Diverse Applications of Statistical Analysis in

Business Fields

Maged Younan
MBA, Six Sigma MBB
 Table of contents

Abstract.............................................................................................................................. 3

Part I - Six Sigma in Brief ................................................................................................ 4

History of Six Sigma....................................................................................................... 4

The meaning of the Word “Six Sigma” .......................................................................... 5
Six Sigma Philosophy ..................................................................................................... 6
The Six Sigma DMAIC .................................................................................................. 7
• Define Phase ................................................................................................... 7
• Measure Phase ................................................................................................ 7
• Analyze Phase................................................................................................. 8
• Improve Phase................................................................................................. 8
• Control Phase .................................................................................................. 8
Roles within six sigma organizations ............................................................................. 9
Critical Success factors of Six Sigma Implementation................................................. 11

Part II - Applications of Six Sigma Methodology and Tools ...................................... 14

Use of six sigma and statistics in Manufacturing: ........................................................ 14

Use of six sigma and statistics in Service Industries .................................................... 20
Statistics and Marketing:............................................................................................... 22
Use of six sigma and statistics in Finance .................................................................... 24
Use of six sigma and statistics in HR............................................................................ 26
Statistics and Legal Judgments ..................................................................................... 28

Recommendations and Conclusion ............................................................................... 30

Abstract

In any business various decisions have to be made each day. Availability of reliable data

facilitates the decision making process. However, when statistical and graphical tools are

used to present and analyze available data, the decision becomes clearer, faster and more

accurate. In this paper we are discussing the Six Sigma methodology, its limitations and

its applications in the business field. Various examples of how six sigma tools can be

very useful will be presented covering the fields of Marketing, HR, Finance,

Manufacturing and Service. A case study will be also discussed in details showing how

the thinking methodology of six sigma was used to uncover the root cause of a problem

that one of the multinational telecommunications service providers was facing.

Part I - Six Sigma in Brief
“Six Sigma is a scientific, data driven approach to uncover the root cause of customer

concerns and drive out defects” This is how Ford Motor company defined six sigma on

presenting their annual report in 2002. In this year Ford enjoyed a total savings of 52

Million USD due to implementation of six sigma. The “six sigma” is a methodology that

organizations deploy to reach perfection. It relies on making several breakthrough

improvements through what is called: ”Six Sigma Projects”. Each “Project” follows a

certain road map for execution called the “DMAIC”. Throughout this road map a variety

of statistical and non statistical tools is used to understand and minimize the variation of a

critical process output. The steps of the “DMAIC” will be explained in a separate section.

Six sigma projects can tackle complex problems in all business areas including

production, finance, HR, sales, supply chain and others

History of Six Sigma

Six Sigma started in Motorola in the mid of the nineteen eighties. When it started it was

mainly used and limited to fixing production defects. Later, the concept was extended to

deal with problems in all business areas, not only those in the shop floor. Motorola did

not invent any of the statistical or non statistical tools that are usually used in a six sigma

project but rather collected these tools and made them available as a kit that can be used

by green and black belts in solving problems. What Motorola did invent however is the

roadmap that should be followed in tackling problems. Six sigma methodology was then

conveyed Allied Signal and General Electric. Jack Welch who was the CEO of General

Electric is the one responsible for the great reputation and fame that the six sigma
methodology had acquired. Six sigma is very popular now in the United States in

manufacturing as well as non manufacturing organizations including banks and hospitals.

Recently, Six Sigma started to be well known in some European countries; mainly in the

United Kingdom. Here in Egypt, around 100 business organizations are adopting and

implementing this new way of thinking.

The meaning of the Word “Six Sigma”

The “sigma level” of a process or a company is a measure of performance. The higher the

performance of a process is the higher the sigma level. At a six sigma quality level the

nearest customer specification limit is six standard deviations away from the process

mean. The figure below explains that at this level the probability of having a defect –

typically any unit outside of the specifications range – is very low.

Chart showing defects produced at 3 Sigma level compared to those at 6 Sigma level

Symbios – Consulting training material

A sigma level of six however is not the optimum for all processes. In applications where

people safety is involved as in aviation industry for example, six sigma quality is not

accepted.

Six Sigma Philosophy

The six sigma philosophy depends on the very basic equation of: Y = f (X). Meaning that

if we are trying to fix any process output “Y”, we should rather try to improve the process

inputs (Xs) and then the output would be automatically fixed.

