Service Quality Attributes for Public Transportation in Indonesian Cities

Transcription

1 Service Quality Attributes for Public Transportation in Indonesian Cities Tri Basuki JOEWONO a, Djoen San SANTOSO b a Graduate School, Parahyangan Catholic University, Bandung, 40141, Indonesia a vftribas@unpar.ac.id b School of Engineering and Technology, Asian Institute of Technology, Pathumthani, 12120, Thailand b djoensan@yahoo.com Abstract: Many developing cities are trying to improve their public transportation systems. Thus, study into the characteristics of each city s public transport services is needed. The motive of this study is to elaborate the specific characteristics of service quality in public transport for three Indonesian cities, i.e. Jakarta, Bandung, and Yogyakarta. The objective is to compare the importance of service quality attributes in these three cities. Study results show that each attribute is perceived as having a different level of importance. Comparisons among cities for each attribute show that six out of 15 attributes are significantly perceived as different. Further analysis shows that these cities have similarities as well as differences. Each city s users are unique, implying that the existing generalized approach for all cities in a big country like Indonesia is doubtful. A unique approach which accounts for each city s specific characteristics seems more rational. Keywords: Service Quality, Attribute, Public Transport, Developing Cities 1. INTRODUCTION Transportation problems have occurred in almost all major cities in Indonesia, such as Jakarta, Bandung, and Yogyakarta. One approach that is believed can overcome the problems in urban areas is the implementation of a strong public transport system, which is also recommended as a way to achieve sustainable transportation (SUTP, 2009). Additionally, many experts have suggested that appropriate approaches in managing the demand of transport will lead to success (VTPI, 2009). At present, there is much concern over the quality and quantity of public transportation in various cities in Indonesia to support economic development and commuting needs. On one side, the rapid motorization, along with economic development, intensifies the competition between private autos and mass transit. On the other side, poor service and insufficient quantity of public transportation due to weak management and the financial burden of public transportation discourages the use of public transport (Susilo et al., 2007). Many big cities in Indonesia have been trying to improve their public transportation system, such as DKI Jakarta with the implementation of BRT, which is known as TransJakarta Busway, since 2004, and the planning of monorails and LRT. Yogyakarta has implemented TransJogja busway since 2008, while Bandung has implemented Trans Metro Bandung since In addition to implementing new transit systems, various cities are trying to organize their public transportation operation systems in many ways. These initiatives are a positive sign for the improvement of public transportation. However, it is also necessary to consider the local situation and the market demand to achieve sustainable transportation. Adequate knowledge of the market determines success in this field 1064

2 (Hensher and Brewer, 2001). Understanding the market can be defined as an effort to understand the expectations, demands, and needs of the community, especially the user. It covers various aspects of the quality of public transport. Measuring the quality of transportation services involves a number of attributes, both qualitative and quantitative (MORPACE Int. Inc. and Cambridge Systematics Inc., 1999; Kittelson & Associates Inc. et al., 2003). Quality loop develop by AFNOR, as cited by QUATTRO (1998), involves two parts of quality measurement. The first one is the measurement of customer satisfaction and the second one is the measurement of the system s performance. Quality of services measured by the user are based on the gap between the expected quality and the perceived experience. As this is qualitative and subjective in nature, there is a chance of difference from one user to the other, or from one city to another. Differences appear as a result of perception, culture, environment, or character. To bring the differences together with the unique situation in each city, unique planning is needed for each city. It represents an effort to produce more suitable and acceptable policies, as well as approaches, for both users and non-users in the community. Indeed, research that elaborates the characteristic of each city in providing public transport services becomes very important. Unfortunately, studies of this topic are very rare in developing countries, like Indonesia. Some studies have been initiated, such as Joewono and Kubota (2007abc, 2008), which mainly focus on one type of mode and are located in one city. A gap still exists for research to explore the service characteristics expected by users for public transport in Indonesian cities which may be unique for each city due to the influences of local culture in the form of behavior or attitudes, geographical and topographical characteristics, or climate among others. Based on these reasons, the motive of this study is to elaborate the specific characteristics of the aspects of service quality of public transport for three Indonesian cities, i.e. Jakarta, Bandung, and Yogyakarta. The objective is to compare the importance of service quality attributes in these three cities. Surveys were conducted by questionnaire distribution to users of public transportation. The introduction of this article is followed by a brief discussion regarding the service quality of public transportation in section two. Section three explains the research method. Detailed description regarding the social demographic and travel characteristics of the respondents is provided in section four. The data and analysis can be found in section five, which consists of a comparison using confidence interval and factor analysis. Concluding remarks are provided in the final section. 2. SERVICE QUALITY OF PUBLIC TRANSPORTATION Performance measurement in public transportation indicates the achievement of the service provider in providing services for the public. Traditionally, measurement is represented by effectiveness and efficiency indicators. As an example, the performance of a bus service is assessed using several indicators (World Bank, 1987). These indicators mainly apply quantitative measurements, while as a matter of fact, the service of public transportation has a qualitative and subjective nature. Thus, to comprehensively measure service performance, a long list of indicators has been developed (MORPACE Int. Inc. and Cambridge Systematics Inc., 1999; Kittelson & Associates Inc. et al., 2003). These indicators cover quantitative and qualitative measurements, i.e. subjective and objective. Although specific aspects of service quality may be particularly positive or negative in a user s perception, it is the total package of services evaluated by a user that suitably measure the overall level of user satisfaction. For 1065

