Perceptual Mapping Using Principal Component Analysis in a Public Sector Passenger Bus Transport Company a Case Study

7/30/2019 Perceptual Mapping Using Principal Component Analysis in a Public Sector Passenger Bus Transport Company a Ca

1/16

International Journal of Production Technology and Management (IJPTM), ISSN 0976 6383

(Print), ISSN 0976 6391 (Online) Volume 3, Issue 1, January- December (2012), IAEME

27

PERCEPTUAL MAPPING USINGPRINCIPAL COMPONENTANALYSIS IN A PUBLIC SECTOR PASSENGER BUS TRANSPORT

COMPANY: A CASE STUDY

M.Vetrivel Sezhian *1, C.Muralidharan

2, T.Nambirajan

3, S.G.Deshmukh

4

1Professor, Department of Mechanical Engineering,Dr. Pauls Engineering College, Villupuram - 605109, TamilNadu, India,

[email protected], Department of Manufacturing Engineering,

Annamalai University, Chidambaram - 608002, TamilNadu, India,

[email protected] Professor, Department of Management Studies,

Pondicherry University, Pondicherry - 605014, Pondicherry, India,[email protected]

4Professor, Department of Mechanical Engineering,

Indian Institute of Technology Delhi, New Delhi - 110016, India,

[email protected]* - corresponding author

ABSTRACT

This study aims at evaluating the customer expectations in a public sector

passenger transport company which is a crucial sector for transportation of people in

developing countries like India. A questionnaire containing eighteen qualitycharacteristics was administered to various customers of three bus depots of one division

of a state road transport undertaking (SRTU) in south India. Two quality dimensions, viz.

customer expectations and company responsibilities, have been identified based onprincipal component analysis. Subsequently, perceptual mapping is used to identify their

strengths and weaknesses of three depots. The findings would help prioritise differentparameters and also provide guidelines to managers to focus on or to improve.

Keywords: Public bus transport, Customer expectations, Perceptual mapping, Principal

component analysis.

INTERNATIONAL JOURNAL OF PRODUCTION TECHNOLOGY AND

MANAGEMENT (IJPTM)

ISSN 0976- 6383 (Print)

ISSN 0976 - 6391 (Online)

Volume 3, Issue 1, January-December (2012), pp. 27-42

IAEME: www.iaeme.com/ijptm.html

Journal Impact Factor (2012): 1.5910 (Calculated by GISI)www.jifactor.com

IJPTM

I A E M E


2/16



28

1. INTRODUCTION

In several Asian countries, the public sector occupies an important position in the

economy. Bus is very popular sector of transportation in India because of its low cost andhas made a great and significant contribution to the national economy. Transport sector

accounts for a share of 6.4 per cent in Indias Gross Domestic Product (GDP). Roadtransport has emerged as the dominant segment in Indias transportation sector with ashare of 5.4 per cent in Indias GDP. Road transport demand is expected to grow by

around 10% per annum in the backdrop of a targeted annual GDP growth of 9% during

the Eleventh Five Year Plan. In India, the state road transport undertaking (STRUs)which has 58 members, which forms the backbone of mobility for urban and rural

population across the country is operating over 1,15,000 buses, serving more than 65

million passengers a day and also providing employment to 0.8 million people (ASTRU).Effectiveness is the extent to which outputs of service providers meet the

objectives set for them. Efficiency is the success with which an organization uses its

resources to produce outputs that is the degree to which the observed use of resources

to produce outputs of a given quality matches the optimal use of resources to produceoutputs of a given quality. This can be assessed in terms of technical, allocative, cost and

dynamic efficiency. Improving the performance of an organizational unit relies on both

efficiency and effectiveness. A government service provider might increase its measuredefficiency at the expense of the effectiveness of its service. For example, a state transport

undertaking might reduce the inputs used like fleet size, cost, bus or day to carry the same

number of passengers. This could increase the apparent efficiency of that state transportundertaking but reduce its effectiveness in providing satisfactory outcomes forpassengers. Therefore, it is important to develop effectiveness indicators also (Bhagavath,

2006).

However as it is serving a large amount of passengers, the quality of service is

the main concern issue. In the road transport sector in India, liberalization of theautomobile industry in parallel to a rapid increase in per capita incomes has led to a shift

towards personal vehicles. The share of public transport, on the other hand, has declinedover time. The economy is now being constrained by the increasing number of vehicles

causing congestion, and thus slower speeds on roads. Transport infrastructure is

recognized as being the critical constraint here (Ramanathan and Parikh, 1999). Efficient

and optimal utilization of the available transport infrastructure would require meetingmobility needs through a greater share of public transport (Planning Commission, 2002).

