sense if the cost of inspecting an item is less than the cost of passing on a
defective item.
As you will see in this section, the goal of acceptance sampling is to determine the
criteria for acceptance or rejection based on the size of the lot, the size of the sample,
and the level of confidence we wish to attain. Acceptance sampling can be used for
both attribute and variable measures, though it is most commonly used for
attributes. In this section we will look at the different types of sampling plans and at
ways to evaluate how well sampling plans discriminate between good and bad lots.
Sampling Plans
A sampling plan is a plan for acceptance sampling that precisely specifies the
parameters
of the sampling process and the acceptance/rejection criteria. The variables to be
specified include the size of the lot (N), the size of the sample inspected from the lot
(n), the number of defects above which a lot is rejected (c), and the number of
samples that will be taken.
There are different types of sampling plans. Some call for single sampling, in which
a random sample is drawn from every lot. Each item in the sample is examined and
is labeled as either good or bad. Depending on the number of defects or bad
items found, the entire lot is either accepted or rejected. For example, a lot size of 50
cookies is evaluated for acceptance by randomly inspecting 10 cookies from the lot.
The cookies may be inspected to make sure they are not broken or burned. If 4 or
Lecture Notes Unit IV EHU-601
more of the 10 cookies inspected are bad, the entire lot is rejected. In this example,
the lot size N _ 50, the sample size n _ 10, and the maximum number of defects at
which a lot is accepted is c _ 4. These parameters define the acceptance
sampling plan.
Another type of acceptance sampling is called double sampling. This provides an
opportunity
to sample the lot a second time if the results of the first sample are inconclusive. In
double sampling we first sample a lot of goods according to preset criteria for
definite acceptance or rejection. However, if the results fall in the middle range, they
are considered inconclusive and a second sample is taken. For example, a water
treatment plant may sample the quality of the water ten times in random intervals
throughout the day. Criteria may be set for acceptable or unacceptable water quality,
such as .05 percent chlorine and .1 percent chlorine. However, a sample of water
containing between .05 percent and .1 percent chlorine is inconclusive and calls for a
second sample of water.
In addition to single and double-sampling plans, there are multiple sampling plans.
Multiple sampling plans are similar to double sampling plans except that criteria are
set for more than two samples. The decision as to which sampling plan to select has
a great deal to do with the cost involved in sampling, the time consumed by
sampling, and the cost of passing on a defective item. In general, if the cost of
collecting a sample is relatively high, single sampling is preferred. An extreme
example is collecting a biopsy from a hospital patient. Because the actual cost of
getting the sample is high, we want to get a large sample and sample only once. The
opposite is true when the cost of collecting the sample is low but the actual cost of
testing is high. This may be the case with a water treatment plant, where collecting
the water is inexpensive but the chemical analysis is costly. In this section we focus
primarily on single sampling plans.