Download as pdf or txt
Download as pdf or txt
You are on page 1of 18

International Journal of Lean Six Sigma

Selecting Six Sigma project: a comparative study of DEA and LDA techniques
A. Hadi-Vencheh, A. Yousefi,
Article information:
To cite this document:
A. Hadi-Vencheh, A. Yousefi, (2018) "Selecting Six Sigma project: a comparative study of
DEA and LDA techniques", International Journal of Lean Six Sigma, https://doi.org/10.1108/
IJLSS-11-2016-0067
Permanent link to this document:
https://doi.org/10.1108/IJLSS-11-2016-0067
Downloaded on: 13 July 2018, At: 18:59 (PT)
References: this document contains references to 87 other documents.
Downloaded by INSEAD At 18:59 13 July 2018 (PT)

To copy this document: permissions@emeraldinsight.com


The fulltext of this document has been downloaded 1 times since 2018*
Access to this document was granted through an Emerald subscription provided by emerald-
srm:178665 []
For Authors
If you would like to write for this, or any other Emerald publication, then please use our Emerald
for Authors service information about how to choose which publication to write for and submission
guidelines are available for all. Please visit www.emeraldinsight.com/authors for more information.
About Emerald www.emeraldinsight.com
Emerald is a global publisher linking research and practice to the benefit of society. The company
manages a portfolio of more than 290 journals and over 2,350 books and book series volumes, as
well as providing an extensive range of online products and additional customer resources and
services.
Emerald is both COUNTER 4 and TRANSFER compliant. The organization is a partner of the
Committee on Publication Ethics (COPE) and also works with Portico and the LOCKSS initiative for
digital archive preservation.

*Related content and download information correct at time of download.


The current issue and full text archive of this journal is available on Emerald Insight at:
www.emeraldinsight.com/2040-4166.htm

Selecting six
Selecting Six Sigma project: sigma project
a comparative study of DEA
and LDA techniques
A. Hadi-Vencheh
Department of Mathematics, Isfahan (Khorasgan) Branch,
Islamic Azad University, Isfahan, Iran, and Received 10 November 2016
Revised 8 July 2017
30 August 2017
A. Yousefi Accepted 13 September 2017
Department of Customer Service,
Esfahan Province Electricity Distribution Co (EPEDC), Isfahan, Iran
Downloaded by INSEAD At 18:59 13 July 2018 (PT)

Abstract
Purpose – Nowadays, most of the organizations have focused through the world on Six Sigma to reduce the
costs, improve the productivity and enhance concerned individuals’ satisfaction, especially customers’
satisfaction. Annually, these organizations define and execute thousands of Six Sigma projects which involve
a great deal of investments. But are all of these projects successful and do the organizations benefit from the
above advantages? The purpose of this study is to proposing a methodology to to answer this question that:
How can we reduce the risk of failure in Six Sigma projects? The first step to reduce the risk of failure in Six
Sigma projects is selecting optimal ones which have the most profits and the least expected risks.
Design/methodology/approach – First, the effective criteria are recognized and defined in selecting Six
Sigma projects. Then, a new data envelopment analysis (DEA) model is proposed for project selection process.
A real example is resolved by the presented model. Finally, the authors use linear discriminate analysis (LDA)
to examine the validity of obtained results from the proposed model.
Findings – The results show that the proposed model is a suitable tool for selecting Six Sigma Projects. The
findings demonstrate that the selected projects by suggested integrated DEA model are those confirmed by
LDA.
Originality/value – The paper, using a real case study, provides a mathematical model to enhance
decision quality in Six Sigma project selection. Applying the specific DEA model is remarkable itself, which
joined to a pioneering procedure to use LDA to validity evaluation of the results.
Keywords Six Sigma, DEA, LDA, Project selecting
Paper type Research paper

1. Introduction
Six Sigma is an innovative method and improves strategically the business. It is a customer-
based, orderly, advantageous and systematically optimizing method which encompasses
the whole organization, is based on measurable methodology and focuses steadily upon the
process. Six Sigma uses statistical methods and the principles of quality management to
optimize processes and products. It is done by Defining, Measuring, Analyzing, Improving
and Controlling (DMAIC) the framework of quality improvement and based on a project-to-
project process to meet the customers’ needs (Goh, 2002; Tang et al., 2007).
International Journal of Lean Six
The authors would like to express their appreciation to the Editor, Prof. Jiju Antony and three Sigma
anonymous reviewers for their careful reading and invaluable and constructive comments that have © Emerald Publishing Limited
2040-4166
improved the presentation of this paper. DOI 10.1108/IJLSS-11-2016-0067
IJLSS In recent years, manufacturing organizations succeeded in using Six Sigma as a strategy
to reduce the numbers of defective unites in the manufacturing process, consequently
reducing the costs and increasing the profits (Kumar et al., 2007). It has been reported the
advantages of Six Sigma in the related literatures (Hahn et al., 1999; Harry, 1998; Hendricks
and Kelbaugh, 1998; Lanyon, 2003; Robinson, 2005; Dhiraj and Deepak, 2014; etc). However,
there are considerable cases in which the Six Sigma has failed in producing the expected
results (Ahadian and Abadi, 2012). The results from a research by the journal “Aviation
Week” among the major aerospace companies showed that about 50 per cent of these
companies were satisfied with the results of Six Sigma projects, 30 per cent of them were
unsatisfied and nearly 20 per cent of them were partially satisfied (Zimmerman and Weiss,
2005). Six Sigma is not an appropriate methodology, while a company is facing radical
changes (Mellat Parast, 2011). Yet, comparing with the process optimizing techniques used
in recent five decades, Six Sigma has emerged evidently as the most influential technique in
quality optimizing (Dusharme, 2003).
Six Sigma is developed from the previous quality methods such as total quality
management (TQM), deeming statistical quality control. Six Sigma can be defined as a
Downloaded by INSEAD At 18:59 13 July 2018 (PT)

strategy, involving TQM, strong focus on customers, enhanced tools for analyzing data,
financial results and project management (Anbari, 2002; Kwak and Anbari, 2004). Although
it is originated from the manufacturing organizations and intended to reduce defects in
producing process, nowadays, all industries use Six Sigma, including service industries
such as health-care management (Antony, 2004; Antony and Fergusson, 2004; Krupar, 2003;
Moorman, 2005; Yousefi and Aqamohammadi, 2014). Contrary to potential of Six Sigma in
extensive applications, none of other quality improvement initiatives has gained such high
application outside the manufacturing organizations. That is because of having a specific
methodology and some clear procedures and roadmaps, as well as a useful applicable bi-
dimensional concept which interweaves both quality management and statistical and
qualitative analysis.
Now the question is not whether to use Six Sigma, but how to execute successfully a Six
Sigma process optimizing project. Abdalhkeim et al. (2017) provided the sources for the
identification of potential projects, top-down and bottom-up approaches, the process of
selection of projects, guidelines that assist the selection of appropriate projects, selection
criteria which must be precisely chosen in accordance with the objectives, needs and
capabilities of the organization.
In this paper, we propose a new DEA model to select Six Sigma projects. Then, we use
linear discriminate analysis (LDA) to examine the validity of results from proposed model.
The rest of this paper unfolds as follows: In the following section, literature review, previous
researches related to six sigma project selection have been glanced. After that, the criteria to
select a Six Sigma project have been determined in Section 3. Then, the basic concepts of the
used approaches have been discussed in Section 4. Besides, we propose our new method in
this section. In Section 5, we illustrate our method by a real-word data. Section 6 includes
managerial implications and limitations of the proposed methodology. Finally, conclusions
are given in Section 7.