The Road passenger transport in India is operated partly by public sector and largely by

private sector comprising about 29% and 71% respectively of the total buses. The publicsector should strive to create customer value just as the private sector does (Rahman and

Rahman 2009). Hence, understanding the customer expectations is of immenseimportance.

Over the last few years, companies have gradually focused on service quality and

customer satisfaction. This strategy is very profitable for both companies and customers,

particularly for transit agencies and passengers. An improvement in the service qualitydelivered will attract further users. This would resolve many problems such as helping to

reduce traffic congestion, air and noise pollution, and energy consumption because


3/16



29

individual transport would be used lesser (Eboli and Mazulla 2007). The development of

techniques for customer satisfaction analysis is necessary as they allow the critical

aspects of the supplied services to be identified and customer satisfaction to be increased.

The measure of how products and services supplied by a company meet or surpass thecustomers expectation is seen as a vital performance indicator. In a competitive market

place where businesses compete for customers, it is seen as a key indicator that providesan indication of how successful the organization is at providing services. The level ofexpectation can also vary depending on other factors, such as other products against

which they can compare the organization's products. The usual measures of getting this

involve a survey with a set of statements using a Likerts scale (1 to 5). The customer isasked to evaluate each statement in terms of their perception and expectation of

performance of the service being measured (Wikipedia). Initially when a customer survey

is to be conducted a questionnaire has to be developed. The identification of parametersto be considered is to be listed. The various inputs could be obtained for the same through

brainstorming. The experience and expertise of both the experts from the concerned

sector or field and / or academicians would form a part such an exercise for evolving an

exhaustive and sufficient list of questions or parameters (Chidambaranathan et al 2009).The list of questions evolved may initially be administered to a group of customers to

ascertain if all of the parameters are important or necessary.

The main objectives of the paper are to:

1. Identify the customer characteristics for the public bus transport and validate the

factors responsible for service quality.2. Identify the position of the three bus depots and as a result find their strengths and

weakness.

3. Identify best performing depot of the bus company.

This paper is organised as follows: The literature review of factor analysis namely

principal component analysis (PCA) and perceptual mapping (PM) has been presented,followed by the case study where a discussion about the customer characteristics and its

validation through PCA is given. Also, the data obtained from PCA factor orcomponent score is used to plot a PM to identify the strength and weaknesses of depots

and to determine the best performing depot are provided. Then the results and discussion

are presented and finally the conclusions are drawn.

2. LITERATURE REVIEW

2.1 Principal Component Analysis

Principal component analysis (PCA) is one of the most widely used multivariatetechnique and popular ranking technique used in modern data analysis - in diverse fieldsfrom neuroscience to computer graphics - because it is a simple, non-parametric method

for extracting relevant information from confusing data sets. With minimal effort PCA

provides a method to reduce a complex data set to a lower dimension to reveal thesometimes hidden, implied structures that often underlie it (Jonathon, 2009). It involves a

mathematical procedure that transforms a number of correlated variables into a lesser


4/16



30

number of uncorrelated variables called principal components (Petroni and Braggia,

2000). Even though the objective of PCA may be to reduce the number of variables of a

dataset it retains most of the original variability in the data. The first principal component

accounts for as much of the data variability as possible and succeeding componentsaccount for as much of the remaining variability as possible (Hair et al, 2006).

The review of the literature on factor Analysis which has been carried out in thetransport sector over the last few years is presented herewith. Rahaman and Rahaman(2009) focussed on the railway transportation sector to develop a model defining the

relationship between overall satisfaction and twenty service-quality attributes. Using

PCA they have found that overall satisfaction depended on eight service qualityattributes. Kolanovic et al (2008) have presented the methods of choosing the possible

attributes affecting the perception of the port service quality. They have used PCA to

reduce number of the port service quality attributes and were grouped in two dimensionsof the port service quality: reliability and competence.

Zoe, (2006) has conducted a study that attempts to contribute to the knowledge of

how customer satisfaction, loyalty and commitment related to each other in the Greek

context, based on responses collected from twenty service providers in four servicesectors including transportation. Both factor and reliability analyses provided the

relationship between customer loyalty and satisfaction as well as between commitment

and customer loyalty. Ching et al., (2009) have used PCA to identify crucial resourcesand logistics service capabilities in container shipping services. Furthermore, factor

analysis was employed to identify the crucial logistics service capability in container

shipping service firms. Lai, (2010) explores the relationships between passengerbehavioural intentions and the various factors that affect them in a new public transitcompany in Taiwan. The data analysis was conducted exploratory factor analyses using

principal component with varimax rotation technique to examine construct

dimensionalities of both public transport involvement and service quality. Vetrivel et al.,

(2011) have applied factor analysis for the bus transport company for data reduction andranking of bus depots.