Inputs Process Output(s)

During the six sigma project, data is collected about the outputs at various levels of the

input. When the data is analyzed, relationships can be established and even quantified

between the inputs and the outputs. Root cause and critical inputs can then be easily

identified and adjusted. As six sigma projects are usually complex in nature, therefore

they are usually expected to have high impact on the customer and/or business financials.

On simple problems when the root cause is already known, a six sigma project is not

actually needed but rather improvement and control efforts to solve the problem.
The Six Sigma DMAIC

The road map of any six sigma project consists of 5 main phases; Define, Measure,

Analyze, Improve and Control.

Define Phase

This is the first phase in the life of a six sigma project. In this phase the problem is well

defined. Objective of the project is determined. Measures for the objective are put in

place. Time Schedule, Customer and Business Impact are also agreed upon during this

phase. Finally and most importantly the scope of the project has also to be defined, the

more limited the project scope is, the higher the probability of success. All the above

mentioned data is documented in the “Project Charter” that has to be signed and approved

by top management before the project is allowed to go on.

Measure Phase

In the second phase of the project, all possible inputs of the process are identified. Data is

then collected about these inputs and their corresponding value of the process output.

This phase is usually one of the longest in the project. Two of the very important things

in this phase are:

• Ensuring that all inputs are considered because if any of them is later discovered

to be missing, data collection has to be repeated.

• Validating the reliability of the data through measurement system analysis, It is

very common that people gather wrong data due to personal, environmental or

equipment related factors. It is obvious that when the gathered data is not

accurate, the results of analysis are consequently mistaken

Analyze Phase

In this phase the power of six sigma and statistics is really evident. During analysis

significant and non significant inputs are recognized. Relationship between significant

inputs and the output can be quantified using a regression model for example. Various

graphical and statistical tools are used in analysis. Graphical tools may include: Pareto,

Box plots, histograms, dot plots, scatter plots and many other tools. Statistical tools such

as multiple regression, logistic regression, Design of experiments, Analysis Of Variance

(ANOVA), paired T, Chi Square and many other tools may be relevant in this phase

Improve Phase

After the critical inputs have been identified in the Analyze phase, it should be now easy

to improve and adjust the process. The equation of the output as a function of the process

inputs that was established in the Analyze phase can now be used to identify the best

settings for the process input parameters that will ensure reaching the desired level of the

output.

Control Phase

The last phase of the project is the control phase. To maintain the results that have been

reached in the improve phase after extensive efforts through out the 3 or 6 months of the

project, control is extremely important. The control is applied on the process critical

inputs that were identified and fixed during the improve phase. Error proofing devices

and techniques could be used if applicable to sustain the new process settings.
Roles within six sigma organizations

In a typical six sigma organization , there exists Yellow Belts “YB”, Green Belts “GB”,

Black Belts “BB” and Champions. Yellow belts are the core team members in any six

sigma project. They are the ones who do most of the work in a project including data

collection and implementation of solutions when identified. Yellow belts are not exposed

to six sigma technical details as they attend only to a one day introductory course.

Green Belts are the team leaders of six sigma projects. They guide yellow belts

throughout the logical steps of the project. Green belts are the ones responsible for

applying six sigma statistical and non

statistical tools whenever relevant or YB YB YB YB YB YB YB YB

YB YB YB YB YB YB YB YB

required. They attend to technical

GB
GB GB
GB
training usually with duration of 10

days.

BB
BB
Theoretically, the presence of Green

and Yellow Belts in an organization

Champion
Champion
may be enough for the execution of
Symbios – Consulting training material
six sigma, however, practical

experience proves that availability of Black Belts and Champions is

essential for successful implementation. Black Belts and champions fulfill two forms of

vital support, which the project leader usually requires. These are technical and

managerial support respectively. Six Sigma Black Belts are very experienced technical
people who often aid green belts during their projects. Black Belts also provide technical

expertise required for project selection and evaluation. In addition to assisting green belts

and selecting projects, BB roles also include leading complex projects and training of

YB, GB and Champions. It is worth mentioning that the black belt is a six sigma full time

job.

The final important role in any six sigma organization is the role of the Champion.

Champions are non technical but rather senior people in an organization. Typically any

senior manager in a six sigma organization is a champion. Champions attend a 2 day

training course that describes their role and its importance in details. The roles of a

champion include adding the strategic dimension in evaluating potential six sigma

projects. Champion roles also include pushing Green Belts to progress quickly in their

projects and removing any possible road blocks in the project path. It has been proven

statistically that most of six sigma implementation trials that were unsuccessful, failed

mainly because of ineffective champions. Very rarely do implementation efforts fail

because of technical incapability of black belts but much more often due to lack of

engagement of management team. This reason is also referred to as “lack of leadership

commitment”. It is a very strong and common reason for failure not only in implementing

six sigma but in any other development effort as well.