2. Literature review
Probably the most difficult step in Six Sigma is to select a process optimizing project and
most of debates in Six Sigma literatures are devoted to selecting Six Sigma projects (Pande
et al., 2000; Goldstein, 2001; Snee, 2001; Fundin and Cronemyr, 2003; Mc Adam and Evans,
2004; Ho et al., 2006; Yang et al., 2008; Brun, 2011; Yousefi and Aqamohammadi, 2014;
Marzagalo and Carvalho, 2016; Yousefi and Hadi-Vencheh, 2016; Yousefi, 2017; Abdalhkeim Selecting six
et al., 2017). sigma project
It necessitates various resources to implement a Six Sigma project, such as investments
and manpower (in the form of green and black belts). Based on these resources, a successful
Six Sigma project can reveal useful outputs in the form of enhancing the quality level of Six
Sigma, increasing customer’s satisfaction, reducing cost of poor quality (COPQ). Banuelas
et al. (2006) using a survey identified what criteria are considered for selecting Six Sigma
projects and how Six Sigma projects are selected in organizations in the UK. Ray and Das
(2010) proposed a top-down approach for project selection, as top management support for
Six Sigma initiatives is absolutely critical to see tangible, significant results. The authors
suggest establishing the linkage with data (either reactive or survey), otherwise through
prioritization tool for project selection. Sharma and Chetiya (2010) study the Six Sigma
project selection approach and the factors responsible in the Indian manufacturing industry
with foreign technical collaboration. Marques et al. (2014) identify the following four kinds
of sources which they put into a four area scheme: external, internal, retroactive and
proactive resources.
Downloaded by INSEAD At 18:59 13 July 2018 (PT)

According to Larson and Gray (2011), the initial project identification should include the
following necessary activities:
 reconsidering the mission of the organization;
 establishing long-term goals and objectives;
 the analysis and development of strategies for achieving desirable goals; and
 the establishing of portfolio of project choice.

Based on Abdalhkeim et al. (2017), the final list of projects is subject to further evaluation
and selection.
The process of project selection generally comprises the following steps:
 compiling a list of potential project candidates (potential projects) – the data about
listed projects should be prepared in an appropriate format and completed;
 establishing criteria for project selection – the criteria should be established
according to the predefined requirements and in compliance with the mission; and
 prioritization of projects – the selected projects are assessed according to the
established criteria to achieve a rated result;
 the choosing of projects – at this stage the best rated projects are chosen and a case
study is developed to get the necessary approval of the top management.

There are plenty of techniques and models have been applied to select the Six Sigma
projects as follows. Adams et al. (2003) suggested Project Ranking Matrix. Nourpanah et al.
(2011) integrated group analytic hierarchy process and linear programming (LP). Ahadian
and Abadi (2012) provided a multi-criteria decision-making (MCDM) approach to select Six
Sigma Pilot Projects. Shanmugaraja et al. (2012) developed a model of project selection based
on quality function deployment (QFD).
Recently, Ortiz et al. (2015) wrote about the application of analytic network process
(ANP) methods and decision-making trial and evaluation laboratory (DEMATEL) in the
selection of projects. Vinodh and Swarnakar (2015) used hybrid fuzzy-based MCDM
approach for optimal Lean Six Sigma (LSS) project. A weighted scorecard method was
presented by Holmes et al. (2015) for Six Sigma project selection in a higher educational
institution.
IJLSS More recently, Altintas et al. (2016) presented a fuzzy ANP method in group decision to
determine the criteria and evaluate the Six Sigma projects. Yousefi and Hadi-Vencheh (2016)
proposed using mathematical models to enhance decision quality in Six Sigma project
selection. They found that by using fewer criteria, the results from AHP and TOPSIS are
very similar. Also, the results from these techniques vary from the ones from the data
envelopment analysis (DEA) in many aspects. Sefair et al. (2017) provided linear solution
schemes and application in the oil and gas industry. Finally, Kalashnikov et al. (2017)
suggested a bi-objective programming for project portfolio selection.
Relevant literature brings a number of other techniques recommended for the selection of
Six Sigma projects. However, in practice, the use of sophisticated methods and techniques is
limited, and numerous organizations still select projects ad hoc (Abdalhkeim et al., 2017).
Table I lists the tools used to select the Six Sigma projects.

3. Defining criteria for selecting the Six Sigma project


Selecting a proper Six Sigma project demands precise analyses. The selected project must be
orientated to strategic goals of organization. Pande et al. (2000) categorized the criteria for
Downloaded by INSEAD At 18:59 13 July 2018 (PT)

Six Sigma project as follows:


 advantages of the project for business;
 accessibility to the project; and
 influence of the project on the organization.

Advantages of the project for business include influence upon the customer, influence
upon the strategy of business and influence upon the major competencies, financial and
immediate influences. The criteria for accessibility to a Six Sigma project are as follows:
needed resources, expertise in access, complexity and possibility to success. Learning
and cross-responsibility are discussed under the influences of the project on the
organization.
Harry and Schroeder (2000) propose the following criteria for a Six Sigma project:

Tools Author

QFD Pande et al. (2000)


Pareto priority index, AHP, QFD, theory of constraints Pyzdek (2000, 2003)
Project assessment matrix Breyfogle et al. (2001)
Project selection matrix Kelly (2002)
Pareto analysis Larson (2003)
Project ranking matrix Adams et al. (2003)
Reviewing data on potential projects against specific criteria De Feo and Barnard (2004)
AHP, DEA Kumar et al. (2006, 2007)
Group AHP, LP Nourpanah et al. (2011)
MCDM approach Ahadian and Abadi (2012)
QFD Shanmugaraja et al. (2012)
ANP, DEMATEL Ortiz et al. (2015)
Hybrid fuzzy-based MCDM Vinodh and Swarnakar (2015)
Weighted scorecard Holmes et al. (2015)
Table I. Fuzzy ANP Altintas et al. (2016)
Methods used to DEA, AHP, TOPSIS Yousefi and Hadi-Vencheh (2016)
select a Six Sigma Linear solution schemes Sefair et al. (2017)
project Bi-objective programming Kalashnikov et al. (2017)
 defect per million opportunities (DPMO); Selecting six
 pure savings in costs; sigma project
 costs of poor quality (COPQ);
 period of cycle;
 customer’s satisfaction;
 capacity; and
 internal performance.
Banuelas et al. (2006) indicated the following six criteria for a Six Sigma project:
(1) influence upon the customers;
(2) financial influences;
(3) top management commitment;
(4) measurability and accessibility;
(5) development and learning; and
Downloaded by INSEAD At 18:59 13 July 2018 (PT)

(6) link to strategy of business and competency.