In all the above discussion on the factor analysis literature the focus has been ondata reduction, grouping and identifying the key factors that need immediate attention for

improvement in service quality, etc. In the present study, it is proposed to employ

principal component analysis (PCA) to not only extract the principal components and

group the (remaining) factors but also to identify the best performing depot of the buscompany.

2.2 Perceptual Mapping (PM)

Perceptual map is atechnique in which consumer's views about an alternative areplotted or mapped on a chart. It is a spatial representation in which competingalternatives and attributes are plotted in a Euclidean space. The respondents are asked

questions about their experience with the alternative in terms of its characteristics. These

qualitative answers are transferred to a chart called a perceptual map using a suitablescale (such as the Likerts scale), and the results are employed in improving the product

or in developing a new one (BuisnessDictionary.com). Mapping methods used can be


5/16



31

broadly classified as shown in Figure 1. The various methods are discussed in (Lilien et

al., (2007); Hair et al., (2006); Lattin et al. (2003)). There are two approaches to

perceptual mapping: attribute based and non-attribute based. They differ in assumptions

they employ, the perspective taken, and the input data used. The respondents evaluatedepots on the basis of attributes perceived by them in making the decision.PM has been

done using Factor analysis and Discriminant analysis.The following is the brief review of perceptual mapping using factor analysis.

Luyang and Weidong, (2010) used have studied the perception of students on some

attributes using attribute based perceptual mapping; Vassiliadis, (2006), has used it to

present the proper tourist product characteristics and market opportunities by therecipients of the tourist market; and Esseghaier, (2010), presents how respondents have

rated six car brands on six attributes. Joel Davis, (2006) has shown how consumers use

various dimensions to evaluate brands and products; Kohli and Leuthesser, (1993),present an overview of perceptual mapping and compare three widely used techniques --

factor analysis, discriminant analysis, and multidimensional scaling; Le Van Huy et al.,

(2007), have shown how an organization can recognize customers demand with their

brand or competitors and build competitive advantage on brand position in customersmind.

In this case study the perceptual mapping has been used with principal component

analysis to ascertain three bus depots and of a public sector bus company in south Indiaand assess their the strengths and weakness.

3. RESEARCH METHODOLOGY: A CASE STUDY

In the case study a state road transport undertaking (SRTU) located in south India,

operating passenger buses has been chosen. It is one of the leading public sector bustransport corporations generating consistent returns as well rendering excellent service

over the years has a fleet strength of about 1350 buses. The present case study has beenconducted in three bus depots of one division of the SRTU. There are various divisionssuch as Chennai, Villupuram, Kumbakonam, Salem, Coimbatore, Madurai, and

Villupuram. The present case study has been conducted in three bus depots of the

Villupuram division. The fleet strength of these three depots is 98, 95 and 94

respectively. These depots each employ around 180 bus crew members, 30 maintenancestaff, 10 managerial staff, and 15 -18 administrative staff. As far as the average number

of passengers travelling every day is approximately one and a half lakh for depot-I,

around a lakh and twenty two thousand for depot-II and less than one lakh passengers fordepot-III.

The data was collected through a survey using a questionnaire developed by a

group of three depot managers, transport officials of the SRTU and academicians throughthe brainstorming methodology. The questionnaire consists of a set of eighteen questions

on the customer characteristics as shown in Table 1. The questionnaire was used toenumerate the responses by 150 passengers through the interview method on these

eighteen criteria through the interview method. They were asked to rate the same on a 1-5Likerts scale (Very Good (5), Good (4), Fair (3), Poor (2), And Very Poor (1)).


6/16



32

Ahead of assessing the data collected, a reliability analysis was conducted, which

gave a Cronbachs value of 0.94 (min value of is 0.7, Nunnally (1978)). This proves

the ability of the survey instrument to produce consistent results. Subsequently based on

the passenger responses received, principal component analysis (PCA) is to be employedto extract principal components (Eigen values >1). The Bartlett test of sphericity and the

Kaiser-Meyer-Olkin measure of sampling were used to validate the use of PCA. Thecomponents with factor loadings greater than 0.5 (Kannan, 2002) are extracted and theones with values lesser than 0.5 are omitted. The resulting components would help in

grouping the remaining criteria. The components are then suitably named pertaining to

the concerned grouped entities. Then using the factor score (obtained from factor orcomponent coefficient matrix of the factor analysis) the perceptual map is developed

(Esseghaier, 2010). This map helps to ascertain the position of the depots with respect to

the two factors identified. If the arrow representing a depot is positioned nearer to eitherof the axis represented by one of the factors, it is an area of strength or better

performance and farther away means sign of weakness or not up to expectation

performance. Thus the best performing depot could be identified.