 Critical Success factors of Six Sigma Implementation

Because six sigma improvement projects save a lot of money, they also require a lot of

time to get accomplished. A typical Six Sigma improvement project may last from 3 to 6

months which is considered quite long if compared to other improvement techniques such

as “Kaizen” 1. In some special cases a six sigma project may continue to be open for

more than one year. Occasionally, candidates may view their project as a boring and

endless task. In addition, six sigma projects also require a lot of effort and attention to

details. Persistent and committed leadership is therefore essential for the completion of

six sigma projects and consequently for the successful implementation of six sigma in the

entire organization.

Reasons for unsuccessful Implementation

Technical Reasons No Priority for 6S

S upport in the
Lack of S trong
w rong direction
M anagerial S upport
Lo
w

H
F

ig
in

h
an

Lack of v ision for 6S

c

om
ia
l

benefits
Im

pl
ex
pa

Unsuccessful
i ty
ct

S electing the w rong

Implementat
ion of Six
Sigma

P rojects are taking too long N ot E nough KP Is for

six sigma
implementation

N o Rew ard and Recognition S ix S Igma is not in the

P olicy in P lace M gmt goal tree

Lack of motiv ation No measures in Place

1. “Kaizen” is a Japanese word meaning: “Change is good”. It is a term that is used to describe small improvement projects that can be completed in 3 to 5 days
There are two main categories for the reasons why six sigma implementation may fail.

These are technical reasons and management related reasons.

Technical factors may include the selection of the wrong projects and missing the better

ones. When a project with very low impact is selected, it gives a bad impression about six

sigma and what it can do. Same is true when a project with a very wide scope is selected,

probability of success of such project will be very low which again de-motivates people

and gives the wrong impression about six sigma. Some times, the implementation of six

sigma fails because of the technical incapability of the black or green belts but this case is

very rare

Management related factors are also very important; The main reason for failure of

implementation is usually the lack of management commitment. If top management is not

really engaged, if they do not put in the effort of pushing and driving people to dig deeper

and deeper for improvement, the organization will remain stagnant. Managers have to be

convinced with what they can achieve using six sigma methodology and tools otherwise

they will not by driving it as they should. A good idea is that on the beginning of

implementation, limited scope or area of the organization is selected as a pilot run. When

the pilot succeeds it would be easier to convince other departments to go ahead and

implement.

Another important factor for successful implementation is having the correct measures in

place. Having six sigma Key Performance Indicators (KPIs) as well as targets for those

KPIs is necessary for implementation. KPIs may include; number of completed projects,

actual savings to date, customer complaints and others. In some organizations such as
General Electric, the bonus and promotion systems of the organization are directly linked

to such six sigma related KPIs and targets.

Last but not least, some times six sigma improvement efforts fail due to lack of

motivation. As mentioned earlier, six sigma projects need a lot of time and effort. They

also generate a good amount of savings to the companies. Savings of some projects may

exceed 1 million pounds annually. It is important to allocate a certain percentage of such

savings to be spent on reward and recognition of concerned employees, other wise the

running flywheel of projects and savings would eventually stop

Part II - Applications of Six Sigma Methodology and
Tools
In this section we will discuss how can six sigma methodology and statistics in general

add valuable insights in the fields of Manufacturing, service industries, Finance,

Marketing, Human Resources and even in court trials

Use of six sigma and statistics in Manufacturing:

When six sigma was first introduced by Motorola it was used only in the manufacturing

field and more specifically to minimize defects. Below is an example of how the very

famous Pareto tool can be very useful in this regard.

Assuming a manufacturing plant producing a certain product in which several defects

may exist, The Pareto helps us see quickly which ones are the most recurring.

Pareto Chart of Type of Defect

40

100

30
80
Percent
Count

20 60

40
10
20

0 0
Type of Defect Color Leak Scratch Loose fit Dirt
Count 15 8 6 4 2
Percent 42.9 22.9 17.1 11.4 5.7
Cum % 42.9 65.7 82.9 94.3 100.0
In the chart above we could see that 43% of our defects are mainly color defects. Color

and Leak defects together compose 65% of our total defects

In this same example we can also see the Machines responsible for most of the defects as

shown below

Pareto Chart of M/C No.