Adebanjo et al. (2016) listed the following criteria:
 feasibility;
 financial impact;
 impact on the customer;
 impact on operation goals;
 impact on the employees; and
 impact on the supplier.

Yousefi and Hadi-Vencheh (2016) selected the following criteria and concluded those are
suitable criteria for selecting Six Sigma projects using analytical hierarchy process (AHP),
technique for order preference by similarity to an ideal solution (TOPSIS) and data
envelopment analysis (DEA):
 project cost;
 reduction of low-quality costs;
 project implementation period;
 customer satisfaction;
 increase of Sigma level (Sigma quality); and
 top management responsibility.

It is obvious in research literatures that it is a MCDM to select a Six Sigma project. Literature
brings other criteria as well. An organization cannot use all the criteria in process selection.
It is necessary to establish a limited number of criteria which are the most relevant in a
particular situation (Abdalhkeim et al., 2017).
Therefore, we consider six most relevant criteria which are mostly used in research
literatures as ones to select a Six Sigma project. In addition, covering all aspects of the
problem was our other attribute to define the criteria. On the other word, determined criteria
should truly cover all properties and impacts of the projects. Based on what is mentioned
above, following six criteria are selected:
IJLSS (1) cost of project (Cost);
(2) drop-off in costs due to poor quality (COPQ);
(3) duration of executing the project (Time);
(4) customer’s satisfaction (Customer);
(5) increasing the sigma level (Sigma); and
(6) responsibility of top management (Top Mngr.)

As it considerable, the six defined criteria completely covered the most dimensions of the projects,
including other non-selected criteria which noticed in literature. Anyway, note that model can
vary according to the decision maker’s vision by adding or eliminating some of criteria.

4. Used techniques
4.1 Data envelopment analysis
DEA is a technique for deciding the relative efficiency of a decision-making unit (DMU) by
comparing it with linear combinations of other DMUs engaged in providing the same
Downloaded by INSEAD At 18:59 13 July 2018 (PT)

outputs from the same inputs. The pioneering work of the economist Farrel (1957) provided
a non-parametric method of determining the relative efficiency of a DMU. He gave a method
of computing the facets of the efficient production frontier from a set of observations from
empirically observed DMUs, rather than by estimating the parameters of postulated
production functions. The celebrated paper of Charnes et al. (1978) introduced DEA, an LP-
based technique, to measure Farrel type efficiency. Banker et al. (1984) developed a model
assumed to be variable returns-to-scale, known as BCC. In recent decades, DEA has rapidly
expanded into new application areas (Cook and Seiford, 2009; Hatami-Marbinia et al., 2011;
Shabani et al., 2015; Hosseini and Abedi, 2015; Wu et al., 2016; Chen et al., 2017; Rosenthal
and Weiss, 2017; etc). The following model shows an input-oriented CCR model.
Miny0 ¼ u
st :
Xn
l j yrj  yr0 ðr ¼ 1; 2; :::; sÞ
j¼1 (12)
X
n
l j xij # u xi0 ði ¼ 1; 2; :::; mÞ
j¼1
l j  0; u : Free
where the value calculated for u indicates the rate of efficiency for examined DMU by j0; yrj
and xij are rth output and ith input of jth DUM, respectively, yr0 and xi0 are rth output and ith
input of examined DMU, respectively.
In this paper to assess the efficiency of units, the following is considered to select an
appropriate model:
 In addition to the inputs and outputs, there are some environmental factors which
must be incorporated into the model.
 As the input-oriented and output-oriented models provide the same results, then an
input-oriented model is used.
 To select the units being improved, non-decreasing returns-to-scale is considered in
the model.
Therefore, to measure the efficiency of Six Sigma projects and considering the Selecting six
environmental factors, we propose an input oriented model with nondecreasing returns-to- sigma project
scale (NDRS) as follows:
X
2 X
3
MinZ0 ¼ u  « : Sni  « : Spr
i¼1 r¼1
s:t :
X 9
yrj l j  Spr ¼ yr0 r ¼ 1; . . . ; 5:
j¼1
X9
xij l j þ Sni  u : xi0 ¼ 0 i ¼ 1; 2; 3:
(13)
j¼1
X9
Elj l j  u : El0 # 0 l ¼ 1; . . . ; 10:
j¼1
X9
Downloaded by INSEAD At 18:59 13 July 2018 (PT)

lj  1
j¼1

l j ; Spr ; Sni  0; u : Free


where variables, homograph with the model (12), have the same definition and « , Sni, and
Spr are added to it to improve the model. Elj shows lth environmental factors for jth DMU, j =
1, . . ., 9.
As noted above, to assess the efficiency of units in DEA, the attributes must be
categorized into input and output variables. Here, there are some attributes which are
neither input, nor output, but they can influence the efficiency of units. These attributes are
environmental factors or variables. The input variables are cost of project (Cost) and
duration of executing the project (Time). The output variables are drop-off in costs due to
poor quality (COPQ), customer’s satisfaction (Customer) and increasing the sigma level
(Sigma). The responsibility of top management (Top Mngr) is considered as an
environmental factor.