The flow chart for the research methodology is presented in Figure 2. The looparound shows that this processes needs to be periodically (every 2 to 3 years) repeated to

review the relevance of the various criteria and add new ones if found necessary. The

principal component analysis (PCA) has been done using the software SPSS 16.0. Theperceptual mapping has been developed using the MS Excel.

4. RESULTS AND DISCUSSION

Generally, it only makes sense to use principal component analysis when the data

are not independent. A method of determining appropriateness of factor analysis is

Bartlett test of sphericity and Kaiser-Meyer-Olkin test. Kaiser-Meyer-Olkin measure of

sampling adequacy tests whether the partial correlations among variables are small ornot.For the present study the Bartlett test of sphericity had returned a value of

2(153) =

4823. Thus, we can conclude that according to Bartletts test, the correlation matrix is notan identity matrix.The Kaiser-Meyer-Olkin test of sphericity has a KMO index value of

0.953 (greater than 0.5, Hair et al, 2006). The significance value for the same (.000) was

less than 0.001, and hence significant. This validates the use of PCA.Thus KMO Statistic

suggests that we have sufficient sample size relative to the number of attributes in ourscale. Hence, the KMO statistic and Bartletts test of Sphericity suggest that the

correlation matrix is factorable and that there are some underlying factor/dimensions that

may explain the variance of 18 items (Parikshat, 2010).Then based on the received responses, PCA has been employed to extracted two

components with Eigen values > 1. Varimax rotation most popular orthogonal factorrotation method is used, which tries to achieve simple structure by focusing on thecolumns of the factor loading matrix and is given in Table 2 (Hair et al., 2006; Lattin et

al., 2003). These two components accounted for a total variance of 58.73%. Variance (the

dispersion of values around a mean) is a measure of the information content of anattribute. The larger the variance, the higher is the information content. This can be

observed in Table 3 and the same be seen in the scree plot. Scree plot is a plot of Eigen


7/16



33

values against the number of factors in the order of extraction is shown in Figure 3. In the

scree plot we look for an elbow in the curve that is, a point after which the remaining

Eigen values decline in approximately linear fashion and retain only those two

components that are above the elbow (Lattin et al., 2003). Communality which is the totalamount of variance a variable shares with all other variables being considered. The

communalities may be viewed as whether the variables meet acceptable levels ofexplanation. A small communality figure shows that the factors taken together do notaccount for the variable to an appreciable extent. On the contrary, large communality

figure is an indication that much of the variable is accounted for by the factors. The

communality value should be greater than 0.5 (Hair et al., 2006). In Table 2, all theeighteen communality values were greater than 0.5 and so significant for further study.

Also, from the analysis, all the factor loadings were greater than 0.5 (Kannan, 2002) and

therefore all the eighteen sub-criteria are significant.The two components extracted with Eigen values > 1, have been designated as

customer expectations and company responsibilities as seen in Table 4. Customer

expectations are the factors which gives details of the facilities that the passengers expect

inside the bus to make the journey comfortable. Company responsibilities are the factorswhich gives details of what responsibilities the passengers expect of the bus company.

The customer expectations factors includes bus punctuality (C11), seat comfort

(C12), cleanliness (C13), lighting & entertainment (C14), new fleet addition (C15), seatingfor handicapped (C16), seating for elderly (C17), issue of proper ticket (C18), in-time issue

of ticket (C19), and issue of proper change (C110) as shown in Table 4. Earlier in the

analysis these factors were known by Q1, Q4, Q5, Q6, Q8, Q10, Q11, Q16, Q17 and Q18respectively as in Table 1.

The company responsibilities factors includes stopping the bus at correct place

(C21), backup service during breakdown (C22), provision for luggage (C23), obey traffic

rules (C24), first aid facilities (C25), driver behaviour (C26), conductor behaviour (C27) and

information to passengers (C28) as shown in Table 4. Earlier in the analysis these factorswere known by Q2, Q3, Q7, Q9, Q12, Q13, Q14 and Q15 respectively as in Table 1.