3 this purpose, Hensher and Prioni (2002) and Hensher et al. (2003) introduced and developed a service quality index (SQI) as a way to capture customer satisfaction with the service and to benchmark service effectiveness. Eboli and Mazzulla (2007) employed a structural equation model to explore the impact of the relationship between global customer satisfaction and service quality attributes. Quality of service can be examined from the users perspective and the providers perspective (MORPACE Int. Inc. and Cambridge Systematics Inc., 1999; QUATTRO, 1998; Latu and Everett, 2000). From the perspective of users, quality is measured by the difference (gap) between expected and experienced services. The other perspective measures system performance, i.e. the gap between the providers targeted and delivered quality levels. These two perspectives are interrelated to each other, establishing a quality loop (QUATTRO, 1998) as shown in Figure 1, which can be applied at the firm s level as well as at the whole urban public transport system s level. The quality loop provides a reference framework for fine-tuning the level of quality standards and for optimizing service provision notably on the basis of market reactions, where it may be used at the level of the urban area, covering all public transport systems (multi-authorities, multi-operators). The quality loop can also be used as a management tool at the level of one route/line or one depot and can be used as a corner stone of standard for quality in collective passenger (QUATTRO, 1998). The loop is based on four distinct quality benchmarks, which are used to implement a dynamic process of service improvement. The larger the gap between two benchmarks, the greater the problem is. Improving service quality and efficiency means reducing the gap among these benchmarks. As there are many components in service quality, it is imperative to identify aspects that are considered important from the users perspective. The improvement of service quality is expected to increase customer satisfaction. Satisfied customers will keep using the service and they can also be expected to market and promote the use of public transportation. These are attitudes and behaviors that are associated with customer loyalty in using public transportation (MORPACE Int. Inc. and Cambridge Systematics Inc., 1999). Expected Targeted Measurement of customer satisfaction Measurement of system performance Perceived Delivered Final Customers (passengers and city dwellers) Service Contributors (operators, authorities, etc.) Figure 1. Quality loop at the level of the public transportation system (QUATTRO, 1998) 3. RESEARCH METHOD This research is based on the data collected from three cities, i.e. Jakarta, Bandung, and Yogyakarta. Jakarta is the capital of Indonesia as representation of a metropolitan area with highly complex transportation problems. Bandung is located around 150 km from Jakarta, representing a growing middle-sized city in Indonesia. Bandung is the capital city of the 1066

