Handling stakeholder uncertain judgments in strategic transport service analyses

Transport Policy 29 (2013) 54–63

Contents lists available at ScienceDirect

Transport Policy journal homepage: www.elsevier.com/locate/tranpol

Handling stakeholder uncertain judgments in strategic transport service analyses Toni Lupo n Università degli Studi di Palermo, Dept. of Chemical, Management, Informatics and Mechanical Engineering, Palermo, Italy

art ic l e i nf o

a b s t r a c t

Available online 18 May 2013

The quality level of services has to be constantly controlled, especially under conditions of competition increasing and limited resources. However, considering that service performance analyses are based on stakeholders' judgments, they can be characterized by possible uncertainties related to incompleteness for partial ignorance, imprecision for subjectivity and even vagueness. Therefore, under these conditions, unreliable results can be obtained by widely used service analysis methodologies. In the present paper, a methodology based on a recent extension of the SERVQUAL model, and that uses in combined manner the fuzzy set theory and the analytic hierarchy process method is proposed to effectively handle uncertainty in service performance analyses. In particular, the fuzzy set theory is considered to deal with such uncertainty, whereas the AHP method is adopted as tool to estimate the importance weights of the strategic service attributes. Subsequently, the Italian public transport service sector is strategically analysed, and its overall service quality structure is described and, finally, the strategic analysis of the public urban transport service delivered in Palermo (Italy) is performed by means of the proposed methodology. The performed service analysis allows the most influencing service factors to be captured and commented upon. The obtained results show that the management's perception of service quality meaningfully influences the overall service performance level. & 2013 Elsevier Ltd. All rights reserved.

Keywords: Uncertainty ServQual model AHP method Fuzzy set theory Customer satisfaction evaluation Transport service analysis

1. Introduction Incorporation of stakeholder judgments is a necessary step to evaluate the performance of a service. In fact, for the peculiar characteristics of services, their performance level is not directly observable and so also does not directly measurable. Services are typically assessed by considering suitable and measurable service characteristics or factors, which performance levels provide an indirect assessment of the service performance. For example, the evaluation of customer satisfaction (CS) represents an indirect measure of the service performance level, since it is performed with relation to proper service factors whose performance levels, quantified by means of the so-called manifest variables, are intended as “latent effects” of the service performance level (Ding, 2006). The relationship between manifest variables and latent effects can be formalized by means of specific conceptual models. In the literature, several conceptual models have been proposed and among these, as pointed out by Büyüközkan et al. (2011), the ServQual model proposed by Parasuraman et al. (1985)

n

Tel.: +399123861879. E-mail addresses: [email protected], [email protected]

0967-070X/$ - see front matter & 2013 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.tranpol.2013.04.002

is still the widely used method for measuring service quality. Table 1 shows the most recent applications of such model in different service sectors. By considering the ServQual model, there are seven major Gaps in the service quality concept, which are shown in Fig. 1. Acording to a recent development of such model (Curry, 1999; Luk and Layton, 2002), the three main Gaps, which are more associated with CS, are: Gap 1, Gap 5 and Gap 6; since they have a direct relationship with customers. More in detail, such Gaps measure the discrepancy between:


customers' expectations and management's perceptions of service quality, for Gap 1;


customers' expectations and employees' perceptions of service quality, for Gap 6.


customers' expectations and their perceptions, for Gap 5; and they are defined with relation to all the service dimensions. By considering the cognitive sphere of the stakeholder, such Gap values can be obtained by the algebraic comparison between (Parasuraman et al., 1985):


management's perceptions of the customers' expectations (PM) and the customers' expectations (E): Gap 1 ¼PM−E;

T. Lupo / Transport Policy 29 (2013) 54–63


employees' perceptions of customers' expectations (PE) and the


customers' expectations (E): Gap 6 ¼PE−E: customers' perceptions (P) and the their expectations (E): Gap 5¼ P−E.

Thus, values assumed by Gap 1 can be considered as a result of the lack of a marketing research orientation, inadequate upward communication and too many layers of management, whereas Gap 6 values represent the result of the differences in the understanding of customer expectations by front-line service providers. Instead, Gap 5 values reflect the result of the influences exerted from the customer side and the shortfalls (Gaps) on the part of the service provider and therefore such values can be considered Table 1 Several recent applications of the ServQual model. Authors

Service sector

Bai et al. (2008) Grigoroudis et al. (2008) Song and Jamalipour (2008) Yun and Chun (2008) Chen et al. (2009) Large and König (2009) Liu et al. (2009) Lin (2010) Büyüközkan et al. (2011)

Public services Web portals Wireless systems Telemedicine service Shipping Purchasing General portals Regular chain service Healthcare

direct indicators of the CS degree. Therefore, customers' dissatisfaction is collected for service aspects in which a negative value of Gap 5 is obtained. Given the financial and resource constraints under which service organizations have to operate, it is crucial that customers' expectations are properly understood and measured and that, from the perspective of customers, any service Gaps are properly identified. In fact, the latter quantities can effectively support the decision maker in identifying suitable “Gaps oriented” service improvement solutions. In the light of the previous considerations, a purpose of the present paper is to estimate the values assumed by the three previously described main Gaps, considering the public urban transport service delivered in the city of Palermo (Italy). The measurement of transport service performance represents a crucial activity with relation to various aspects. First of all, to assess community expectations and perceptions related to the service main attributes, and to single out management problems regarding costs of the service (Transportation Research Board, 1994). In addition, the service performance measures can be used as monitoring tool to on-going control the service quality level and to compare the obtained performance service level over time and/ or across space (De Borger et al., 2002). Moreover, the considered service is one of the most important and critical public services delivered in Palermo, and it is currently characterized by a process of facilities modernization and overall quality improvement. Under such conditions, the proposed service analysis can assume a

Customer needs

Word of mouth communications

55

Past experience

Expected service Consumer Side Gap5 Gap6 Perceived service

Service delivery (including pre-and post contacts)

Gap1

Gap4

External communications to customers

Gap3 Employee perceptions of consumer expectation

Translation of perceptions into service quality specifications

Gap 2 Gap7 Management perceptions of consumer expectations Fig. 1. The ServQual conceptual model (Parasuraman et al., 1985).