40

100

30
80

Percent
Count

20 60

40
10
20

0 0
M/C No. M/C No. 2 M/C No. 5 M/C No. 9 M/C No. 1
Count 14 12 6 3
Percent 40.0 34.3 17.1 8.6
Cum % 40.0 74.3 91.4 100.0

Machine numbers 2 and 5 are the ones responsible for almost 75% of our defects.

More interesting than that, we can also see which M/C was responsible for which defect

to better scope our improvement efforts

 Pareto Chart of Type of Defect by M/C No.
C olor Leak S cratch Loose fit Dirt
M/C No. = M/C No. 1 M/C No. = M/C No. 2 Ty pe of
16
Defect
Color
12 Leak
Scratch
8 Loose fit
Dirt
4
Count

0
M/C No. = M/C No. 5 M/C No. = M/C No. 9
16

12

0
C olor Leak S cratch Loose fit Dirt
Type of Defect

Now we have a much better and detailed view. Although we already knew that M/C

numbers 2 and 5 were responsible for 75% of our defects, we now understand what is the

exact problem with each M/C. We can conclude that M/C no. 2 has a lot of problems and

probably needs a full repair or replacement, whereas M/C no 5 has mainly a problem with

adjusting the color of the product. Attention has to be made however to whether it is an

operator fault or M/C fault.

Another example in manufacturing may be comparing different M/C productivity for a

certain product. Doing same exercise for several products may help identify the best

product / Machine mix. i.e which product to work on which M/C assuming we are having

several similar machines. In the chart below it is clear that for product A, machine

number 19 has the best productivity with a performance level higher than 90%

Boxplot of Percent Theoretical / Total Time for Product A

1.0
Percent Theoretical / Total Tim

0.9 0.9

0.8

0.7

0.6

0.5

0.4
MC11 MC12 MC16 MC19 MC20 MC21
Machine
Other applications in manufacturing may involve evaluation of a substitute material for

example. In the chart below the average M/C speed is compared based on the material

type used. We can conclude that types C and F are two of the best options in terms of

machine speed.

Main Effects Plot for Max Machine speed (RPM)

Data Means

100

90

80
Mean

70

60

50

A B C D E F
Material Type

Six sigma and statistical tools can still have more powerful and complex application like

in the following example. In this situation we are trying to model the tension of paper

coming out from a paper production machine based on the machine speed and

temperature. Graphical and statistical analysis are conducted for two types of paper
 3D Scatterplot of Output: Paper tension vs M/C speed - RPM vs M/C Temp
Product
No.
A
B

100

90
O utput: P aper tension

80

70
280
260
345 240 M /C T emp
330
315 220
300
M /C speed - RP M

As can be seen from the above scatter plot, the paper tension increases with the increase

of speed or temp. We can also notice that product A (Shown in black) usually has higher

tension than product B (In red). Further analysis could be done using regression where

we can actually model the relation ship between the paper tension against temperature

and speed. Using a regression equation we can actually estimate the output paper tension

given the M/C settings. Making it even more useful, we can calculate the required M/C

settings that will enable us to reach a required value of paper tension.

Output: Paper tension = - 86.7 + 0.267 M/C speed - RPM + 0.322 M/C Temp

Predictor Coef SE Coef T P

Constant -86.71 11.38 -7.62 0.000
M/C speed - RPM 0.26689 0.02920 9.14 0.000
M/C Temp 0.32211 0.02835 11.36 0.000

S = 2.62329 R-Sq = 89.0% R-Sq(adj) = 88.2%

According to the available data, 88 % of the variation in paper tension is now explained

by the equation above or in other words by the variation in machine speed and temp.
Use of six sigma and statistics in Service Industries

Six sigma could be also used in service industries. Banks, hospitals, call centers and other

service industries make very good use of six sigma methodology and tools. In part 3 of

this document we will discuss in detail how the methodology was useful in solving a

problem experienced by a multinational company in the field of telecommunication.

Here we are going to discuss how small service providers like restaurants can make good

use of six sigma tools.

We are assuming a restaurant manger who is interested in the voice of the customer. He

is having complaints about his service level. Upon collecting some data, he was able to

see client satisfaction level as a function of the client waiting time to make an order

Fitted Line Plot

Customer satisfaction = 121.9 - 7.375 Waiting time
+ 0.1060 Waiting time**2
S 8.36570
100
R-Sq 91.6%
R-Sq(adj) 91.2%

80
Customer satisfaction

60

40

20

0 5 10 15 20 25 30
Waiting time

Looking at the above graph, he could easily notice that satisfaction rating is heavily

affected by the waiting time.