4 province of West Java, while Yogyakarta is the capital city of the Special Region of Yogyakarta. Yogyakarta is selected as a representative of the middle class which is located far from the capital, where eastern traditional ambience and culture are still strongly maintained. Two types of questionnaire have been distributed off-board to collect the responses, i.e. travel diary and evaluation of service quality. Several places were selected to randomly distribute the questionnaire and to cover all possible users of public transportation modes under study, e.g. common meeting places, terminals, stations, or bus stops. The respondents were limited to public transport users. Public transport in this research refers to land-based public transport. The questionnaire form and detailed questions can be found in Santosa et al. (2008). This article only reports the results of the evaluation of service quality. The questionnaire consists of 50 questions which are contained in five parts, i.e. respondent characteristics, usage of public transport, negative experiences while using public transport, evaluation of service quality, and loyalty. In the service quality evaluation section, respondents were asked to rate variables or attributes related to service quality based on level of importance and level of satisfaction. The list of attributes is presented in Table 1. For the level of importance, respondents could choose highly important, important, not important, and absolutely not important. Weights of 1, 2, 3 and 4 were assigned respectively to those choices. The same weights of 1, 2, 3 and 4 were also assigned to satisfaction levels of very satisfied, satisfied, not satisfied, and totally not satisfied, respectively. However, this study focuses mainly on the level of importance to compare the significance of service quality among the three cities. After examining the completed questionnaires, the total number of valid questionnaires that can be used for analysis are 494, 499, and 498 for Bandung, Jakarta and Yogyakarta, respectively. Table 1. List of variables Code Variable/Attribute Example of Potential Measures* Q1 Cleanliness in the vehicle Presence and amount of trash and dirt, graffiti, or offensive odors. Q2 Cleanliness in the station/stop Q3 Condition of the vehicle Quality of vehicle, physical condition of vehicle and infrastructure, maintenance profile or breakdowns, number of failures, Q4 Condition of the station/stop Quality of seats, physical condition of infrastructure, presence and quality of lightning, seating and telephone, maintenance. Q5 Comfort in the vehicle Ease to enter the vehicle (opening doors), temperature, quality of seats and materials, quality of lightning, ergonomics, space width and heights, noise level, acceleration and deceleration, Q6 Comfort in the station/stop Ease to enter the station, availability of handrails, quality of seats and materials, space width and heights, noise level, Q7 Route coverage Number and length of routes, number of stops, distance between stops, Q8 Ease of payment Available fare types, fare restrictions (exact fare only, surcharges, etc), flexibility of payment methods (cash or non-cash), amount of fare, Q9 Safety and security in the vehicle Number and type of accidents, system rules violations, vehicle travel speeds, driver training, Q10 Security in the station/stop Presence of police, transit staff, emergency-phones, Q11 Access to the vehicle Heigth of platforms, transfer time and number, frequency of connecting services, number of elevators, heigth of stairs, presence and quality of curb cuts, ramps, tactile surfaces. Q12 Driver s skill Driver training and accident records, acceleration and deceleration, punctuality. Q13 Punctuality Waiting time, delay time, vehicle speed, schedule adherence. Q14 Attitude of the staff Presence and knowledge of staff, level of responsiveness 1067

5 Q15 Ease of getting the mode Frequency of service (headway) during weekdays, operating time (hours of service), distance to stop, frequency of service on weekends *most of the explanation of potential measures are followed explanation by MORPACE International, Inc., and Cambridge Systematics, Inc. (1999) 4. RESPONDENTS CHARACTERISTICS 4.1 Socio Demographics Table 2 shows the characteristics of the respondent in these study areas. There is a clear difference characteristic between Jakarta and the other two cities. Jakarta has a higher representation, around twice as many, of full-time workers and people with higher education. On the other side, Bandung and Yogyakarta have higher proportions of students than Jakarta. Yogyakarta and Bandung are well-known for having many good quality universities and, compared with the size of the city, these two cities have more universities, colleges, and other higher education institutions than other cities in Indonesia. Yogyakarta is even known as the Student City (Kota Pelajar). Since the highest level of education attained shows that around half of the respondents in Bandung and Yogyakarta completed high school, given the educational reputations of these two cities, it is reasonable to assume that the majority of the students are undergraduate students. These undergraduate students are usually more attracted to work in big cities such as Jakarta which offer higher salaries and more promising careers after their graduation. This explains the higher number of respondents with bachelor s degree in Jakarta compared to Bandung and Yogyakarta. Table 2. Characteristics of the respondents Characteristics Group Proportion (%) Jakarta Bandung Jogjakarta Gender Female Male years old years old Age years old years old years old > 65 years old Marital Status Single/divorce Married Elementary school Junior High School Highest Level Senior High School of Education Diploma Attained Undergraduate Post-graduate Student Civil servant/military Employment Full-time worker Status Self-employed Housewife Retirement/non-worker N