Provider Side

56

T. Lupo / Transport Policy 29 (2013) 54–63

meaningful importance since it allows the identification of the critical to quality service attributes on which can be opportune to carry out improvement actions in order to improve the overall service satisfaction. In the literature, there is a variety of methods regarding the performance measures about the different transit service aspects (Gatta, 2006, 2008; Marcucci, 2005 ), specifically applied to a local public transport (Valeri et al., 2012; Marcucci and Gatta, 2006, 2007) and to possible methodological advantages (Felici and Gatta, 2008; Marcucci and Gatta, 2012). Such methods can be mainly classified as stated importance methods, in which customers are asked to rate each service attribute on an importance scale, or derived importance methods, in which the importance measure of each service attribute is statistically derived considering relationships among individual attributes with overall satisfaction. In addition, the perception level can be compared with the zone of expectations tolerance, i.e. the interval defined by the values of the maximum desired and minimum acceptable levels of expectations (Figini, 2003). However, for different reasons, with these methods, judgments provided by the customer can be affected by possible uncertainties related to incompleteness for partial ignorance, imprecision for subjectivity and even vagueness. The latter, can be also related to the considered questionnaire structures that can bring the customer to simultaneously judge all the considered service items, since the series arrangement of the service items can force the customer to simultaneously evaluate all the considered items (Simon, 1983; Miller, 1956). In order to overcome the previously described limitations, in the present paper the analytic hierarchy process (AHP) method (Saaty, 1980) is considered as tool to estimate the importance weights of the strategic service attributes, since it uses a sequential approach in which only a couple of attributes is compared at a time. The AHP is a multi-criteria decision making (MCDM) method that helps the decision-maker facing a complex problem with multiple conflicting and subjective criteria (e.g. location or investment selection, projects ranking, and so forth). Several papers have compiled the AHP success stories in very different fields. For example, as stressed by Berrittella et al. (2008), AHP has been widely applied to assess numerous complex environmental and economic problems (Duke and Aull-hyde, 2002; Ferrari, 2003). In the project management field, AHP has used for the assessment

Table 2 Uncertainty categories and theories (Ferdous et al., 2012). Type

Nature

Theory

Aleatory uncertainty Epistemic uncertainty

Irreducible, objective, random, stochastic, Ambiguous, ignorance, incomplete, inconsistent, imprecise, subjective, vague

Probably theory and Evidence theory Possibility theory, fuzzy set theory, and evidence theory

The customer Satisfaction

Goal

Service Dimensions Service Attributes

and allocation of human resources. For example Dweiri and Kablan (2006) propose a fuzzy decision making system (FDMS) for the evaluation of the project management internal efficiency by considering as evaluation criteria the project cost, the project time and the project quality and they suggest the use of AHP to find the relative weights of criteria. Also Certa et al. (2009) propose the use of AHP in the field of the project management. In the field of service quality assessment, AHP method has been suggested by Lupo and Passannanti (2008) to find out the relative weights of student requirements in high educational sector. Moreover, AHP presents several advantages as seeking consistency in judgments, easiness to use, etc. It also allows to structure complex problems in the form of a hierarchy or a set of integrated levels and can be combined with operations research techniques to handle more difficult problems. But, AHP in its original formulation can be unreliable in handling ambiguity of the concepts related to the human knowledge. In fact unfortunately, the latter is often incomplete, inconsistent and even vague or imprecise. As a consequence, the latter introduces uncertainty in service performance analyses. The choice of the technique to be used to minimize uncertainty effects is usually based on the type and nature of uncertainty as stated in Table 2. Since uncertainty related to service performance analyses is of epistemic type: It is generally ascribed to the coexistence of three relevant aspects, i.e. vagueness, imprecision and subjectivity in stakeholders' judgments (Curcurù et al., 2012), in the present paper the fuzzy set theory (FST) introduced by Zadeh (1965) is considered to deal with such uncertainty type. The FST allows the mathematical representation of uncertain knowledge and provide formalized tools for dealing with intrinsic imprecision of real life problems. In particular, it is particularly useful in the quantification of linguistic categories since it allows the representation for different “membership degrees” of a concept (Negoita, 1985). This feature is well represented by the idea expressed by Zadeh (1996) on the fuzzy set as tool to compute with words, which highlights the need for an effective interface between the crisp world of the numbers and linguistic categories in order to improve the understanding and utilization capability of real-life information. The FST has been applied in many fields of the management science (Büyüközkan et al., 2011a and Büyüközkan and Cifci, 2011b), but it is still quietly used in the field of the service quality assessment (Tseng, 2009a, 2009b). In the light of the previous considerations, in the present paper a methodology able to effectively handle uncertainty in service performance analyses, based on a recent development of the ServQual model and that uses in a combined manner the AHP method and the FST is proposed. Subsequently, the strategic analysis of the public urban transport service delivered in Palermo by using the proposed methodology is performed, and the possible implications for the service improvements are given. The remainder of the present paper is organized as follows: in the next section the theoretical issues of the proposed composite methodology are described; in Section 3, the Italian public

D1

A1,1 A1,2 …

D2

A1,C1

A2,1 A2,2 … A2,C2

Fig. 2. Structure of the service quality.

Dw

Aw,1 Aw,2

…

Aw,Cw

T. Lupo / Transport Policy 29 (2013) 54–63

transport service sector is analysed, and its strategic service quality structure is described; in Section 4 the evaluation of the public transport service delivered in Palermo (Italy) is performed by means of the developed composite methodology, and the obtained results are commented and, finally, the conclusions, with a summary and directions for future researches, close the work.