4.2 Linear discriminate analysis


LP methods of discriminate analysis have received a great deal of attention, for example, see
Glover et al. (1988), Glover (1990), Ragsdale and Stam (1991), Yang and Wu (2014) and
recently Hou and Riley (2015), Wang et al. (2016), Safo and Ahn (2016) and most recently
Shao and Sang (2017), Xu (2017), among others. DA is a method for determining group
classification for a set of similar units or observations. For example, DA may be used to
classify loan applicants as either good or bad credit risks and hence determine whether to
accept or reject new applicants. The same relevant factors are measured for all units in the
set, and there is some distinct outcome or occurrence which determines group membership
for each unit. For example, whether the loan recipient defaults on the loan. The objective of
DA is to find a set of factor weights which best separates the groups, given a set of units for
which group membership is already known. The resulting set of weights may then be used
to predict group membership for new units.
In general, DA involves two or more groups, but the two-group scenario has received the
most attention, as in Glover (1990). The two-group case shall be focused on exclusively here
because this is the scenario which relates most directly to DEA. Assume n units (j = 1, . . ., n)
IJLSS and k independent factors or variables (i = 1, . . ., k) have been recorded for each unit and
denoted zij. No previous knowledge of factor effects or relations is used, only previous
knowledge of group membership is used. It is known that n1 units are in group 1 (denoted
j [ G1) and n2 units are in group 2 (denoted j [ G2) and n1 þ n2 = n. Then a basic version of
the DA model is:
Xn
Min dj (1.0)
j¼1

X
k
S:t wi zij  dj < T; j 2 G1 ; (1.1)
i¼1

X
k
wi zij þ dj  T; j 2 G2 ; (1.2)
i¼1
Downloaded by INSEAD At 18:59 13 July 2018 (PT)

dj  0; j ¼ 1; :::; n; (1.3)

wi ; T unrestricted: (1.4)
This model results in an optimal vector of factor weights w* as well as T*, which together
define a hyper plane that attempts to separate the two groups. In essence, each unit receives
a “score” which determines group classification by being either above or below the
“threshold” or boundary value T*. Here, dj is the external deviation variable, which is the
amount by which unit j is misclassified. Subsequent to obtaining (w*, T*), new units can be
evaluated to predict their future group membership.
This model cannot be used as shown because there is nothing to prevent the trivial
solution (w* = 0, T* = 0) from occurring. There are two basic approaches to dealing with
this problem. One involves the use of a small positive number, « , to impose a “gap” of
separation between the groups, used by Ragsdale and Stam (1991), Stam and Ragsdale
(1992) and Glover et al. (1988). A formulation of this type of DA model is:
Xn
Min dj (2.0)
j¼1

X
k
S:t wi zij  dj # T  « ; j 2 G1 ; (2.1)
i¼1

X
k
wi zij þ dj  T; j 2 G2 ; (2.2)
i¼1

dj  0; j ¼ 1; :::; n (2.3)

wi ; T unrestricted: (2.4)
The other approach involves the use of normalizations, as in Koehler (1989), Freed and Selecting six
Glover (1986) and Glover (1990). The recommended normalization has changed repeatedly sigma project
as weaknesses were found with earlier versions. These weaknesses relate to unacceptable
solutions and solutions not being invariant under data transformation, which involves data
rotation and/or translation. In DA, the solution invariance issue is considered important and
earlier DA methods were criticized for their inability to deal with this problem, for example,
see Markowski and Markowski (1985), Bajgier and Hill (1982) and Freed and Glover (1986).
The most recent formulation of this type is that of Glover (1990), which involves a
normalization of the deviation variables, as shown below.
X
n X
n
Min h0 d0 þ hj dj  k0 B0  kj Bj (3.0)
j¼1 j¼1

X
k
S:t wi zij  d0  dj þ B0 þ Bj ¼ T; j 2 G1 (3.1)
i¼1
Downloaded by INSEAD At 18:59 13 July 2018 (PT)

X
k
wi zij þ d0 þ dj  B0  Bj ¼ T; j 2 G2 (3.2)
i¼1

X  X 
2n1 n2 ðB0  d0 Þ þ n2 Bj  dj þ n1 Bj  dj ¼ 2n1 n2 (3.3)
j2G1 j2G2

d0 ; B0 ; dj ; Bj  0; j ¼ 1; :::; n (3.4)

wi ; T unrestricted (3.5)
Here, the internal deviation variables (Bj) are included, as well as d0 and B0, which in a sense
represent the maximum external and internal deviations, respectively. However, as all four
deviation variables are included, this is not an exact interpretation. This model effectively
deals with both translations and rotations. In fact, Glover et al. (1988) show that any
normalization on the deviation variables would suffice in this regard, for this form of DA
model. The parameters in the objective function allow different priority to be given to
different goals, as well as different priority for different units.

5. Analysis, application and results


The problem is ranking some homogenous projects (DMUs) which have the same input and
output factors. Also, maybe the projects are influenced by some environmental factors.
Therefore, calculating the productivity of the projects is a remarkable parameter for decision
maker. In addition, decision makers need a technique to provide sensitivity analysis and
benchmark the DMUs of the problem and be able to compare values of input and output
factors of DMUs after rudiment analysis. Furthermore, the specific model should be enough
flexible to analyze crisp data as well as imprecise ones.
DEA is a model that satisfies all the properties above and even something more. By the
way, LDA, as one of the bests categorizing techniques, has been applied to examine validity of
the result of proposed DEA model. Here is a complete flexible mathematical package to rank,
IJLSS categorize, compare, improve and select the Six Sigma projects. The entire characteristics
above made our suggested integrated model to a dominant model in comparison of other
presented models in the literature and justified why we implement the following procedure.
In this section, we apply our model to a real-world data. Data are a development of an
example provided by Kumar et al. (2007), which is a case study of the cycle DMAIC on
starter batteries of locomotives in an Asian country. In current step, decision maker
determines values of the projects’ criteria base on the primary Gant chart and CPM plan,
resource and material plan, voice of customers, calculating the sigma situation before and
after each project, as well as interview with top manager. Obviously, it can be the most
challenging step to forecast the values of the six criteria by crisp numbers, exactly. So, a
perfect opportunity will be to use imprecise numbers such as Fuzzy numbers (Yousefi, 2009)
for the future study. Table II shows the values of six variables for projects.
At first, the relative efficiency of projects is assessed by DEA in this example which
includes 20 Six Sigma projects. Then, after ranking the projects according to their efficiency,
LDA is used to assess the results.
Considering the environmental factors, we propose an input oriented model of DEA with
Downloaded by INSEAD At 18:59 13 July 2018 (PT)

non decreasing returns-to-scale (NDRS) in this paper, (13). The input attributes are: cost of
project (Cost) and duration of executing the project (Time). The output attributes are drop-off in
costs due to poor quality (COPQ), customer’s satisfaction (Customer) and increasing the sigma
level (Sigma). The only environmental factor is responsibility of top management (Top Mngr).
After primary execution of the model, we observe that four projects (7, 12, 17 and 19) are on
the efficient frontier and their efficiency is equal to 1. To ranking the efficient projects, the
methods of ranking efficient units are used, and the values of super-efficiency are obtained
(Anderson and Peterson, 1993; Esmaeilzadeh and Hadi-Vencheh, 2015). The values of super-
efficiency for efficient units (which their efficiency is less than 1) are the same values of their
efficiency. In input oriented models, the values of super-efficiency for efficient units are higher
than 1, and it allows ranking the efficient units. To solve the problems, the DEA Solver software