6 Meanwhile, there are similar distributions of both gender and marriage status. The majority of respondents from the three cities are within the age range of 18 to 29 years old, but the proportions of this age range among the cities are quite different. A similar trend is also observed for respondents under 18 years old. The differences in respondent characteristics among the three cities will be taken into consideration in the analysis. Detailed information on the relationship between respondents socio demographic and travel behaviour is available in Joewono et al. (2009). 4.2 Mode of Public Transport Descriptive analysis is employed to compare the mode from each city to the others. Figure 2 shows that whilst Jakarta and Yogyakarta have significant fractions of bus users, Bandung is dominated by paratransit (angkot) users. This is understandable since, unlike Jakarta and Yogyakarta, Bandung is a city with relatively small and hilly roads which only has 11 bus routes (compared to 33 operated paratransit routes). Paratransit in this article refers to fixed route services (and relatively fixed fare can be negotiated based on distance) but no schedule. Usually it is operated using a minibus-sized vehicle (12-16 passenger capacity), and the service quality (travel time, stopping time, frequencies, etc.) fully depends on the drivers (see Joewono and Kubota, 2007abc, 2008 for more discussion regarding paratransit in Bandung). Jakarta and Yogyakarta have implemented TransJakarta Busway and TransYogya Busway, since 2004 and 2008, respectively. Besides bus rapid transit, both cities have many bus routes and operators, which explains the dominance of buses in Jakarta and Yogyakarta. Paratransit is the second most used mode of transport in Jakarta and Yogyakarta. It is also interesting to note the operation of non-motorized vehicles in Yogyakarta. This city is well known as a cultural city with high traditional ambience. Different than the other two cities, Yogyakarta still maintains its traditional culture to a high degree. Thus, the operation of non-motorized transport is believed to be supportive for this city. 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% Bandung Jakarta Yogyakarta other rickshaw tri-cycle motorcycle taxi taxi rail bus paratransit Figure 2. Type of public transportation modes 1069

7 4.3 Frequency of Usage During weekdays, there are more daily users in Yogyakarta than in Jakarta and Bandung, especially once and twice per day (Figure 3). On the contrary, Yogyakarta has fewer frequent users (three or more times per day) of public transport on weekends. This can be explained by its characteristics as an educational and cultural city; Yogyakarta has fewer industrial and trade activities than Jakarta or Bandung. Bandung and Jakarta have similar characteristics in the number of daily usage on weekdays. Bandung and Jakarta as metropolitan cities rely heavily on economic activities, therefore the number of frequent usages (three or more per day) are high. Most respondents use public transportation twice per day, which indicates that most users are people who commute to office or school. On average, about 40% of respondents (users of public transport on weekdays) do not use public transport during weekends (Figure 4). Jakarta s respondents use public transport slightly more than respondents in the two other cities during weekends. Compared to Yogyakarta, Bandung and Jakarta have higher usage numbers on weekends, maybe due to more economic activities in these two cities than in Yogyakarta Percentage Jakarta Bandung Yogyakarta x 2x 3x 4x or more Number of trips per day in weekdays Figure 3. Frequency of public transport usage in weekdays 1070

8 Percentage Jakarta Bandung Yogyakarta 0 0x 1x 2x 3x or more Number of trips per day in Weekends Figure 4. Frequency of public transportation usage in weekends 5. ANALYSIS OF SERVICE QUALITY ATTRIBUTES 5.1 Analysis of Variance In this section, comparison among variables of service quality is conducted by employing analysis of variance (ANOVA). ANOVA is termed a univariate procedure to assess group differences on a single metric dependent variables (Hair et al., 2006). It provides the tool necessary to judge the differences on a variable. In this study, the variable is a likert-scale rating with large sample size, thus a normal distribution assumption can be hold. The mechanics to do the calculation of ANOVA can be found in many literatures, for example Walpole et al. (2012). Figures 5, 6, and 7 show the ANOVA results for each city, i.e. Bandung, Jakarta, and Yogyakarta, respectively, as well as the individual 95% confidence intervals for each attribute of service quality. These results show that there are significant differences among attributes in each city. This means that each attribute has a different rating, i.e. each attribute has different levels of importance. Thus, it can be inferred that the users in each city have different perceptions regarding the level of importance of each attribute. They do not perceive each attribute as having the same influence to the overall service quality. Different ratings or levels of importance do not necessarily mean that each pair of attributes is different. Significant difference in this ANOVA result means that at least one pair of attributes is different. As the analyses show that there are significant differences, it is possible to conduct further analysis to find out which pair of attributes is different. The author chooses to conduct further analysis by applying factor analysis to find out which attribute has the highest weight or is rated as the most important. Moreover, as there are three different datasets, it is possible to examine whether or not there is a significant difference in the level of importance of the service quality attributes among the three cities by applying ANOVA. The results are presented in Table 3 where, with 5% significance level, six out of 15 attributes have significantly different importance. The six attributes are: cleanliness in the vehicle (Q1), ease of payment (Q8), safety and security in the vehicle (Q9), security in the station/stop (Q10), driver s skill (Q12), and attitude of staff (Q14). This shows that there are different perspectives in the significances of these six 1071