Perceptual Map

The perceptual map (Figure 4) shows the positioning of the three depots which

have been dully labelled. A vector on the map (shown by a line segment with an arrow)indicates both magnitude and direction in the Euclidean space. Vectors are used to

geometrically denote attributes of the perceptual maps. The length of the vectors

indicates how well or decisively the attributes can distinguish between the products. Along vector indicates that the attribute is decisive in consumers minds. The further the

product is from the centre of the map the more decisive is its differentiation based on thatattribute. The vector pointing in the opposite gives low degree of association. The size ofthe angle between the vectors also gives important information. A narrow angle indicates

that the attributes are closely related since the correlation between them is high. The axes

of the map are special set of vectors suggesting the underlying dimensions that bestcharacterize how customers differentiate between alternatives. Many similar methods


8/16



34

exist (Sinclair, and Stalling, 1990; Bijmolt, and Wedel, 1999; Kohli, and Leuthesser,

1993).

The mean values of customer responses for the eighteen attributes of the three

depots are presented in Table 4. The factor or component coefficients obtained for theeighteen attributes from the principal component analysis is presented in last two

columns of Table 4. The factor or component score is obtained by multiplying thecoefficients with the respective mean values of the various attributes for each depot asfollows:

Factor score = (coefficient of attribute * mean of the attribute)

Factor 1 = 0.178 C11+0.155 C12+0.261 C13+0.197 C14+0.199 C15+0.169 C16+0.200

C17+0.059 C18+0.042 C19+0.103 C110-0.121 C21-0.063 C22-0.014 C23-0.011 C24-0.152C25-0.165 C26-0.070 C27-0.092 C28

Factor 2 = -0.080 C11-0.049 C12-0.186 C13-0.110 C14-0.102 C15-0.068 C16-0.111

C17+0.055 C18+0.083 C19+0.017 C110+0.245 C21+0.194 C22+0.136 C23+0.129 C24+0.270C25+0.285 C26+0.194 C27+0.214 C28

Loadings that have high absolute value (high absolute values of correlations)make interpretation easy. In a perceptual map the factor-loading matrix is represented

visually as attribute vectors, where correlation between any attribute and a factor is equal

to the cosine of the angle between that attribute vector and the corresponding factor(Lilien et al., 2007).

The factor scores thus calculated using the above equations for the three depots

are displayed in Table 5. The factors F1 and F2 are taken as the two axes in a graph

(Figure 4). Figure 4 is a perceptual map derived from factor analysis, where the length of

each attribute vector indicates the proportion of the variance of that attribute recovered bythe map. The factors may be rotated to aid interpretation (Table 2), forcing attributes to

have either big or small cosines with the transformed factors. The result is that a set ofattributes tends to line up closely with each factor. In this way, attributes tend to be

closely aligned with a single factor. We can then better identify the attributes most

closely associated with the transformed factors. Although rotation changes the variance

explained by each factor, it does not affect the total variance explained by the set ofretained factors. To further aid interpretation, we can draw each attribute vector on the

map with a length that is proportional to the variance of that attribute explained by the

retained factors (Lilien, et al., 2007).Using the values of factors in Table 5 as X and Y axis components, a scatter

graph is presented and the three points are obtained for the three depots. Vectors aredrawn to these as seen in perceptual map. The length of the vector representing depot-1 islonger of the three followed by depot-2 and depot-3. The vector for depot-1 is nearer

(narrower angle) to F1, i.e., customer expectations factor whereas vector for depot-2 is

nearer (narrower angle) to F2, i.e., company responsibilities factor. The vector for depot-3 is farther from both F1 and F2. It is apparent that the depot-1 is delivering better in the


9/16



35

customer expectation factor whereas depot-2 is faring better in company responsibility

factor. As such depot-3 is better in none.

5. CONCLUSIONSThis paper has presented perceptual mapping using principal component analysis

for a public sector bus transport company in south India, aimed at helping the companymanagement to formulate their strategies. Todays world is there is thrust forprivatisation of government owned sectors; particularly the public sector bus companies

in India face stiff competition from the private sector. It therefore becomes relevant that

they formulate policies and strategies to suit the needs of the situation. This present studybecomes relevant in getting to know the customers or passengers perspective of the

quality of service rendered. The advantages of factor analysis are that both subjective and

objective attributes can be used.Here eighteen service criteria that were taken up for study as customer

characteristics have all been found significant for further analysis by the PCA. The

managerial implication of this methodology is that it has given a clear insight into the

customer preferences and perspective. The eighteen criteria enlisted in this work have allfound importance from the passengers i.e., to the facilities and comfort as well as the

responsibilities which the company ought to take up. Subsequently, when the perceptual

mapping using PCA was done it clearly showed the following:

i. Depot-1 is performing better in customer expectations (10

attributes/criteria).ii. Depot-2 is performing better in company responsibilities (8

attributes/criteria).

iii. Depot-3 is performing better in none.