Project no. Cost COPQ Time (day) Costumer Sigma Top Mngr

1 212 331 70 11 0.24 0


2 199 342 63 29 0.77 1
3 214 333 88 28 0.33 1
4 280 303 77 29 0.48 0
5 263 240 72 19 0.41 1
6 203 306 70 21 0.52 1
7 196 345 61 31 0.78 1
8 215 264 79 22 0.27 0
9 281 239 71 17 0.66 1
10 233 338 66 10 0.30 1
11 263 310 84 27 0.88 1
12 198 341 60 32 0.51 0
13 220 308 80 15 0.31 1
14 284 325 79 23 0.31 0
15 214 314 87 19 0.54 0
16 235 236 80 27 0.34 1
Table II. 17 200 339 63 33 0.83 1
Values of six 18 217 313 75 13 0.74 1
variables for 20 Six 19 198 343 63 31 0.87 1
Sigma projects 20 227 317 70 10 0.46 1
was used. Table III shows the values of efficiency and super-efficiency for projects, as well as Selecting six
their ranking. It shows that projects 12, 19, 17, 7 and 2 are in the rank 1 to 5, respectively. sigma project
Now it is time to test the results from DEA by LDA. We divide 20 projects into two
groups. In the first group, there are four projects which were considered as efficient in DEA
and the remaining projects (inefficient) are in the second group. To analyze the data, we used
SPSS software. By LDA, the following results were observed.
Table IV shows the correlation coefficient of the criteria. As we see, no criteria have the
positive, significant correlation coefficient.
The special values are shown in the Table V, in which the focus correlation is 0.866.
According to the Table VI, the performed categorization is fully significant, because it is
Sig. = 0.002. It is necessary Sig. < 0.05 to be significant.
We see in the Table VII that all four projects dedicated to the first group by DEA
(efficient projects) were confirmed to be in the first group by LDA. But LDA transferred one
of 16 projects, which DEA put in the second group, to the first group. Finally, initially
categorization by DEA was confirmed by LDA for 95 percent of the projects.
Table VIII shows the categorizations by DEA and LDA, survey scores for each project
and the possible membership of each project in Groups 1 and 2. We observe here that the
Downloaded by INSEAD At 18:59 13 July 2018 (PT)

Project no. Efficiency Super efficiency Ranking by DEA

1 0.93396 0.93396 8
2 0.98507 0.98507 5
3 0.91667 0.91667 10
4 0.77922 0.77922 19
5 0.83333 0.83333 18
6 0.96585 0.96585 7
7 1 1.02725 4
8 0.92093 0.92093 9
9 0.85290 0.85290 15
10 0.90909 0.90909 11
11 0.84825 0.84825 16
12 1 1.56298 1
13 0.89189 0.89189 13
14 0.75949 0.75949 20
15 0.97966 0.97966 6
16 0.83544 0.83544 17
17 1 1.04342 3 Table III.
18 0.90411 0.90411 12 Values and ranking
19 1 1.05878 2 for Six Sigma
20 0.86649 0.86649 14 projects using DEA

Correlation COST COPQ TIME CUSTOMER SIGMA TOPMNGR

COST 1.000 0.386 0.112 0.183 0.082 0.129


COPQ 0.386 1.000 0.086 0.118 0.049 0.105
TIME 0.112 0.086 1.000 0.386 0.054 0.204
CUSTOMER 0.183 0.118 0.386 1.000 0.227 0.078 Table IV.
SIGMA 0.082 0.049 0.054 0.227 1.000 0.455 Pooled within-groups
TOPMNGR 0.129 0.105 0.204 0.078 0.455 1.000 matrices
IJLSS Project 2 is the one transferred by LDA from the second group into the first group. It may
appear that this transfer indicates a wrong ranking caused by DEA, but in fact, the DEA
ranking is confirmed by this transfer because the review of efficiency of projects in DEA
reveals that the Project 2 has the most efficiency among the inefficient projects (0.98507) and
is located after four projects, in the fifth rank. Hence, five dominant projects are the same in
both DEA and LDA; they are projects 12, 19, 17, 7 and 2.

6. Discussion, managerial implications and limitations


The first step to reduce the risk of failure in Six Sigma projects is selecting optimal projects
which have the most profits and the least expected risks. The current paper provides a
mathematical model to enhance decision quality in Six Sigma project selection. In fact,
unique excellence of suggested model in comparison to existing literature – according to
Table I – is that there is a complete flexible mathematical package to rank, categorize,
compare, improve and select the Six Sigma projects. Applying the specific DEA model is
remarkable itself, which joined to a pioneering procedure to use LDA to validity evaluation
of the results.
Downloaded by INSEAD At 18:59 13 July 2018 (PT)

Managers and decision makers should pay attention to current approach because
suggested model is able to provide their remarkable parameter such as calculating and
forecasting the productivity of the projects. Furthermore, the specific model is enough
flexible to analyze crisp data as well as imprecise ones. Moreover, the model presents
sensitivity analysis and benchmark the DMUs of the problem and is able to compare values
of input and output factors of DMUs after rudiment analysis.
The entire characteristics above made our suggested integrated model to a dominant
model in comparison of other presented models in the literature and justified why we
implement the following procedure.

Function Eigenvalue % of variance Cumulative % Canonical correlation


a
1 3.004 100.0 100.0 0.866
Table V.
a
Eigenvalues Note: One canonical discriminate function was used in the analysis

Test of function(s) Wilks’ Lambda Chi-square df Sig.


Table VI.
Wilks’ lambda 1 0.250 20.811 6 0.002

Predicted group membership Total


Project no. GROUP 1 2

Original
Count 1 4 0.0 4
2 1 15 16
% 1 100.0 0.0 100.0
Table VII. 2 6.3 93.8 100.0
Classification
resultsa Note: a95.0% of original grouped cases correctly classified
Project DEA LDA Discriminant Probabilities of Probabilities of
Selecting six
no. categorizing categorizing scores membership in Group 1 membership in Group 2 sigma project
1 2 2 0.80143 0.00023 0.99977
2 2 1 2.70817 0.99768 0.00232
3 2 2 1.16600 0.00005 0.99995
4 2 2 0.18955 0.00287 0.99713
5 2 2 1.40581 0.00002 0.99998
6 2 2 0.36621 0.02753 0.97247
7 1 1 3.25612 0.99976 0.00024
8 2 2 0.79452 0.00024 0.99976
9 2 2 1.46688 0.00002 0.99998
10 2 2 1.10074 0.00007 0.99993
11 2 2 0.89840 0.00016 0.99984
12 1 1 3.54314 0.99992 0.00008
13 2 2 2.04943 0.00000 1
14 2 2 1.34611 0.00002 0.99998
15 2 2 1.50738 0.00001 0.99999
Downloaded by INSEAD At 18:59 13 July 2018 (PT)

16 2 2 1.06450 0.00008 0.99992


17 1 1 3.25104 0.99975 0.00025 Table VIII.
18 2 2 1.07830 0.00007 0.99993 Categorizing and
19 1 1 3.10458 0.99954 0.00046 possible membership
20 2 2 1.36021 0.00002 0.99998 of projects

There is a considerable challenge for all methods and techniques of the literature that is
estimating exact values of the projects’ criteria. Mentioned limitation can be tackled by
using the imprecise numbers such as fuzzy numbers.