2. The methodology To adopt the proposed methodology, the first step concerns the identification of the service quality structure. The latter consists of several hierarchical levels: The first one shows the general objective or goal of the analysis, i.e. the CS; in the second level the service dimensions that allow to the service the capability to satisfy the general objective are reported. Subsequently, in the third level the service attributes for each service dimension are identified. The number of the considered hierarchical levels depends on the detail degree that one wants to carry out with the analysis. Fig. 2 shows a general three levels hierarchical quality structure composed by w service dimensions, D1, D2, …, DW, each one composed by C1, C2,…, Cw attributes. In particular, in Fig. 2 the generic service dimension i is indicated with the term Di, whereas its generic attribute j with the term Aij. In order to evaluate the service main Gaps, a suitable questionnaire structure has to be adopted. In particular, it is composed by two parts; in the first part, interviewees indicate the relative importance of all the pairwise comparisons of service attributes and dimensions. These evaluations are used to construct several pairwise comparison matrices that, processed by the considered fuzzy extension of the AHP method hereinafter described, allow the evaluation of the importance weight of each service attribute and dimension. On the contrary, in the second part of the questionnaire customers are asked to assess their perceptions related to service attributes. In both the questionnaire parts, interviewees point out the levels of their preferences by using suitable linguistic-fuzzy scales described below. In the next section, a brief overview about FST and its theoretical principles useful for the aim of the present work are given. Subsequently, the considered approaches for the measurement of customers' importance weights and perceptions level are described.

57

rule that associates to each linguistic category x of T(X) its meaning M(x). Such semantic rule may be defined by a fuzzy number M(x) in U. Thus the meaning of M(x) of a linguistic category x is defined by a membership function μx: U-[0,1] that associates to each u of U its compatibility with x (Klir and Yuan, 1999). A positive triangular fuzzy number (TFN), denoted as A~ ¼ ðxL ; xM ; xU Þ, where xL ≤xM ≤xU , has the following triangulartype membership function: 8 x−xL for xL ≤x ≤xM > < xM −xL U μA~ ðxÞ ¼ xx−x ð1Þ for xN ≤x ≤xU −x > : U M 0 otherwise Alternatively, by defining the interval of confidence level α (α-cut), a TFN can be characterized as ∀α∈½0; 1 A~ α ¼ ½aαL ; aαU  ¼ ½ðxM −xL Þα þ xL ; −ðxU −xM Þα þ xU 

ð2Þ

Service performance analyses often articulate stakeholders' knowledge/judgments in term of linguistic variables such as: very bad, poor, average, good, excellent, etc, and the use of TFNs as a way to compare fuzzy judgments has been proposed by van Laarhoven and Pedrycz (1983), and it is largely adopted in more recent works (Chang, 1996; Cheng, 1996; Kwong and Bai, 2002). In particular, Ayyub and Klir (2006) provided a chart to define the lower and upper boundary for such linguistic variables based on experts' assessment. Considering the most likely value as an average of these two boundaries, TFNs can be used to represent such linguistic variables. The fuzzy boundaries of a TFN may also be defined by means of the fuzzy Delphi method that is a typical multi-experts procedure for combining views and opinions (Kaufmann and Gupta, 1988). Finally, the FST allows the extension of arithmetic operations from crisp numbers to fuzzy numbers. By considering the membership degree α (α-cut) of positive fuzzy numbers, some main operations useful for the aim of the present work are given by the following expressions (Klir and Yuan, 1999): ∀aL ; aU ; bL ; aU ∈Rþ ; α α α Aα ⊕Bα ¼ ½aL þ bL ; aαU þ bU  α α α α Aα ΘBα ¼ ½aL −bL ; aU −bU  α α Aα ⊗Bα ¼ ½aαL  bL ; aαU  bU  α α α α Aα =Bα ¼ ½aL =bL ; aU =bU 

∀α∈½0; 1;

Aα ¼ ½aαL ; aαU ;

α

α

Bα ¼ ½bL ; bU  ð3Þ

2.1. Fuzzy set theory and linguistic-fuzzy scales In the FST, the concept of convexity of a set differs from that applied in the classical set theory: a fuzzy set is said convex, if and only if the degree of membership μA of an element x2 between two elements x1 and x3 is not less than the minimum value among the membership degrees of x1 and x3. Therefore, a fuzzy number A~ is a convex fuzzy set defined in R and such that 1. ∃x0 jmA ðx0 Þ ¼ 1 2. the membership function μA(x) continues. These properties are necessary to properly represent the considered conceptions. In particular, the general principle of the fuzzy assessment approach is that a linguistic variable can be seen as variable whose values are words or structured combinations of words whose meaning is defined by semantic rules. In particular, a linguistic variable is characterized by five elements (X, T(X), U, G, M), where X is the name of the variable; T(X) is the set of linguistic categories of the variable, U is the universe of discourse, G is a syntactic rule that generates the terms in T(X) and M a semantic

2.2. The evaluation of the importance weights of the service attributes As before said, in the present paper a composite approach between the AHP method and the FST is considered to effectively handle uncertainty related to service performance analyses. According to this purpose, in the literature several applications of fuzzy extensions of the AHP method have been proposed in different research fields (Chamodrakas et al., 2010; Fu et al., 2006 and Huang et al., 2008). A broad review on fuzzy AHP methods is presented by Demirel et al. (2008). On the contrary, in the service quality field, relatively few works have been proposed considering such composite approach and most of them only recently (Büyüközkan et al., 2011 and Ayag˘, 2005). To adopt the proposed fuzzy AHP approach a multi-step procedure has to be applied. The first step is to compare the importance score: Linguistic terms are used to indicate the relative importance of each pair of elements in the same hierarchy (see Fig. 2). The next step is related with the construction of the fuzzy comparison matrix. In particular, considering the attributes of the

58

T. Lupo / Transport Policy 29 (2013) 54–63

generic dimension k, the generic element a~ ij of the pairwise comparison matrix A~ k , represents the relative importance value, expressed in fuzzy form, of the attributes i vs the attribute j. Moreover, not all the Ck2 pairwise comparison coefficients have to be detected, but only Ck (Ck−1)/2 since a~ i;j ¼ 1=a~ j;i