We can hence conclude that depot-1 is the best performing followed by depot-2and Depot-3 respectively in that order. It is also apparent that not only depot-3 needs to

improve considerably but the other two as well have show to improvements as well. Theentire above mentioned are the strengths of the depots respectively and it is to be noted

that the others factors (a group of criteria / attributes) are weak or grey areas for these

depots. This gives a clear indicator to the management that it needs to come up withaction plans both for short term (within 3 months) and for long term (greater than 3

months and less than 2 years) for all the depots to meet the expectations of customers.

When the customers feedback is appropriately acted upon, it in turn may be a customerretention strategy (short-term benefit) and will bring in more customers to patronise the

transport in future, i.e., customer development strategy (long-term benefit).


10/16



36

Table 1 Customer characteristics

Q.No Criteria Description

Q1 Bus punctuality arrival of the buses as per timings

Q2 Stopping bus at correct place bus stopping at the assigned bus stops

Q3 Backup service on breakdown in case of breakdown is a spare or backup bus is providedQ4 Seat comfort the seats provide comfort for travel

Q5 Cleanliness upkeep of the buses such as dusting, cleaning, etc.

Q6 Lighting & entertainment provision of a lights, TV, radio/FM, DVD, etc.

Q7 Provision for luggagewhether the loft is provided or arrangement to put the

luggage carried

Q8 New fleet addition whether new buses are added periodically

Q9 Obey traffic rules bus driver /crew following traffic signal, stop-lines, etc.

Q10 Seating for handicapped are physically challenged people provide separate seating

Q11 Seating for elderly are aged & elderly people provide separate seating

Q12 First aid facility whether first aid facility available in the bus

Q13 Driver behaviour whether driver behaviour is kind or courteous

Q14 Conductor behaviour whether conductor behaviour is kind or courteous

Q15 Information to passengersadequate information about change of route, stops, schedule

given to the passengers

Q16 Issue of proper ticket whether conductor issues proper ticket

Q17 In time issue of ticketwhether conductor issues ticket immediately on boarding

the bus

Q18 Issue of proper changewhether conductor issues proper change (correct amount)

for the ticket taken

Table 2 Principal component analysis with varimax rotation

Q.No. CommunalitiesRotated component matrix

Component-1 Component-2

Q1 0.640 0.737

Q2 0.520 0.682

Q3 0.587 0.667

Q4 0.661 0.724

Q5 0.692 0.820

Q6 0.603 0.737

Q7 0.539 0.585

Q8 0.692 0.779

Q9 0.508 0.564

Q10 0.645 0.731

Q11 0.623 0.746Q12 0.501 0.683

Q13 0.510 0.702

Q14 0.512 0.635

Q15 0.502 0.648

Q16 0.546 0.558

Q17 0.633 0.571 0.554

Q18 0.672 0.668


11/16



37

Table 3 Total Variance explained by the components

Component Total % Variance Cumulative %

1 5.882 32.677 32.677

2 4.690 26.057 58.734

Table 4 Grouping of factors (sub-criteria) under the two components (criteria); the

mean values for the three depots and Factor Coefficients Matrix.