7. Conclusion
Several process improvement techniques have come and gone in the past five decades.
Although these techniques did improve quality and productivity of products manufactured
by some companies, many failed to get any benefit out of these quality improvement
programs. One of the main reasons for the failure of a process improvement initiative in
most cases can be identified as “wrong project selection”. Although it is claimed that the
impact of Six Sigma has been much higher than any of the previous quality improvement
programs, failure to select the right projects will also result in failure of Six Sigma
implementation. In this paper, a DEA-based project selection model was proposed for Six
Sigma project selection. The efficient DMUs (projects) based on proposed DEA model are
selected for implementation. Then we used LDA to examine the validity of obtained results
from proposed DEA model. The results show that the proposed model is a suitable tool for
selecting Six Sigma Projects. Our findings demonstrate that the selected projects by
suggested integrated DEA model are those confirmed by LDA. The future study will be
using the imprecise numbers such as fuzzy numbers to smooth the challenge of exact
forecasting the values of the projects’ criteria.

References
Abdalhkeim, F.A.F., Zakic, T. and Tornjanski, A. (2017), “Identification and selection of Six Sigma
projects”, Journal of Process Management – New Technologies, Vol. 5 No. 2, pp. 10-17.
Adams, C., Gupta, P. and Wilson, C. (2003), Six Sigma Deployment, Butterworth-Heinemann, Oxford.
IJLSS Adebanjo, D., Samaranayake, P., Mafakheri, F. and Laosirihongthong, T. (2016), “Prioritization of six-
sigma project selection: a resource-based view and institutional norms perspective”,
Benchmarking: An International Journal, Vol. 23 No. 7, pp. 1983-2003.
Ahadian, B. and Abadi, A.G.M. (2012), “Six sigma pilot project selections using an MCDM approach”,
Management Science and Engineering, Vol. 6 No. 1, pp. 34-43.
Altintas, M., Erginel, N. and Kucuk, G. (2016), “Determining the criteria and evaluating six sigma projects
via fuzzy ANP method in group decision”, IFAC-PapersOnLine, Vol. 49 No. 12, pp. 1850-1855.
Anbari, F.T. (2002), “Six sigma method and its applications in project management”, Proceedings of the
Project Management Institute Annual Seminar and Symposium, San Antonio, TX.
Anderson, P. and Peterson, N.C. (1993), “A procedure for ranking efficient units in data envelopment
analysis”, Management Sciences, Vol. 39 No. 10, pp. 1261-1264.
Antony, J. (2004), “Six sigma in the UK service organizations: results from a pilot survey”, Managerial
Auditing Journal, Vol. 19 No. 8, pp. 1006-1013.
Antony, J. and Fergusson, C. (2004), “Six sigma in a software industry: results from a pilot study”,
Managerial Auditing Journal, Vol. 19 No. 8, pp. 1025-1032.
Bajgier, S.M. and Hill, A.V. (1982), “An experimental comparison of statistical and linear programming
Downloaded by INSEAD At 18:59 13 July 2018 (PT)

approaches to the discriminant problem”, Decision Sciences, Vol. 13 No. 4, pp. 604-618.
Banker, R.D., Charnes, A. and Cooper, W.W. (1984), “Some models for estimation technical and scale in
efficiencies in data envelopment analysis”, Management Science, Vol. 30 No. 9, pp. 1078-1092.
Banuelas, R., Tennant, C., Tuersley, I. and Tang, S. (2006), “Selection of Six Sigma projects in UK”, The
TQM Magazine, Vol. 18 No. 5, pp. 514-527.
Breyfogle, F., Cupello, J. and Meadws, B. (2001), Managing Six Sigma, Wiley Inter-Science, New York, NY.
Brun, A. (2011), “Critical success factors of Six Sigma implementations in Italian companies”,
International Journal of Production Economics, Vol. 131 No. 1, pp. 158-164.
Charnes, A., Cooper, W.W. and Rhodes, E. (1978), “Measuring the efficiency of decision making units”,
European Journal of Operational Research, Vol. 2 No. 6, pp. 429-444.
Chen, N., Xu, L. and Chen, Z. (2017), “Environmental efficiency analysis of the yangtze river economic
zone using super efficiency data envelopment analysis (SEDEA) and tobit models”, Energy,
Vol. 134, pp. 659-671.
Cook, W.D. and Seiford, L.M. (2009), “Data envelopment analysis (DEA) – thirty years on”, European
Journal of Operational Research, Vol. 192 No. 1, pp. 1-17.
De Feo, J.A. and Barnard, W. (2004), Juran Institute’s Six Sigma Breakthrough and Beyond, Quality
Performance Methods, McGraw-Hill, New York, NY.
Dhiraj, K. and Deepak, K. (2014), “A review of Six Sigma approach: methodology, obstacles and
benefits”, Global Journal of Engineering, Design & Technology, Vol. 3 No. 4, pp. 1-5.
Dusharme, D. (2003), “Six sigma survey: big success . . . what about other 98 percent?”, (February),
Quality Digest.
Esmaeilzadeh, A. and Hadi-Vencheh, A. (2015), “A new method for complete ranking of DMUs”,
Optimization, Vol. 64 No. 5, pp. 1177-1193.
Farrel, M.J. (1957), “The measurement of productive efficiency”, Journal of Royal Statistics Society:
Series A, Vol. 120 No. 3, pp. 253-290.
Freed, N. and Glover, F. (1986), “Resolving certain difficulties and improving the classification power of
LP discriminant analysis formulations”, Decision Sciences, Vol. 17 No. 4, pp. 589-595.
Fundin, A. and Cronemyr, P. (2003), “Use customer feedback to choose Six Sigma projects”, ASQ Six
Sigma Forum Magazine, Vol. 3 No. 1, pp. 17-22.
Glover, F. (1990), “Improved linear programming models for discriminant analysis”, Decision Sciences,
Vol. 21 No. 4, pp. 771-785.
Glover, F., Keene, S. and Duea, B. (1988), “A new class of models for the discriminant problem”, Decision Selecting six
Sciences, Vol. 19 No. 2, pp. 269-280.
sigma project
Goh, T.N. (2002), “A strategic assessment of Six Sigma”, Quality and Reliability Engineering
International, Vol. 18 No. 5, pp. 403-410.
Goldstein, M.D. (2001), “Six sigma program success factors”, ASQ Six Sigma Forum Magazine, Vol. 1
No. 1, pp. 36-45.
Hahn, G.J., Hill, W.J., Hoerl, R.W. and Zinkgraf, S.A. (1999), “The impact of Six Sigma improvement – a
glimpse into the future of statistics”, The American Statistician, Vol. 53 No. 3, pp. 208-215.
Harry, M.J. (1998), “Six sigma: a breakthrough strategy for profitability”, Quality Progress, Vol. 31
No. 5, pp. 60-64.
Harry, M.J. and Schroeder, R. (2000), Six Sigma: The Breakthrough Management Strategy
Revolutionising the World’s Top Corporations, Currency Publishers, Sydney.
Hatami-Marbinia, A., Emrouznejad, A. and Tavana, M. (2011), “A taxonomy and review of the fuzzy
data envelopment analysis literature: two decades in the making”, European Journal of
Operational Research, Vol. 214 No. 3, pp. 457-472.
Hendricks, C.A. and Kelbaugh, R.L. (1998), “Implementing Six Sigma at GE”, The Journal of Quality
Downloaded by INSEAD At 18:59 13 July 2018 (PT)