ð4Þ

∀i≠j a~ i;i ¼ 1

ð5Þ

∀i; j ¼ 1; 2:::; C k

For the aggregation of multiple stakeholders' judgments in the form of a fuzzy number, a number of methods, e.g., max–min. arithmetic averaging, symmetric sum, t-norm, etc., are available. Among these, the geometric mean is the aggregator operator herein considered since, as pointed out by Enea and Piazza (2004), such operator allows the respect of the constraint expressed in Eq. (4). In fact, if p~ ijk is the fuzzy preference of the generic kth customer and t the total number of judgments to be aggregated, it is possible to write !1=t !1=t t

a~ i;j ¼

a~ j;i ¼

∏ p~ ijk

k¼1

t

∏ 1=p~ ijk

ð6Þ

k¼1

and consequently: a~ i;j ¼ 1=a~ j;i

ð7Þ

The next step is related to the computing of both the maximum fuzzy eigenvalue and the related fuzzy eigenvector of A~ k . The maximum fuzzy eigenvalue λ~ max of A~ k is a fuzzy number solution of the following fuzzy relationship: ~ ¼ λ~ max ⋅w ~ A~ k ⋅w

ð8Þ

~ is a fuzzy vector (Ck  1) composed by Ck fuzzy in which w numbers representing the importance weights of the Ck compared attributes of the dimension K. Considering the relationships reported in (2), for the generic attribute i, Eq. (8) is equivalent to ½ðaαL Þi;1 ⋅ðwαL Þ1 ; ðaαU Þi;1 ⋅ðwαU Þ1 ⊕:::⊕½ðaαL Þi;C K ⋅ðwαL ÞC K ; ðaαU Þi;C K ⋅ðwαU ÞC K  ¼ ½λαL ⋅ðwαL Þi ; λαU ⋅ðwαU Þi 

~ αi ¼ ½ðwαL Þi ; ðwαU Þi ; w

i; j ¼ 1; 2; :::; C k

α λ~ max ¼ ½λαL ; λαU 

ð10Þ

The α-cut is known to include the respondents' confidence over their preferences. In the case herein considered, it incorporates the customers' confidence and uncertainty over their judgments and, in such condition, the pairwise comparison coefficient aαij at the confidence level α, can be obtained with respect to the degree of satisfaction of the judgments, estimated by the index of optimism m (Cheng and Mon, 1994 and Chang, 1996). Such index is a linear convex combination defined as aαij ¼ μ⋅ðaαU Þi;j þ ð1−μÞ⋅ðaαL Þi;j ∀α∈½0; 1

ð11Þ

when α is fixed and after setting the index of optimism m, the following matrix (12) can be obtained in order to estimate the local importance weights of the considered attributes. 2

1

6 aα 6 21 AαK ¼ 6 6 ::: 4 aαn1

Fuzzy perception related to the attribute i of the generic service dimension k, at the confidence level α (α-cut), ðP~ α Þk;i , can be obtained with a reference to the judgments satisfaction degree. The latter is estimated by the index of optimism m. The larger value of the index m indicates the higher degree of optimism. Such index is a linear convex combination defined as ðP α Þk;i ¼ μðpαU Þk;i þ ð1−μÞ⋅ðpαL Þk;i ∀α∈½0; 1

ð13Þ ðpαU Þk;i

ðpαL Þk;i

and are the upper and lower bounds in which in Eq. (13) of fuzzy aggregated judgments at the confidence level α (α-cut), considering as aggregator operator the arithmetic mean. While α is fixed, after setting the index of optimism m, Eq. (13) gives the crisp value of customers' perception for the considered attribute.

3. Quality in transport service sector in Italy The transport service sector in Italy presents a significant economic size with about the 24.9% of the population over 14 years that uses the transport services for their displacements. The related market is affected by complex interactions among different economic subjects that give a particular configuration to its structure:

position;


a series of secondary operators connected with the main operator;

~ 1 ; :::; w ~ ck Þ; ~ t ¼ ðw w

α a~ ij ¼ ½ðaαL Þi;j ; ðaαU Þi;j ;

∀α∈½0; 1;

2.3. The evaluation of service perceptions


a main national operator that assumes a legal monopoly ð9Þ

in which A~ k ¼ ½a~ i;j ;

Eqs. (8), (10) and (12) correspond to the fuzzification of the Lambda-max method, initially introduced by Saaty (1980) in crisp term with the AHP method, which has been introduced by Csutora and Buckley (2001). Subsequently, the same procedure is adopted also to compare the service dimensions. The last step is to determine the global importance weights of the service attributes. The latter can be obtained by multiplying the local importance weights of each attribute by the importance weight or the related service dimension.

aα1;2

:::

1

:::

:::

:::

:::

:::

aα1n

3

aα2n 7 7 7 ::: 7 5 1

ð12Þ


many small operators organized, in general, in trade associations;
supply chains;
Regulatory authorities in the sector. The Italian public transport sector is characterized by a crisis condition that by now persists by several years. Over the last five years, the reduction of the users' number is equal to 19% and, at the same period, the increasing of the kilometres number performed by means of private vehicles is equal to 28%. There are no doubts that the widespread increasing of the life quality has contributed to establish such situation. In addition, such crisis condition can be also associated with the fact that customers, on average, perceived public transports characterized by a low overall quality level (European Commission, 2011). To the contrary, the excessive use of private vehicles has led to the traffic congestion phenomenon with other harmful consequences such as: increased number of accidents, air and noise pollution, energy consumption and therefore with consequences also for the environment. For these reasons, the regulations at the levels of European Union (EU), national and local, encourage the development of policies that discourage the use of private vehicles and that aim to the improvement of the public transit service quality. In Italy, public transit service transformation is mainly related to the deep normative reform that is affecting the entire sector. As

T. Lupo / Transport Policy 29 (2013) 54–63

59

pointed out by Marcucci and Gatta (2007), the relevant key factors of such transformation concern:


customer expectations identification, with respect to both





those explained out by customers and those implicitly considered satisfied by the service; service delivery system design, in order to correctly “translate” customers' expectations in service specifications; service delivery system that comprises operations standardization and the control of the critical to quality service factors; internal and external communication of achieved quality results, with the aim to involve both employees and customers in the continuous improvement process of the service; service performance evaluation.