Now, knowing how important the client waiting time is, the restaurant manger tried to

model this waiting time as a function of the number of available staff members and the

hour of the day.

Surface Plot of Waiting time vs No of Staff members, Time of the day

30

20
Waiting time

10
8

0
6 No of Staff members
2
4 4
6
8
Time of the day

Regression Analysis: Waiting time versus Time of the , No of Staff

The regression equation is

Waiting time = 11.7 + 1.69 Time of the day - 1.65 No of Staff
members

Predictor Coef SE Coef T P VIF

Constant 11.728 1.833 6.40 0.000
Time of the day 1.6869 0.1533 11.00 0.000 1.2
No of Staff members -1.6505 0.1916 -8.61 0.000 1.2

S = 2.13341 R-Sq = 87.3% R-Sq(adj) = 86.8%

Looking at the graph and using the equation created through linear regression, he was

able to know how many staff members should be available in his restaurant at each hour

of the day so that the client waiting time would be equal to 10 minutes or less. He then

hired people on part time basis, so that even on the peak of deman he always had the

appropriate number of staff members

Statistics and Marketing:

Marketing is of course one of the fields where statistics and six sigma tools could be used

the most. A simple example is illustrated here where a marketer is interested to see

whether customer preference for different car brands depends on the area where the

customer lives or not. She looks at the data of the cars sold during the last 3 months in

three areas: Giza, Maadi and Heliopolis. She is interested mainly in 3 brands: Audi,

Mercedes and Jeep.

Pareto Chart of Brand by Dealer Name (Customer)

Jeep Mercedes Audi
Dealer N ame (C ustomer) = G iza Branch D ealer N ame (C ustomer) = H eliopolis Branch Brand
30
Jeep
Mercedes
20 Audi

10
Count

0
Dealer N ame (C ustomer) = M aadi Branch
30

20

10

0
Jeep Mercedes Audi
Brand

Looking at the chart above, she could easily see the dependence of product type on dealer

location. People living in Giza prefer to have a Mercedes where as those living in Maadi

prefer driving a Jeep. Heliopolis customers do not have a certain preference among the

three brands. The existence of the relationship between preferred product and area could

be proved using a Chi square test which has a wide use in the field of marketing
Chi Square Test: Dealer Name (Customer), Brand

Rows: Dealer Name (Customer) Columns: Brand

Audi Jeep Mercedes All

Giza Branch 5 8 16 29
6.56 11.39 11.05 29.00

Heliopolis Branch 10 8 9 27
6.11 10.61 10.29 27.00

Maadi Branch 4 17 7 28
6.33 11.00 10.67 28.00

All 19 33 32 84
19.00 33.00 32.00 84.00

Cell Contents: Count

Expected count

Pearson Chi-Square = 12.277, DF = 4, P-Value = 0.015

In this case the Chi square test was able to prove the existence of such relationship.

Similar marketing applications studying the effect of demographics such as gender, age

group, social class or others on product preference or packaging color and design for

example do exist.
Use of six sigma and statistics in Finance

Finance departments in any company are usually good sources for reliable data. The data

when analyzed however becomes much more useful. A simple example is a cost break

down using Pareto analysis. Pareto chart could be used for expenses break down can be

done for expenses by department or for the percent of budget compliance by department,

or as in the below case; cost breakdown by business unit.

Pareto Chart of Country

4000000 100

80
3000000

60

Percent
Cost

2000000
40

1000000
20

0 0
Country Morroco Algeria Sudan Egypt Ethopia Other
Cost 1205930 1184768 826791 615568 107003 107003
Percent 29.8 29.3 20.4 15.2 2.6 2.6
Cum % 29.8 59.1 79.5 94.7 97.4 100.0

Another example would be analysis of delayed payments to improve collection period. In

the below example the finance manager is interested to know whether the delay is

customer dependent or not.