and Participation, Vol. 21 No. 4, pp. 43-48.


Ho, S., Xie, M. and Goh, T. (2006), “Adopting Six Sigma in higher education: some issues and challenges”,
International Journal of Six Sigma and Competitive Advantage, Vol. 2 No. 4, pp. 335-352.
Holmes, M.C., Jenicke, L.O. and Hempel, J.L. (2015), “A framework for Six Sigma project selection in
higher educational institutions, using a weighted scorecard approach”, Quality Assurance in
Education, Vol. 23 No. 1, pp. 30-46.
Hosseini, S.A. and Abedi, M. (2015), “Data envelopment analysis: a knowledge-driven method for
mineral prospectivity mapping”, Computers & Geosciences, Vol. 82, pp. 111-119.
Hou, S. and Riley, C.B. (2015), “Is uncorrelated linear discriminant analysis really a new method?”,
Chemometrics and Intelligent Laboratory Systems, Vol. 142, pp. 49-53.
Kalashnikov, V., Benita, F., López-Ramos, F. and Hernández-Luna, A. (2017), “Bi-objective project portfolio
selection in Lean Six Sigma”, International Journal of Production Economics, Vol. 186, pp. 81-88.
Kelly, M. (2002), “Three steps to project selection”, ASQ Six Sigma Forum Magazine, Vol. 2 No. 1, pp. 29-33.
Koehler, G.J. (1989), “Characterization of unacceptable solutions in LP discriminant analysis”, Decision
Sciences, Vol. 20 No. 2, pp. 239-257.
Krupar, J. (2003), “Yes Six Sigma can work for financial institutions”, ABA Banking Journal, Vol. 95
No. 9, pp. 93-94.
Kumar, D.U., Crocker, J., Chitra, T. and Saranga, H. (2006), Reliability and Six Sigma, Springer, Berlin.
Kumar, D.U., Saranga, H., Ramírez-Márquez J.E. and Nowicki, D. (2007), “Six sigma project selecting
using data envelopment analysis”, The TQM Magazine, Vol. 19 No. 5, pp. 419-441.
Kwak, Y.H. and Anbari, F.T. (2004), “Benefits, obstacles and future of Six Sigma”, Technovation: The
International Journal of Technological Innovation, Entrepreneurship and Technology
Management, Vol. 26 Nos 5/6, pp. 708-715.
Lanyon, S. (2003), “At raytheon Six Sigma works, too, to improve HR management processes”, Journal
of Organizational Excellence, Vol. 22 No. 4, pp. 29-42.
Larson, A. (2003), Demystifying Six Sigma, American Management Association, New York, NY.
Larson, E., W. and Gray, C., F. (2011), Project Management: The Managerial Process, McGraw-Hill,
New York, NY.
Mc Adam, R. and Evans, A. (2004), “The organizational contextual factors affecting the implementation
of Six Sigma in a high technology mass-manufacturing environment”, International Journal of
Six Sigma and Competitive Advantage, Vol. 1 No. 1, pp. 29-43.
IJLSS Markowski, E.P. and Markowski, C.A. (1985), “Some difficulties and improvements in applying linear
programming formulations to the discriminant problem”, Decision Sciences, Vol. 16 No. 3, pp. 237-247.
Marques, P., A., Saraiva, P., M., Requeijo, J., G. and Guerreiro, F., F. (2014), “Six sigma life cycle”, in
Henriques, E., Pecas, P., Arlindo, S. (Eds), Technology and Manufacturing Process Selection,
Springer-Verlag, London.
Marzagalo, D.S.L. and Carvalho, M.M. (2016), “Critical success factors for Six Sigma projects”,
International Journal of Project Management, Vol. 34 No. 8, pp. 1505-1518.
Mellat Parast, M. (2011), “The effect of Six Sigma projects on innovation and firm performance”,
International Journal of Project Management, Vol. 29 No. 1, pp. 45-55.
Moorman, D.W. (2005), “On the quest for Six Sigma”, American Journal of Surgery, Vol. 189 No. 3,
pp. 253-258.
Nourpanah, P., Ansary, N., Nourpanah, R. and Amini, S. (2011), “Selection six sigma project use gahp-
lp”, Australian Journal of Basic and Applied Sciences, Vol. 5 No. 12, pp. 1917-1922.
Ortiz, M.A., Felizolla, H.A. and Isaza, S.N. (2015), “A contrast between DEMATEL-ANP and ANP
methods for six sigma project selection: a case study in healthcare industry”, BMC Medical
Informatics and Decision Making, p. 1.
Downloaded by INSEAD At 18:59 13 July 2018 (PT)