The latter is characterized by significant aspects of complexity, given that the service performance evaluation has necessarily to reflect the point of views of different service stakeholders: the transport company, the local community, directly or indirectly involved in the transit service and customers. The company point of view essentially tends to focus on costs efficiency/effectiveness (Bertini and El-Geneidy, 2003). A measure of cost efficiency is typically defined as produced services (e.g. vehicle kilometers), while a measure of service effectiveness is defined as consumed service (e.g. passenger kilometers) (see Fig. 3). On the contrary, the community point of view is affected by matters relating to equipment, in terms of quantity, quality and safety, and the environmental impact of the service. Finally, the customers' point of view is related to their perceived quality level of the delivered service and can be considered the main driver of the investment choices to improve the service quality. In fact, apart from certain essential aspects of the service, the investment choices should strategically take into primary consideration the customers' point of view, considering their needs with the related importance levels. Therefore, it is clear the need to define the quality structure of the transit service, i.e. the set of the critical to quality service dimensions and attributes, with respect to which to evaluate the CS level and to consider the use of the other quality cycle tools (Fig. 4), in order to allow an effective and efficient quality improvement of the service. The previously described transformation process is also affecting the public transport service delivered in Palermo (Italy). Such service is one of the most important and critical public services provided in Palermo and it is currently characterized by a process of facilities modernization and quality improvement. For such reasons, the analysis reported below has been performed.

Service Inputs: Labor, Capital, Fuel

Service Outputs: Vehicle Hours, Vehicle Kilometers, Capacity Kilometers

Service Effectiveness

Quality Structure of Service

Customer Satisfaction

Service Assurance

Benchmarking Standardization and Certification Fig. 4. The quality cycle (Figini, 2003).

Table 3 The overall structure of the public transport service (Eboli and Mazzulla, 2007).

Transit Service Quality

Service Dimension

Service Attribute

Route characteristics

Path Number of bus stops Distance between bus stops Bus stops location;

Service characteristics

Service frequency Daily service time

Service reliability

Reliability of the scheduled runs Punctuality of the runs;

Information

Availability of schedule/maps on bus Availability of schedule/maps at bus stops Availability of information by phoneinternet;

Personnel

Personnel appearance Personnel helpfulness;

Customer service

Easiness of purchasing a ticket Administration of complaints

Comfort

Bus crowding Comfort of bus seats Air condition on bus Level of vibration on bus Availability of shelter and beaches at bus stops

Safety and security

Bus reliability Competence of drivers Security against crime on bus Security against crime at bus stops;

Cleanliness

Cleanliness of bus interior, seats and windows Cleanliness of bus exterior

3.1. The performance evaluation of the transport service of Palermo.

Cost Effectiveness

Cost Efficiency

Quality Cycle

Service Consumptions: Passengers, Passenger Kilometers, Operating Revenue

Fig. 3. Relationship between efficiency and effectiveness indicators (Bertini and ElGeneidy, 2003).

The public transport service of Palermo is supplied by the Palermo Public Urban Transport Company (AMAT S.p.a.) and covers the entire urban territory by means of about 90 bus lines distributed over 20 service hours for day. The daily customer's basin is of about 600.000 potential customers, mainly composed by citizens. The overall structure of the public transit service quality stated in Table 3 has been considered to single out the relevant elements of the quality structure of the considered service. In particular, such elements have been selected from the overall structure by using the Critical Cases Approach (CCA) (Cronin and Taylor, 1992), on the basis of preliminary interviews to both service experts (decision makers group) and a limited number of customers (Table 4). In particular, with respect to the overall structure of the public transport service (Table 3), the dimensions Customer Service and

60

T. Lupo / Transport Policy 29 (2013) 54–63

Table 4 Relevant elements of the Palermo transit service quality structure. Goal:

Customer satisfaction

Dimension:

D1 Route characteristics

D2 Information

D3 Personnel

D4 Comfort

D5 Safety and security

Service attribute:

A11: Path;

A21: Availability of schedule/maps on bus; A22: Availability of schedule/maps at bus stops; A23: Availability of information by phone-internet;

A31: Personnel appearance; A32: Personnel helpfulness;

A41: Bus crowding;

A51: Bus reliability;

A42: Comfort of bus seats;

A52: Competence of drivers; A53: Security against crime on bus; A54: Security against crime at bus stops;

A12: Number of bus stops; A13: Distance between bus stops; A14: Bus stops location;

A43: Air condition on bus; A44: Level of vibration on bus; A45: Availability of shelter and beaches at bus stops

Table 5 Extract of the adopted questionnaire. First part: How important is:

Compared with

Availability of schedule/maps on bus

Availability of schedule/maps at bus stops

Availability of information by phone-internet;

Please, write the letter related to your judgement.

Please, write the letter related to your judgement.

Availability of schedule/maps at bus stops

Please, write the letter related to your judgement.

D: C: B: A: e: a: b: c: d:

Extremely more important Very strongly important Strongly important Moderately important Equally important Moderately less important Strongly less important Very strongly less important Extremely less important

Second part: Information:

Very bad

Poor

Average

Good

Excellent

□ □ □

□ □ □

□ □ □

□ □ □

□ □ □

Indicate the performance level of the following service attributes Availability of schedule/maps on bus Availability of schedule/maps at bus stops Availability of information by phone-internet

Cleanliness are not considered relevant for the service delivered in Palermo. Subsequently, a suitable questionnaire structure has been developed. In particular, the questionnaire is composed by two parts; in the first one, customers are asked to indicate the relative importance of all the pairwise comparisons of service attributes and dimensions. Instead, in the second one, customers are asked to assess their perceptions related to the service attributes. In both the questionnaire parts, customers point out the levels of their judgments by using suitable fuzzy-linguistic evaluation scales. Table 5 shows the first and the second part of the questionnaire related to the service dimension Information. Finally, for both, the index of optimism m and the confidence level α (α-cut,) a value equal to 0.5 have been assumed and the linguistic-fuzzy scales reported in Table 6 have been considered. The survey has been conducted for three months, between May and July 2012, and about 300 customers and a total of 50

respondents between decision makers and service employers have been interviewed. Table 7 reports the obtained levels of customers' expectations and perceptions. As it can be seen from Table 7, from customers perspective, the most important service dimension is D1 (0.404), and its most important attributes are in order A11 (0.127), A13 (0.119) and A12 (0.091); subsequently, the second service dimension for importance is D5 (0.215), and A54 (0.084) is its most important attribute. Lastly, the other service dimensions are characterized by similar importance levels. In particular, the least important service dimension is D4 (0.118) and its least important attribute is A43 (0.009). The same table also shows the service perception levels: The most powerful attribute for is A51 (0.088) followed by the attributes A52 (0.077) and A31 (0.076). The attributes A14, A23, A12, A11 and A13 are characterized by similar perception levels, within the range (0.075; 0.071). Follow in order all the other attributes that present a gradual decreasing of the perception levels, within the range (0.063; 0.022).