 Boxplot of Payment Duration (in Days)

80

70
Payment Duration (in Days)

60

50

40

30

20
Adawy Corporation Auto Maged El Masry & Sons
Dealer Name (Customer)

One-way ANOVA: Payment Duration (in Days) versus Dealer Name (Customer)

Source DF SS MS F P
Dealer Name (Customer) 2 1479 740 5.04 0.009
Error 69 10130 147
Total 71 11609

S = 12.12 R-Sq = 12.74% R-Sq(adj) = 10.21%

Level N Mean StDev

Adawy Corporation 24 57.00 14.19
Auto Maged 24 46.54 11.88
El Masry & Sons 24 48.54 9.90

Individual 95% CIs For Mean Based on Pooled StDev

Level -+---------+---------+---------+--------
Adawy Corporation (-------*-------)
Auto Maged (--------*-------)
El Masry & Sons (-------*-------)
-+---------+---------+---------+--------
42.0 48.0 54.0 60.0

In this case, the ANOVA test was able to prove that the customer named “Adawy

Corporation” is paying in more time than the other two customers. Finance manager can

now focus his efforts on collecting payments from this particular customer on time.
Use of six sigma and statistics in HR

Graphs and statistical tests could be also used here to study the effect of post, age group

or years of experience on job satisfaction for example. In the below example the effect of

gender and department on organizational commitment are studied

Boxplot of Mean Org Commitment

6
Mean Org Commitment

Female Male
GENDER

Boxplot of Mean Org Commitment

6
Mean Org Commitment

Business Dev Finance Operations Security Supply Chain WH

Dept 2
In the above example, females seem to be more committed than males. Similar to all

other examples, hypothesis testing should be made -based on representative sample- to

prove whether such difference exist or not. It is worth mentioning that especially in the

field of HR data source could be a questionnaire rather than observations like in the other

applications. This sometimes has its negative effect on the quality of data and

consequently the quality of analysis and conclusions reached.

In the example below, the organizational commitment is modeled as a function of the

state of hope of employees. Two separate models are created for males and females in the

organization. When a strong relationship of this kind is noticed, HR could use the state of

hope of interviewees – provided they are able to assess it - as a predictor for their

expected organizational commitment

Scatterplot of Mean Org Commitment vs Mean state of Hope

GENDER
6
Female
Male
Mean Org Commitment

3 4 5 6 7 8
Mean Dispositional Hope

Analysis of employee turn over is another famous application of six sigma tools in the

field of HR.
Statistics and Legal Judgments

Statistical tests are widely used in court in the western countries. Imagine a juice

manufacturer who states on his package that it contains 1 liter. Due to normal process

variation, quantity in each package may not be exactly equal to 1 liter. Some packages

may have 1.05 liters and others may have 9.9 for example. However, if the manufacturer

is honest, the over all average of 100 packages is expected to be very close to 1 Liter.

Assuming that the overall average turns out to be 0.96 liters, some people may claim that

the manufacturer is cheating whereas others may say no he is honest, its only sampling

errors. To prove whether the real average is below one litter or not, a hypothesis test has

to be made.

One-Sample T: Actual Litters

Test of mu = 1 vs not = 1

Variable N Mean StDev SE Mean 95% CI T P

Actual Litters 50 0.97036 0.04660 0.00659 (0.95712, 0.98360) -4.50 0.000

Histogram of Actual Liters

(with Ho and 95% t-confidence interval for the mean)
12.5

10.0

7.5
Frequency

5.0

2.5

0.0 _
X
Ho

0.875 0.900 0.925 0.950 0.975 1.000 1.025 1.050

Actual Liters
 Boxplot of Actual Liters
(with Ho and 95% t-confidence interval for the mean)

_
X
Ho

0.85 0.90 0.95 1.00 1.05

Actual Liters

In this example the t-test has proved that the true average of the population is somewhere

between 0.957 and 0.983 Litters with 95% confidence. In other words the true average

cannot be 1 litter at 95 % confidence level. In western countries such result may be

enough to convict this juice manufacturer.

Recommendations and Conclusion

In this paper we have discussed how six sigma and statistics in general could be very

useful in various business fields. Examples from marketing, Finance, manufacturing and

human resources management were also discussed. A detailed case study from

telecommunication industry was presented in detail. It is important to note that all the

tools presented in this paper represent less than 25% of the tools that could be used in a

real six sigma project. Six sigma thinking methodology is also important not only the

tools. The concept of “Y = f (X)” is a very important one that people usually miss when

looking into their professional and their personal life problems as well.

The success or failure of implementing six sigma in an organization depends mainly on

how committed the leadership is. Their commitment depends usually on how convinced

they are with what a tool like six sigma can do for their business. Six sigma is of course

not the only way for improvement, but rather one among several useful ways.