Criteria Sub-CriteriaMean values Components

Depot-1 Depot-2 Depot-3 1 2

Customerexpectations

Bus punctuality (C11) 4.00 3.11 2.57 0.178 -0.080

Seat comfort (C12) 4.15 3.36 2.51 0.155 -0.049

Cleanliness (C13) 4.41 3.37 2.48 0.261 -0.186

Lighting & entertainment (C14) 4.35 3.41 2.50 0.197 -0.110

New fleet addition (C15) 4.13 2.95 2.50 0.199 -0.102

Seating for handicapped (C16) 4.09 3.11 2.41 0.169 -0.068

Seating for elderly (C17) 3.94 2.83 2.45 0.200 -0.111

Issue of proper ticket (C18) 3.85 3.49 2.45 0.059 0.055

In time issue of ticket (C19) 3.73 3.29 2.39 0.042 0.083

Issue of proper change (C110) 3.77 3.18 2.41 0.103 0.017

Company

responsibilities

Stopping bus at correct place (C21) 3.30 3.43 2.50 -0.121 0.245

Backup service during breakdown

(C22)3.53 3.62 2.50 -0.063 0.194

Provision for luggage (C23) 3.44 3.46 2.51 -0.014 0.136

Obey traffic rules (C24)3.45 3.33 2.50 -0.011 0.129

First aid facility (C25) 2.77 2.77 2.43 -0.152 0.270

Driver behaviour (C26) 3.21 3.31 2.56 -0.165 0.285

Conductor behaviour (C27) 3.50 3.23 2.42 -0.071 0.194

Information to passengers (C28) 3.56 3.70 2.41 -0.092 0.214

Table 5 Factor Scores for the three depots

F1 F2

D1 4.207 3.140

D2 2.715 3.833

D3 2.170 2.754


12/16

International Journal of Production Technology and Management (IJPTM), ISSN 0976 6383 (Print), ISSN 0976

January- December (2012), IAEME

38

Figure 1 Approach to Mapping Methods

Preference MapsSimilarity Based Map

Vector ModelIdeal Point Model

Discriminant

Analysis

Factor

Analysis

Correspondence

Analysis

Joi

External Analysis

Wit

Prefe

With Ideal Point

Preference Map

Mapping Methods

Perceptual maps

Attribute Based Non- Attribute Based


13/16

International Journal of Production Technology and Management (IJPTM), ISSN 0976 6383 (Print), ISSN

0976 6391 (Online) Volume 3, Issue 1, January- December (2012), IAEME

39

Figure 2 Flow chart for the principal component analysis (PCA)

Figure 3 Eigen value Scree plot

Identify customer characteristics

Develop the Perceptual Map

Apply Principal component analysis

Identification of factors

Determine the Component score

Identify the best performing depot

By Brainstorming

For grouping attributes

For data reduction

Using factor coefficients

and mean values

To position the depots


14/16



40

Figure 4 Perceptual map of three depots represented by vectors.

REFERENCES

1. Petroni .A and Braggia .M, (2000), Vendor selection using principal component analysis,The Journal of Supply Chain Management, Vol.36, pp 63-69.

2. ASTRU Association of state road transport undertakings, India (www.astru.org).3. Bijmolt .T and Wedel .M. (1999). A Comparison of Multidimensional Scaling Methods

for Perceptual Mapping, Journal of Marketing Research, Vol.36, pp 277-285.4. BusinessDictionary.com, Online Business Dictionary. (www.businessdictionary.com)

5. Chidambaranathan .S, Muralidaran .C and Deshmukh .S.G, (2009), Analyzing theinteraction of critical factors of supplier development using ISM an empirical study,

International Journal of Advanced Manufacturing Technology, Vol. 43, pp 1081-1093.

6. Ching .C.Y, Marlow .P.B, and Chin-Shan Lu, (2009), Assessing resources, logisticsservice capabilities, innovation capabilities and the performance of container shipping

services in Taiwan, International Journal of Production Economics Vol. 122, pp 420.

7. Kohli .C.S, and Leuthesser .L, (1993), Product positioning: A comparison of perceptualmapping techniques, The Journal of product and brand management, Vol: 2, No.4, pp 1-

20.

8. Chris A. Vassiliadis, George J. Siomkos, Aikaterini Vassilikopoulou, and John

Mylonakis, (2006), Product design decisions for developing new tourist destinations: Thecase of Rhodopi mountain TOURISMOS: An International multidisciplinary journal of

Tourism, Vol. 1, No.1, 2006, pp 93-110.

9. Eboli L and Mazzulla G, (2007), Service quality attributes affecting customer satisfactionfor Bus Transit, Journal of Public Transportation, Vol. 10, No.1, pp 21-34.

10.Lilien. G.L, Rangaswamy .A, and Bruyn .A.D, (2007), Principles of Marketing

Engineering, by Trafford Publishing, Victoria, BC, Canada.

F1

F2


15/16



41

11.Hair. J.F, Black.W.C, Babin B.J., Anderson.R.E, and Tatham.R.L, (2006), Multivariate

data analysis, sixth edition, Pearson Education, New Delhi.

12.Lattin .J.M, Carroll .J.D, and Green .P.E, (2003), Analyzing Multivariate Data, CengageLearning, New Delhi.

13.Davis .J, (2006), Perceptual mapping, Communication 560 Advertising research course

material, School of communication, San Diego University, pp 1-17.14.Shlens .J, (2009), A tutorial on principal component analysis,(www.snl.salk.edu/~shlens/pca.pdf)

15.Kannan V.R, and Tan K.C, (2002), Supplier selection and assessment, their impact on

business performance, Journal of Supply Chain Management, Vol. 38, pp 11-21.16.Kolanovic I, (2008), Defining the port service quality using factor analysis, Pomorstvo,

Vol. 22, No. 2, pp 83-97.