T. Lupo / Transport Policy 29 (2013) 54–63

On the contrary, the obtained levels of service quality perceptions related to the management and employers point of views are listed in Table 8. The table also shows the obtained service main Gaps values.

Table 6 Linguistic-fuzzy evaluation scales. Perception evaluation scale

Importance evaluation scale

Linguistic category

Triangular fuzzy number

Linguistic category

Triangular fuzzy number

Very bad Poor

(0, 1, 3) (2, 3, 5)

(1, 1, 3) (1, 3, 5)

Average Good

(3, 5, 7) (5, 7, 9)

Excellent

(7, 9, 11)

Equal importance Moderate importance Strong importance Very strong importance Extremely more importance

(3, 5, 7) (5, 7, 9) (7, 9, 11)

Table 7 Customers' expectation and perception levels. Dimension Importance weight

Attribute Local importance weight

Global importance weight

Perception

D1

0.404

D2

0.137

D3

0.126

D4

0.118

A11 A12 A13 A14 A21 A22 A23 A31 A32 A41 A42 A43 A44 A45 A51 A52 A53 A54

0.127 0.091 0.119 0.067 0.022 0.069 0.046 0.042 0.084 0.030 0.024 0.009 0.017 0.038 0.025 0.042 0.065 0.084

0.073 0.072 0.071 0.074 0.033 0.044 0.074 0.076 0.044 0.041 0.063 0.061 0.031 0.034 0.088 0.077 0.022 0.022

D5

0.215

0.314 0.226 0.295 0.165 0.163 0.503 0.334 0.335 0.665 0.258 0.203 0.073 0.142 0.324 0.114 0.195 0.301 0.39

61

As can be seen from Table 8, considering Gap 5, it emerges that the most satisfied attribute is A51 (0.063) followed by the attributes A43 (0.052) and A42 (0.039). The attributes A52, A31, and A23 are characterized by similar satisfaction levels, within the range (0.035; 0.028). Follow in order the attributes A44, A21, A41, and A14 that present a gradual decreasing of the satisfaction levels, within the range (0.014; 0.007). Conversely, the most dissatisfied attribute is A54 (−0.062) followed by the attributes A11 (−0.054), A13 (−0.048), A53 (−0.043) and A32 (−0.040). Finally, the attributes A22, A12 and A45 present similar dissatisfaction levels, within the range (−0.025; −0.004). The latter results can be conveniently used to support the decision maker in identifying a suitable strategy for the service quality improvement. In particular, in the light of the obtained results, the strategic “Gaps oriented” implications for the service improvement should take into account the service dimensions Route Characteristics (D1) and Safety and Security (D5), and, in particular, the service attributes Security against crime at bus stops (A54), Path (A11), Distance between bus stops (A13) and Security against crime on bus (A53). Finally, in order to investigate on effects on CS levels (Gap 5) by the variables Gap 1 and Gap 6, a regression model has been developed. Table 9 summarizes the obtained results. The regression model presents a corrected R2 equal to 0.42, that is the 42% of the variation of the CS level (Gap 5) is explained by the independent variables Gap 1 and Gap 6. In particular, as shown in Table 9, the variable Gap 1 meaningfully affects the CS level. The latter result highlights the meaningful importance assumed by correct management's perceptions of service quality on the achieved CS level. Furthermore, the assumed value by the statistic F, equal to 7.036, permits to reject, at the 1% level, the null statistic

Table 9 Main effects. Independent variable

Estimated coefficient

Description

Value

t Statistic

(Intercept) Gap 1 Gap 6

4.95E−5 0.631 0.203

0.007 2.573a 1.255

a

Significant at the 2% level (two-tailed test).

Table 8 Management and employers service quality perception levels and service main Gap values. Dimension

Importance weight

Attribute

Decision makers

Employers

D1

0.358

0.129

D2

0.106

0.251

D3

0.174

0.119

D4

0.077

0.168

D5

0.285

0.333

A11 A12 A13 A14 A21 A22 A23 A31 A32 A41 A42 A43 A44 A45 A51 A52 A53 A54

Local importance weight

Global importance weight

Main gap

Decision makers

Employers

Decision makers

Employers

Gap 1

Gap 6

Gap 5

0.202 0.315 0.238 0.245 0.304 0.267 0.429 0.443 0.557 0.086 0.236 0.307 0.107 0.264 0.345 0.236 0.295 0.124

0.258 0.234 0.284 0.224 0.486 0.274 0.24 0.348 0.652 0.125 0.361 0.274 0.098 0.142 0.135 0.284 0.393 0.188

0.072 0.113 0.085 0.088 0.032 0.028 0.045 0.077 0.097 0.007 0.018 0.024 0.008 0.020 0.098 0.067 0.084 0.035

0.033 0.030 0.037 0.029 0.122 0.069 0.060 0.041 0.078 0.021 0.061 0.046 0.016 0.024 0.045 0.095 0.131 0.063

−0.055 0.021 −0.034 0.021 0.010 −0.041 0.000 0.035 0.013 −0.024 −0.006 0.015 −0.009 −0.018 0.074 0.025 0.019 −0.049

−0.094 −0.061 −0.083 −0.038 0.100 0.000 0.014 −0.001 −0.006 −0.009 0.037 0.037 0.000 −0.014 0.020 0.053 0.066 −0.021

−0.054 −0.019 −0.048 0.007 0.011 −0.025 0.028 0.034 −0.040 0.011 0.039 0.052 0.014 −0.004 0.063 0.035 −0.043 −0.062

62

T. Lupo / Transport Policy 29 (2013) 54–63

hypothesis of no significant relations among the independent variables and the Gap 5 and Gap 6 values.

