3. Quality economics
• Montgomery (1985) indicates that there are
several reasons why the cost of quality should be
explicitly considered in an organization:
1) An increase in the cost of quality because of
increases in technology use.
2) Increasing sophistication of end users in their
consideration of lifecycle costs.
3) Internally, an increase of the use of quality cost
data to make quality management-related
decisions.
4. What are quality related costs?
• There is no universal consensus upon just what
constitutes quality costs.
• Traditionally, costs of quality (COQ) have been
considered as the cost of running a quality assurance
system (complete or in development), with perhaps
the inclusion of other costs, such as scrap and warranty
costs.
• Quality costs are incurred in the design,
implementation, maintenance and improvement of a
quality system.
• Cost of quality (COQ) crosses inter- and
intradepartmental boundaries, much as the process for
developing and producing the product or services does
in that organization.
5. What are quality related costs?
• No department or group is isolated and
therefore is not immune to the constraints
and opportunities of managing quality costs.
• Cost of quality is not confined to the internal
environment of the organization, as the
activities of, say, suppliers, etc., can affect the
outcome of costs related to quality.
6. Classification of quality costs
BS 4778-part 2 (British standard institution
standards) classifies quality costs into the
following:
1. Quality related costs: the expenditure incurred in
defect prevention and appraisal activities plus the
losses due to internal and external failure.
2. Prevention costs: the cost of any action taken to
investigate, prevent or reduce defects and failures.
Prevention costs can include the cost of planning,
setting up and maintaining the quality concern
system. They also include process design, product
and service design and employee training schemes.
7. Classification of quality costs
3) Appraisal costs: the cost of assessing the quality
achieved. Appraisal costs can include the cost of
inspecting, testing, etc. carried out during and on
completion of manufacture of product or service.
4) Failure costs-internal: the costs arising within the
manufacturing processes of the organization of the
failure to achieve specified quality. This can include
the cost of scrap, rework and re-inspection.
5) Failure costs-external: the costs arising from outside
the manufacturing organization of the failure to
achieve quality specified. The term can include the
costs of claims against warranty, replacement and
consequential losses of customs and goodwill.
8. Importance of quality costs to the quality oriented
organization
• According to Dale and Plunkett(1991) the quality
related costs commonly range from 5%-20% of
company annual sales turnover.
• Generally 95% of the total quality related costs are
expended on appraisal and failure elements.
• Failure costs must be regarded as avoidable and a
reduction in such costs in usually attributable to such
activities as eliminating causes of non conformance,
which may lead to a reduction in appraisal costs.
There is therefore a demonstrated need to balance
future failure costs with appraisal costs.
• Managerial requirements suggests that what can be
measured, can be managed.
9. Importance of quality costs to the quality
oriented organization
• Robertson states that for the average UK
organization the analysis of quality related costs are
65% failure costs, 30% appraisal costs and 5%
prevention costs, and further indicates that 4-20 %
costs are attributed to quality related costs of total
sales turnover.
• Garvin (1983) compares Japanese air-conditioning
manufacturers with their American counterparts,
focusing on warranty claims. Garvin indicates that
Japanese company warranty claims are 0.6% of sales
turnover, the best an American company could
report 1.8% with the worst being 5.2%.
10. Quality costs – why measure them?
• Measuring quality costs will provide a means to
quantify in management terms the effect that quality
related activities have on organizational
performance.
• It should influence employees and their attitudes
towards the quality system, TQM and related
continuous quality improvement schemes and
practices.
• Measurement of quality costs will focus attention
upon such areas as appraisal, prevention and failure
and therefore provide opportunities for cost
reductions.
11. Quality costs – why measure them?
• Performance across a wide range of quality related
activities may need to be measured and this will
provide a basis for internal cost comparisons
between departments, processes, services and
products.
• The measurement of quality costs can be clearly
seen as a major step towards quality control, quality
improvement and TQM.
12. Cost of quality versus cost of non quality
• The cost of quality can be divided into three main
aspects:
1. Failure costs
2. Appraisal costs
3. Prevention costs
from these only prevention can be regarded as a cost
of quality, whereas the other two are in essence
the cost of non-quality- inspection and rework of
errors; rather than the principle of working to
attain zero defects.
13. Cost of quality versus cost of non quality
• Finding shows that costs of non-quality in the service
industry may be very high.
• The intangible nature of the service products means
that the mistakes, which can never be undone, will
have to be regarded as a cost factor.
• Client, who have been dissatisfied may not only
never return, but may even influence other people
into not using the defective services that the
organization provide.
• What losses are incurred because of this process are
very difficult to substantiate, many organizations
have turned a blind eye to these losses.
14. Cost of quality versus cost of non quality
• However if they are truly committed to TQM, it will
be recognized that one aspect able to reduce this
cost of non quality is to ensure that the incidences
are reduced to an absolute minimum.
• This can be accomplished through proper training of
staff in the communications skills required to be able
to asses clients, need better and to bring customer
satisfaction to the optimum- whether internal and
external customers.