17.Lai. W.T., and Chen. C.F., (2010), Behavioural intentions of public transit passengers -The roles of service quality, perceived value, satisfaction and involvement, Transport

Policy, doi:10.1016/j.tranpol.2010.09.003.

18.Le Van Huy, Chinh-Yuan Huang, Connie, C.F. Hsu, and Truong Thi Dieu Thu, (2007),

Establishing brand positioning strategy based on the integration of MDS tool and radardiagram of the satisfaction consumer: Research with the card service bank, International

conference on business and information, Tokyo, Japan.

19.Nunnally J.C, (1978), Psychometric Theory, McGraw Hill, New York.20.Parikshat S. Manhas, (2010),Strategic Brand Positioning Analysis through Comparison

of Cognitive and Conative Perceptions, Journal of Economic Finance Administrative

Science, Vol.15, No.29, pp 15-33.21.Planning Commission (2002). 10th Five Year Plan (2002-2007) - Volume II: Sectoral

Policies and Programmes. New Delhi, Planning Commission, Government of India.

22.Rahaman R.K. and Rahaman Md.A., (2009), Service Quality attributes affecting the

satisfaction of passengers of a selective route in south-western Bangladesh, Theory andEmpirical Research in Urban Management, Vol. 3, No.12, pp 115-125.

23. Ramanathan .R. and Parikh .J.K., (1999) Transport sector in India: An analysis in the

context of sustainable development, Transport Policy, Vol. 6, No.1, pp 35-46.24.Sinclair .S.A. and Stalling .E.C, (1990). Perceptual Mapping; A Tool for Industrial

Marketing; A Case Study, The Journal of Business & Industrial Marketing, Vol. 5, pp 55-

6625.Esseghaier .S, (2010), Perceptual Mapping, a presentation, KOC University, Istanbul,

Turkey. (http://tmsbiz.com/wp-content/uploads/2010/02/perceptual-

mapping_mostimp.pdf)

26.Bhagavath .V, (2006), Technical efficiency measurement by data envelopment analysis:an application in transportation, Alliance Journal of Business Research, pp 60-72.

27.Vetrivel Sezhian M., Muralidharan C., Nambirajan T., and Deshmukh S.G., (2011),

Ranking of a public sector passenger bus transport company using principal componentanalysis: A case study, Management Research and Practice, Vol. 3, No.1, pp 60-71.

28.Wikipedia, (http://en.wikipedia.org/wiki/Principal_component_analysis).

29.Gillette .W, and Evans .J.H., (1975), " Service marketing: A bank marketing exampleusing Perceptual mapping", in Advances in Consumer Research, Vol. 02, pp 525-534.


16/16



42

30.Luyang .Z, and Weidong .W, (2010), Marketing science innovations and economic

development, Proceedings of 2010 Summit international marketing science and

management Technology Conference, China, pp 109- 115.31.Dimitriades .Z.S., (2006), Customer satisfaction, loyalty and commitment in service

organizations - some evidence from Greece, Management Research News, Vol. 29,

No.12, pp 782-800.

ACKNOWLEDGEMENT

The authors would like to thank the Managing Director, Tamil Nadu State TransportCorporation (TNSTC), Villupuram and his subordinates for giving us the necessary

permission for data collection in the Villupuram division.

AUTHORS PROFILE

1. M.Vetrivel Sezhian is a Professor in the Department of Mechanical Engineering, Dr.

Pauls Engineering College (Anna University), Villupuram, TamilNadu. His areas ofinterests are Production and Operations management.

Email: [email protected]

2. Dr.C.Muralidharan is a Professor in the Department of Manufacturing Engineering,Annamalai University, Chidambaram. His areas of interests are Quality Management

and Operations Management.

Email: [email protected]. Dr.T.Nambirajan is a Associate Professor in Management Studies, Pondicherry

University. His areas of interests are Operations Management and Supply Chain

Management.

Email: [email protected]. Dr.S.G.Deshmukh is a Professor in the Department of Mechanical Engineering,

Indian Institute of Technology, Delhi. At present (on deputation) Director, Indian

Institute of Information Technology, Gwalior. His areas of interests are Supply ChainManagement and Operations Management.

Email: [email protected].

Documents

Perceptual Mapping Using Principal Component Analysis in a Public Sector Passenger Bus Transport Company a Case Study