4. Conclusions In the present paper, a methodology able to effectively handle uncertainty related to service analyses based on stakeholders' judgments has been developed. The application of such methodology has been shown in a strategic transport service analysis related to the public urban transport service delivered in Palermo. From such analysis, the service main Gaps have been evaluated and a suitable “Gaps oriented” strategy for the service quality improvement has been identified. Moreover, also the effects of the discrepancies between customers' expectations and management's perceptions of service quality (Gap 1) and customers' expectations and employees' perceptions of service quality (Gap 6) on the customer satisfaction level (Gap 5) have been investigated and quantified by means of a regression model. Future researches concerning transport service analyses will involve: (i) the connection between the attribute “cost service” and the other quality attributes, in order to obtain more reliable results; (ii) the evaluation of quality perception of transport services from non-users standpoint, to single out the service improvements configuration to make it attractive for more usercategories; (iii) the further development of the proposed methodology by considering the Fuzzy Logic approach.

Acknowledgements The Author would like to thank AMAT S.p.a. Palermo, for the willingness shown during the data collection phase and the reviewers for their constructive suggestions. References Ayag˘, Z., 2005. A fuzzy AHP-based simulation approach to concept evaluation in a NPD environment. IIE Transactions 37, 827–842. Ayyub, B.M., Klir, G.J., 2006. Uncertainty modeling and analysis in engineering and the sciences. Chapman & Hall, Boca Raton, FL, US 33487-2742. Bai, C.H., Lai, F.J., Chen, Y., Hutchinson, J., 2008. Conceptualising the perceived service quality of public utility services: a multi-level, multi-dimensional model. Total Quality Management & Business Excellence 19 (10), 1055–1070. Berrittella, M., Certa, A., Enea, M., Zito, P., 2008. Transport policy and climate change: how to decide when experts disagree. Environmental Science & Policy 11 (4), 307–314. Bertini, R.L., El-Geneidy, A., 2003. Using archived data to generate transit performance measures. In: 82th TRB Annual Meeting, Washington, DC, 12–16 January, 2003. Büyüközkan, G., Çifçi, G., Guleryuz, S., 2011. Strategic analysis of healthcare service using fuzzy AHP methodology. Expert Systems with Applications 38, 9407–9424. Büyüközkan, G., Feyzioglu, O., Cifci, G., 2011a. Fuzzy multi-criteria evaluation of knowledge management tools, An International Journal on Computational Intelligence Techniques, Methods and Applications, 4 (2), 184–195. Büyüközkan, G., Cifci, G. (2011b) A novel fuzzy multi-criteria decision framework for sustainable supplier selection with incomplete information, Computers in Industry, 62 (2), 164–174. Certa, A., Enea, M., Galante, G., La Fata, C.M., 2009. Multi-objective human resources allocation in R&D projects planning. International Journal of Production Research 47 (13), 3503–3523. Chamodrakas, I., Batis, D., Martakos, D., 2010. Supplier selection in electronic marketplaces using satisficing and fuzzy AHP. Expert Systems with Applications 37 (2010), 490–498. Chang, Da-Yong, 1996. Application of the extent analysis method of fuzzy AHP. European Journal of Operational Research 95, 649–655. Chen, K.K., Chang, C.T., Lai, C.S., 2009. Service quality gaps of business customers in the shipping industry. Transport Research 45, 222–237. Cheng, Ching-Hsue, Mon, D.L., 1994. Evaluating weapon system by analytical hierarchy process based on fuzzy scales. Fuzzy Sets And Systems 63, 1–10. Cheng, C.H., 1996. Evaluating naval tactical missile systems by fuzzy AHP based on the grade value of membership function. European Journal of Operational Research 96, 343–350.