9 attributes in determining the service quality of respondents of the three cities. This also means that at least two cities are different, while the results cannot show which pair of cities is different. The authors choose to analyze further by applying factor analysis in determining the pattern of service quality s attributes for each city. 5.2 Factor Analysis To easily understand how the service quality attributes work together to influence the quality of public transportation and to find out the structure in the relationship between attributes, factor analysis is employed. Factor analysis is an interdependence technique which provides the tool for analyzing the structure of the interrelationships (correlations) among a large number of variables by defining sets of variables that are highly interrelated (Hair et al., 2006). The mechanics and discussions of calculation can be found in literatures, for example Hair et al. (2006) or Warner (2008). This analysis is judged as suitable to show which attribute has the highest priority, i.e. the most importance, in each city. The analysis is also able to show the different patterns among these three cities. Preliminary examination based on correlation matrix results shows that all attributes correlate to each other but not perfectly, as there is no single attribute has a significance value greater than 0.05 and a correlation coefficient higher than 0.9. The determinant value for each city must also considerably high to be confident that multicollinearity is not an issue for these data. One-way ANOVA Source DF SS MS F P Factor Error Total S = R-Sq = 7.41% R-Sq(adj) = 7.24% Individual 95% CIs For Mean Based on Pooled StDev Level N Mean StDev Q (--*--) Q (--*--) Q (--*-) Q (--*--) Q (--*--) Q (--*--) Q (--*--) Q (--*--) Q (--*--) Q (-*--) Q (--*--) Q (--*--) Q (--*--) Q (-*--) Q (--*--) Pooled StDev = Figure 5. Comparison among attributes: Bandung dataset 1072

10 One-way ANOVA Source DF SS MS F P Factor Error Total S = R-Sq = 6.07% R-Sq(adj) = 5.90% Individual 95% CIs For Mean Based on Pooled StDev Level N Mean StDev Q (--*---) Q (--*---) Q (--*--) Q (--*--) Q (--*--) Q (---*--) Q (--*--) Q (--*--) Q (--*--) Q (--*--) Q (--*--) Q (--*--) Q (---*--) Q (--*---) Q (--*---) Pooled StDev = Figure 6. Comparison among attributes: Jakarta dataset One-way ANOVA Source DF SS MS F P Factor Error Total S = R-Sq = 6.18% R-Sq(adj) = 6.01% Individual 95% CIs For Mean Based on Pooled StDev Level N Mean StDev Q (---*--) Q (--*--) Q (--*--) Q (--*--) Q (--*--) Q (---*--) Q (--*--) Q (--*--) Q (--*--) Q (--*--) Q (--*--) Q (--*---) Q (--*--) Q (--*--) Q (--*--) Pooled StDev = Figure 7. Comparison among attributes: Yogyakarta dataset. 1073