15. Hidden costs of quality
• Where errors in manufacturing produce waste, scrap or rework,
then the hidden cost of quality or rather non quality can be seen
as:
1. The extra material needed to be supplied to accommodate this
extra wastage.
2. The extra manpower costs of labour and perhaps overtime.
3. The opportunity cost of working on a part the second time round
or, in the case of a scrapped item, on a completely new part.
4. Possible delays in the ultimate shipment of the order.
5. Increase risk of machine breakdown.
6. Increase machine maintenance and repair costs.
7. Reduced production capacity resulting from the need to
overproduce in order to manufacture a given quantity of
production items.
16. Lifecycle costs
• Juran and Gryna (1993) discuss the impact of the
lifecycle costs theory.
• All products/markets/services have lifecycles.
• Fashion for example is a cycle.
• According to Juran and Gryna’s application is that the
cost of the product/service should not just be limited
to the cost at purchase. It should also include the
cost of maintenance and the running cost of the
product.
• Designing a product that lowers the overall lifecycle
cost may mean that the initial cost may be higher
than originally anticipated, but the consumer would
benefit in the long run.
17. Lifecycle costs
• An example is that of laser printer.; advertised
by Kyocera Electronics that the cost of their
laser printer – over a three year period- was
not the cheapest to purchase, but the
cheapest to run over that time in costs/page
printed.
18. The management of quality costs
• Survey conducted by Roche and Duncalfe and Dale
suggested that only about one-third of the companies
studied acutely collected quality cost data and that these
findings indicated that less than 40% of companies collect
and analyze quality costs data in systematic manner.
• The costs most measured were suggested to be those for
cost of scrap, rework and warranty claims.
• Although the importance of not placing complete reliance
upon the data from quality costing as a means of
improving quality and of reducing costs, quality cost data
should be used as some basis for the quantification of
quality related activities, but not solely to be used as a
weapon by top management to cut costs.
19. The management of quality costs
• Throwing vital resources into appraisal rather than
prevention is not good practice, as many
organizations have found to their folly.
• It is just that point that has made Japanese
manufacturers more effective than their American
and European counterparts during the 60’s and 70’s.
• American and Europeans funded appraisal rather
than prevention; they were essentially targeting the
symptom rather than the Japanese approach of
targeting the core problem and developing an
effective solution to it.
21. Statistical tools in quality
• Statistics is the collection, organization,
analysis, interpretation, and presentation of
data. The body of knowledge of statistical
methods is an essential tool of the modern
approach to quality. Without it drawing
conclusions about data becomes lucky at best
and disastrous in some cases.
22. Concept of variation
• The concept of variation states that no two items will
be perfectly identical. Variation is a fact of nature
and a fact of industrial life. For example even
identical twins vary slightly in height and weight at
birth.
• The cans of tomato soup vary slightly from can to
can; the time required to assign a seat at an airline
check-in counter varies from passenger to passenger.
To disregard the existence of variation (or to
rationalize falsely that it is small) can lead to
incorrect decisions on major problems. Statistics
helps to analyze data properly and draw conclusions,
taking into account the existence variation.
23. Concept of variation
• Data summarization can take several forms:
tabular, graphical, and numerical. Sometimes
one form will provide a useful, complete
summarization. In other cases, two or even
three forms are needed for complete clarity.
24. Tabular summarization of data: Frequency
distribution
• A frequency distribution is a tabulation of data arranged
according to size. The raw data of the electrical resistance
of 100 coils are given in the table.
3.37 3.34 3.38 3.32 3.33 3.28 3.34 3.31 3.33 3.34
3.29 3.36 3.30 3.31 3.33 3.24 3.34 3.36 3.39 3.34
3.35 3.36 3.30 3.32 3.33 3.25 3.35 3.34 3.32 3.38
3.32 3.37 3.34 3.38 3.36 3.27 3.36 3.31 3.33 3.30
3.35 3.33 3.38 3.37 3.44 3.22 3.36 3.32 3.29 3.35
3.38 3.39 3.34 3.32 3.30 3.29 3.36 3.40 3.32 3.33
3.29 3.41 3.27 3.36 3.41 3.37 3.36 3.37 3.33 3.36
3.31 3.33 3.35 3.34 3.34 3.34 3.31 3.36 3.37 3.35
3.40 3.35 3.37 3.35 3.32 3.36 3.38 3.35 3.31 3.334
3.35 3.36 3.39 3.31 3.31 3.30 3.35 3.33 3.35 3.31
25. Tabular summarization of data: Frequency
distribution
Resistance Frequency Cumulative
frequency
3.415-3.445 1 1
3.385-3.415 8 9
3.355-3.385 27 36
3.325-3.355 36 72
3.295-3.325 23 95
3.265-3.295 5 100
total 100
26. Graphical summarization of data: the histogram
• A histogram is a vertical bar chart of a frequency
distribution. Figure shows the histogram for the electrical
resistance data.
• Note that as in the frequency distribution, the histogram
highlights the center and amount of variation in the sample
of data. The simplicity of construction and interpretation of
the histogram makes it an effective tool in the elementary
analysis of data.