Cronin, J., Taylor, S.A., 1992. Measuring service quality: a reexamination and extension. Journal of Marketing 56, 55–68. Csutora, R., Buckley, J.J., 2001. Fuzzy hierarchical analysis: the Lambda-max method. Fuzzy sets and Systems 120, 181–195. Curcurù, G., Galante, G.M., La Fata, C.M., 2012. Epistemic uncertainty in fault tree analysis approached by the evidence theory. Journal of Loss Prevention in the Process Industries 25, 667–676. Curry, A., 1999. Innovation in public service management. Managing Service Quality 9 (3), 180–190. De Borger, B., Kerstens, K., Costa, A., 2002. Public transit performance: what does one learn from frontier studies? Transport Reviews 22 (1), 1–38. Demirel, T., Demirel, N.C., Kahraman, C., 2008. Fuzzy analytic hierarchy process and its application. In: Kahraman, C. (Ed.), Fuzzy Multi-Criteria Decision Making. Springer Science+Business Media, LLC, pp. 53–83. Ding, C.S., 2006. Multidimensional scaling modeling approach to latent profile analysis in psychological research. International Journal of Psychology 41 (3), 226–238. Duke, J.M., Aull-hyde, R., 2002. Identifying public preferences for land preservation using the analytic hierarchy process. Ecological Economics 42, 131–145. Dweiri, F.T., Kablan, M.M., 2006. Using fuzzy decision making for the evaluation of the project management internal efficiency. Decision Support Systems 42, 712–726. Enea, M., Piazza, T., 2004. Project selection by constrained fuzzy AHP. Fuzzy Optimization and Decision Making 3, 39–62. Eboli, L., Mazzulla, G., 2007. Service quality attributes affecting customer satisfaction for bus transit. Journal of Public Transport 10 (3), 21–34. European Commission, DG Energy and Transport, Statistical Coverage and Economic Analysis of the Logistics Sector in the EU, 2011. Felici, G., Gatta, V., 2008. The analysis of aervice quality through stated preferences and rule-based classification. In: Felici, G., Vercellis, C. (Eds.), Mathematical for Knowledge Discovery and Data Mining. Information Science Reference, IGI Global, pp. 65–81. Ferrari, P., 2003. A method for choosing from among alternative transport projects. European Journal of Operational Research 150, 194–203. Ferdous, R., Klan, F., Sadiq, R., Amyotte, P., Veitch, B., 2012. Handling and updating uncertain information in bow-tie analysis. Journal of Loss Prevention in the Process Industries 25, 8–12. Figini, M., 2003. Dare valore alle esigenze dei clienti e dei dipendenti dell'azienda. Franco Angeli, Milano. Fu, H.P., Ho, Y.C., Chen, R.C.Y., Chang, T.H., Chien, P.H., 2006. Factors affecting the adoption of electronic marketplaces: a fuzzy AHP analysis. International Journal of Operations & Production Management 26 (12), 1301–1324. Gatta, V., 2006. Valutare la qualità dei servizi. Un nuovo approccio basato sulla Conjoint Analysis. Statistica, anno LXVI 1, 85–113. Gatta, V., 2008. La qualità e la customer satisfaction di un servizio aeroportuale: una metodologia da affiancare a quella tradizionale. Rivista di Economia e Statistica del Territorio 3, 137–159. Grigoroudis, E., Litos, C., Moustakis, V.A., Politis, Y., Tsironis, L., 2008. The assessment of user-perceived web quality: application of a satisfaction benchmarking approach. European Journal of Operational Research 187, 1346–1357. Huang, C.C., Chu, P.Y., Chiang, Y.H., 2008. A fuzzy AHP application in government sponsored R&D project selection. Omega 36, 1038–1052. Kaufmann, A., Gupta, M.M., 1988. Fuzzy Mathematical Models in Engineering and Management Science. North-Holland, Amsterdam 2008. Klir, G.J., Yuan, B., 1999. Fuzzy sets and fuzzy logic, Theory and Applications. Prentice Hall P T R, Upper Saddle River, NJ. Kwong, C.K., Bai, H., 2002. A fuzzy AHP approach to the determination of importance weights of customer requirements in quality function deployment. Journal of Intelligent Manufacturing 13, 367–377. Large, R.O., König, T., 2009. A gap model of purchasing's internal service quality: concept, case study and internal survey. Journal of Purchasing & Supply Management 15 (1), 24–32. Liu, C.T., Du, T.C., Tsai, H.H., 2009. A study of the service quality of general portals. Information & Management 46 (1), 52–56. Lin, H.-T., 2010. Fuzzy application in service quality analysis: an empirical study. Expert Systems with Applications 37, 517–526. Luk, ShTK, Layton, R., 2002. Perception gaps in customer expectations: managers versus service providers and customers. The Service Industries Journal 22 (2), 109–128. Lupo, T. Passannanti, A., 2008. Proceedings of the evaluating of the importance weights of student requirements using analytic hierarchy process. In: The International Manufacturing Conference Dublin, 3–5 September 2008, pp. 341–348. Marcucci, E., 2005. Introduzione ai modelli a scelta discreta per lo studio dei trasporti, Carocci Editore, Roma. Marcucci, E. Gatta, V., 2006. Valutare la qualità di un servizio pubblico differenziato territorialmente, 〈http://www.3.unipv.it/websiep/wp/554.pdf〉. Marcucci, E., Gatta, V., 2007. Modelli a scelta discrete per il benchmarking della qualità nel trasporto pubblico locale. In: Polidori, G., Borruso, G., Danielis, R. (Eds.), I trasporti e il mercato globale. Franco Angeli, pp. 63–72, a cura di. Marcucci, E., Gatta, V., 2012. Dissecting preference heterogeneity in consumer stated choices. Transportation Research E 48 (1), 331–339. Marcucci, E., Gatta, V., 2007. Quality and public transport service contracts. European Transport 36, 92–106. Miller, G.A., 1956. The magical number seven plus or minus two: some limits on our capacity for processing information. Psychological Review 63, 81–97. Negoita, C.V., 1985. Expert Systems and Fuzzy Systems. Benjamin/Cummings, Menlo Park, CA.

T. Lupo / Transport Policy 29 (2013) 54–63

Parasuraman, A., Zeithaml, V.A., Berry, L.L., 1985. A conceptual model of service quality and its implications for future research. Journal of Marketing 49, 41–50. Saaty, T.L., 1980. The Analytical Hierarchy Process. McGrawHill International Book Compagny, New York. Simon, H.A., 1983. Reason in Human Affairs. Stanford University Press, Stanford. Song, Q., Jamalipour, A., 2008. A quality of service negotiation-based vertical handoff decision scheme in heterogeneous wireless systems. European Journal of Operational Research 191, 1059–1074. Transportation Research Board, 1994. The Role of Performance-Base Measures in Allotting Funding for Transport Operations, TCRP Synthesis 6. National Academy Press, Washington, DC. Tseng, M.-L., 2009a. A causal and effect decision making model of service quality expectation using grey-fuzzy DEMATEL approach. Expert Systems with Applications 36, 7738–7748.

63

Tseng, M.-L., 2009b. Using the extension of DEMATEL to integrate hotel service quality perceptions into a cause–effect model in uncertainty. Expert Systems with Applications 36, 9015–9023. Valeri, E., Stathopoalos, A., Marcucci, E., Gatta, V., 2012. Local public transport: service quality and tendering contracts. In: Venezia, E. (Ed.), Urban Sustainable Mobility. Franco Angeli, pp. 161–172, a cura di. van Laarhoven, P.J.M., Pedrycz, W., 1983. A fuzzy extension of Saaty's priority theory. Fuzzy Sets And Systems 11, 229–241. Yun, E.K., Chun, K.M., 2008. Critical to quality in telemedicine service management: application of DFSS (design for six sigma) and SERVQUAL. Nursing Economics 26 (6), 384–388. Zadeh, L.A., 1996. Fuzzy logic computing with words. IEEE Transactions on Fuzzy Systems 4, 103–111. Zadeh, L.A., 1965. Fuzzy set. Information and Control, 8 (3), 338–353.