Pande, P., Neuman, R. and Cavanagh, R. (2000), The Six Sigma Way: How GE, Motorola and Other Top
Companies Are Honing Their Performance, McGraw-Hill, New York, NY.
Pyzdek, T. (2000), “Selecting Six Sigma projects”, (September), Quality Digest.
Pyzdek, T. (2003), The Six Sigma Project Planner, McGraw-Hill, New York, NY.
Ragsdale, C.T. and Stam, A. (1991), “Mathematical programming formulations for the discriminant
problem: an old dog does new tricks”, Decision Sciences, Vol. 22 No. 2, pp. 296-307.
Ray, S. and Das, P. (2010), “Six sigma project selection methodology”, International Journal of Lean Six
Sigma, Vol. 1 No. 4, pp. 293-309.
Robinson, B. (2005), “Build a management system based on Six Sigma”, ASQ Six Sigma Forum
Magazine, Vol. 5 No. 1, pp. 28-34.
Rosenthal, E.C. and Weiss, H.J. (2017), “A data envelopment analysis approach for ranking journals”,
Omega, Vol. 70, pp. 135-147.
Safo, S.E. and Ahn, J. (2016), “General sparse multi-class linear discriminant analysis”, Computational
Statistics & Data Analysis, Vol. 99, pp. 81-90.
Sefair, J.A., Méndez, C.Y., Babat, O., Medaglia, A.L. and Zuluaga, L.F. (2017), “Linear solution schemes
for mean-semivariance project portfolio selection problems: an application in the oil and gas
industry”, Omega, Vol. 68, pp. 39-48.
Shabani, A., Torabipour, S.M.R., Farzipoor Saen, R. and Khodakarami, M. (2015), “Distinctive data
envelopment analysis model for evaluating global environment performance”, Applied
Mathematical Modelling, Vol. 39 No. 15, pp. 4385-4404.
Shao, G. and Sang, N. (2017), “Regularized max-min linear discriminant analysis”, Pattern Recognition,
Vol. 66, pp. 353-363.
Shanmugaraja, M., Nataraj, M. and Gunasekaran, N. (2011), “Six sigma project selection via quality
function deployment”, International Journal of Productivity and Quality Management, Vol. 10
No. 1, pp. 85-111.
Sharma, S. and Chetiya, A.R. (2010), “Six sigma project selection: an analysis of responsible factors”,
International Journal of Lean Six Sigma, Vol. 1 No. 4, pp. 280-292.
Snee, R.D. (2001), “Dealing with the Achilles’ Heel of Six Sigma initiatives – project selection”, Quality
Progress, Vol. 34 No. 3, pp. 66-72.
Stam, A. and Ragsdale, C.T. (1992), “On the classification gap in mathematical-programming-based
approaches to the discriminant problem”, Navel Research Logistics, Vol. 39 No. 4, pp. 545-559.
Tang, L.C., Goh, T.N., Lam, S.W. and Zhang, C.W. (2007), “Fortification of Six Sigma: expanding the Selecting six
DMAIC toolset”, Quality and Reliability Engineering International, Vol. 23 No. 1, pp. 3-18.
sigma project
Vinodh, S. and Swarnakar, V. (2015), “Lean Six Sigma project selection using hybrid approach based on
fuzzy DEMATEL–ANP–TOPSIS”, International Journal of Lean Six Sigma, Vol. 6 No. 4, pp. 313-338.
Wang, Z., Shao, Y.H., Bai, L., Li, C.N. and Deng, N.Y. (2016), “MBLDA: a novel multiple between-class
linear discriminant analysis”, Information Sciences, Vol. 369, pp. 199-220.
Wu, Y., Hu, Y., Xiao, X. and Mao, C. (2016), “Efficiency assessment of wind farms in China using two-
stage data envelopment analysis”, Energy Conversion and Management, Vol. 123, pp. 46-55.
Xu, J. (2017), “A weighted linear discriminant analysis framework for multi-label feature extraction”,
Neurocomputing, available at: https://doi.org/10.1016/j.neucom.2017.05.008
Yang, K., Yeh, T., Pai, F. and Yang, C.-C. (2008), “The analysis of the implementation status of Six
Sigma: an empirical study in Taiwan”, International Journal of Six Sigma and Competitive
Advantage, Vol. 4 No. 1, pp. 60-80.
Yang, W. and Wu, H. (2014), “Regularized complete linear discriminant analysis”, Neurocomputing,
Vol. 137, pp. 185-191.
Yousefi, A. (2009), “Letter to the editor”, Omega, Vol. 37 No. 1, p. 252.
Downloaded by INSEAD At 18:59 13 July 2018 (PT)

Yousefi, A. (2017), “EPEDC ELEGAME: an integrated model using gamification and data mining to
revolutionise utilities’ functions”, Cired – Open Access Proceedings Journal, Vol. 2017 No. 1,
pp. 2740-2743 doi: 10.1049/oap-cired.2017.0874.
Yousefi, A. and Aqamohammadi, A.R. (2014), “A new DEA model for Six Sigma project selecting: case
study on Esfahan province electricity distribution co. (EPEDC)” Proceeding Of 2014 IEEE
International Conference on Industrial Engineering and Engineering Management (IEEM 2014),
Bandar Sunway, (9-12 December 2014), pp. 627-631, doi: 10.1109/IEEM.2014.7058714.
Yousefi, A. and Hadi-Vencheh, A. (2016), “Selecting Six Sigma projects: MCDM or DEA?”, Journal of
Modelling in Management, Vol. 11 No. 1, pp. 309-325.
Zimmerman, J.P. and Weiss, J. (2005), “Six sigma’s seven deadly sins”, Quality, Vol. 44 No. 1, pp. 62-66.

Further reading
Merve, A., Erginel, N. and Kucuk, G. (2016), “Determining the criteria and evaluating Six Sigma projects
via fuzzy ANP method in group decision”, IFAC- Papers on Line, Vol. 49 No. 12, pp. 1850-1855.
Seiford, L.M. (1996), “Data envelopment analysis: the evaluation of the state of the art (1978-1995)”,
Journal of Productivity Analysis, Vol. 7 Nos 2/3, pp. 99-137.
Vyacheslav, K., Benita, F., Lopez-Ramos, F. and Hernández-Luna, A. (2017), “Bi-objective project portfolio
selection in lean Six Sigma”, International Journal of Production Economics, Vol. 186, pp. 81-88.
Wai-Ki, C., Chu, D., Liao, L.Z. and Wang, X. (2012), “Regularized orthogonal linear discriminant
analysis”, Pattern Recognition, Vol. 45 No. 7, pp. 2719-2732.
Zhecheng, J., Wang, G., Zhang, S. and Qiu, C. (2017), “Building Tianjin driving cycle based on linear
discriminant analysis”, Transportation Research Part D: Transport and Environment, Vol. 53,
pp. 78-87.

Corresponding author
A. Hadi-Vencheh can be contacted at: ahadi@khuisf.ac.ir

For instructions on how to order reprints of this article, please visit our website:
www.emeraldgrouppublishing.com/licensing/reprints.htm
Or contact us for further details: permissions@emeraldinsight.com

You might also like