• Graphical methods are essential to effective data analysis
and clear presentation of results.
• The vividness of a picture when compared to the cold logic
of numbers has practical benefits, e.g. identifying subtle
relationships and presenting results in clear form.
Experience dictates that the first step in data analysis is:
Plot the data.
28. Quantitative methods of summarizing data:
Numerical indices
• Data can also be summarized by computing 1) a
measure of central tendency to indicate where most
of the data are centered and 2) the measure of
dispersion to indicate the amount of scatter in the
data, often these two measures provide an adequate
summary.
• The key measure of the central tendency is the
arithmetic mean, or average. The definition of the
average is X= ∑ x
n
29. Quantitative methods of summarizing data:
Numerical indices
• Another measure of central tendency is the median-
the median value when the data are arranged
according to size. The median is useful for reducing
the effects of extreme values.
• Two measures of dispersion are commonly
calculated. When the amount of data is small (ten or
fewer observation). The range is useful. The range is
the difference between the maximum value and the
minimum value in the data. As the range is based on
only two values, it is not as useful when the number
of observation is large.
30. Quantitative methods of summarizing data:
Numerical indices
• In general the standard deviation is the most useful
measure of dispersion. Like the mean, definition of the
standard deviation is a formula:
s= √ ∑ (x- x)2
n-1
• A problem that sometimes arises in the summarization
of data is that one or more extreme values are far from
the rest of the data. A simple but not necessarily correct
solution is available. Drop such values. The reasoning is
that a measurement error or some other unknown
factor makes the values unrepresentative.
31. Probability distribution: General
• A distinction is made between a sample and a population.
A sample is a limited number of items taken from a larger
source. While a population is a large source of items from
which the sample is taken.
• Measurements are made on the items. Many problems are
solved by taking the measurement results from a sample
and based on these results, making predictions about the
defined population containing the sample.
• It is usually assumed that the sample is a random one i.e.
each possible sample of n items has an equal chance of
being selected (or the items are selected systematically
from material that is itself random due to mixing during
process)
32. Probability distribution: General
• A probability distribution function is a
mathematical formula that relates the values
of the characteristic with their probability of
occurrence in the population.
• The collection of these probabilities is called a
probability distribution. Some distribution and
their functions are summarized as:
33. Probability distribution: General
• Normal distribution: applicable when there is a
concentration of observations about the average and
it is equally likely that observations will occur above
and below the average. Variations in observations is
usually the result of many small causes.
• Exponential distribution: applicable when it is likely
that more observations will occur below the average
than above.
• Weibull distribution: applicable in describing a wide
variety of patterns in variation, including departures
from the normal and exponential.
34. Probability distribution: General
• Poisson distribution: same as binomial but
particularly applicable when there are many
opportunities for occurrence of an event, but a low
probability less than 0.10 on each trial.
• Binomial distribution: applicable in defining the
probability of r occurrences in n trials of an event
which has a constant probability of occurrence on
each independent trial.
35. Probability distribution: General
• Distribution are two types:
1. Continuous (for variable data): when the
characteristics being measured can take on any
value ( subject to the fineness of the measuring
process); its probability distribution is called a
continuous probability distribution. For example ,
the probability distribution for the resistance data
is an example of a continuous probability
distribution because the resistance could have any
value, limited only by the fineness of the
measuring instrument.
36. Probability distribution: General
• Discrete (for attribute data): when the
characteristics being measured can take on
only certain specific values (e.g. integers 0, 1,
2, 3, 4, 5 etc.), its probability distribution is
called a discrete probability distribution. The
common discrete distributions are the Poisson
and binomial.
37. Basic theorems of probability
• Probability is expressed as a number which lies
between 1.0 ( certainly that an event will occur) and
0.0 (impossibility of occurrence).
• A convenient definition of probability is one based
on a frequency interpretation: if an event A can
occur in s cases out of a total of n possible and
equally probable cases, the probability that the event
will occur is
P (A)= s = number of successful cases
n total number of possible cases
38. Basic theorems of probability
• Example: a lot consists of 100 parts. A single
part is selected at random, and thus each of
the 100 parts has an equal chance of being
selected. Suppose that a lot contains a total of
8 defective. Then the probability of drawing a
single part that is defective is then 8/100 or
0.08.
39. Basic theorems of probability
• The following theorems are useful in solving
problems:
Theorem1: If P(A) is the probability that an event A
will occur, then the probability that A will not occur
is 1-P(A)
Theorem2:If A and B are two events, then the
probability that either A or B will occur is
P(A or B)= P(A) + P(B) – P(A and B)
In case A and B are mutually exclusive then the
P(A or B)= P(A) + P(B)
40. Basic theorems of probability
• Theorem3: if A and B are two events then the
probability that events A and B occur together is:
P(A and B)= P(A) x P(B|A)
Where P(B|A) means probability that B will occur
assuming A has already occurred.
A special case in this theorem occurs when the two
events are independent, i.e. when the occurrence of
one event has no influence on the probability of the
other event. If A and B are independent, then the
probability of both A and B occurring is
P(A and B)= P(A) x P(B)