11 Table 3. Comparison among cities for each variable Code Variables F-value p-value Q1 Cleanliness in the vehicle Q2 Cleanliness in the station/stop Q3 Condition of the vehicle Q4 Condition of the station/stop Q5 Comfort in the vehicle Q6 Comfort in the station/stop Q7 Route coverage Q8 Ease of payment Q9 Safety and security in the vehicle Q10 Security in the station/stop Q11 Access to the vehicle Q12 Driver s skill Q13 Punctuality Q14 Attitude of the staff Q15 Ease of getting the mode Two tests were conducted to check the appropriateness of the factor analysis, namely Kaiser-Meyer-Olkin (KMO) measure of sampling adequacy and Bartlett s test of sphericity. Small values for the KMO measure indicate that a factor analysis of the variables may not be a good idea, since correlation between pairs of variables cannot be explained by the other variables. Meanwhile, the Bartlett s test of sphericity can be used to test the hypothesis that the correlation matrix is an identity matrix. The significance level should be smaller than 0.05 to reject the hypothesis that the population correlation matrix is an identity. Table 4 shows the results of these two test measurements for the three cities. The values of KMO for Jakarta, Bandung, and Yogyakarta are exceptionally high to be in the category of meritorious (in the 0.90s) or marvelous (in the 0.80s) according to Kaiser (1974). These two tests indicate that applying factor analysis is appropriate for this data in the three cities. Table 4. KMO and Bartlett s Test Test Jakarta Bandung Yogyakarta Kaiser-Meyer-Olkin Measure of Sampling Adequacy Approx. Chi-square Bartlett s Test of df Sphericity Sig By using the principal component analysis method as the extraction method, two factors with eigenvalues greater than 1 were extracted for Jakarta and Yogyakarta and four factors were extracted for Bandung. Furthermore, the varimax rotation method was used to analyze and indicate which attributes were highly related to each factor. The linear combinations of the factors are explained in Equations (1)-(3) and the relations of the attributes to the factors are tabulated in Table 5. Level of importance for Jakarta = 3.856F1J F2J (1) Level of importance for Bandung = 2.875F1B F2B F3B F4B (2) Level of importance for Yogyakarta = 3.923F1Y F2Y (3) 1074

12 Table 5. Relations and loadings of attributes to each factor Code Variable Jakarta Bandung Yogyakarta F1J F2J F1B F2B F3B F4B F1Y F2Y Q1 Cleanliness in the vehicle Q2 Cleanliness in the station/stop Q3 Condition of the vehicle Q4 Condition of the station/stop Q5 Comfort in the vehicle Q6 Comfort in the station/stop Q7 Route coverage Q8 Ease of payment Q9 Safety and security in the vehicle Q10 Security in the station/stop Q11 Access to the vehicle Q12 Driver s skill Q13 Punctuality Q14 Attitude of the staff Q15 Ease of getting the mode Eigenvalue % of variance The grouping of the attributes with the names allocated to the factors can be tabulated for each city to provide better picture as shown in Tables 6, 7, and 8. For Jakarta as well as Yogyakarta, the first factor is the quality of stations and vehicles. The second factor is comfortable and ease of public transportation. These two factors account for about 50% of the variance both for Jakarta and Yogyakarta. The grouping of variables into factors for these two cities are exactly the same only the loading of each variable is slightly different. For Bandung, the first factor is indexed as physical quality of public transportation, the second factor is accessibility and reliability, the third one is security and safety, and the last factor is information and payment. All these four factors explain almost 60% of the variance. Table 6. Grouping and ranking of the aspects of service quality for Jakarta Stations and Vehicles Quality Comfortable and ease of public transport 1 Cleanliness in the station/stop 1 Access to the vehicle 2 Comfort in the station/stop 2 Ease of payment 3 Condition of the station/stop 3 Route coverage 4 Comfort in the vehicle 4 Ease of getting the mode 5 Cleanliness in the vehicle 5 Attitude of the staff 6 Security in the station/stop 6 Punctuality 7 Condition of the vehicle 7 Driver s skill 8 Safety and security in the vehicle 1075

13 Table 7. Grouping and ranking of the aspects of service quality for Bandung Physical quality of public transportation Accessibility and reliability 1 Cleanliness in the station/stop 1 Ease of getting the mode 2 Condition of the station/stop 2 Access to the vehicle 3 Cleanliness in the vehicle 3 Punctuality 4 Comfort in the station/stop 4 Attitude of the staff 5 Condition of the vehicle Security and safety Information and payment 1 Safety and security in the vehicle 1 Ease of payment 2 Security in the station/stop 2 Route coverage 3 Comfort in the vehicle 4 Driver s skill Table 8. Grouping and ranking of the aspects of service quality for Yogyakarta Station and vehicle quality Comfortable and ease of public transportation 1 Cleanliness in the station/stop 1 Ease of payment 2 Condition of the vehicle 2 Access to the vehicle 3 Comfort in the vehicle 3 Attitude of the staff 4 Condition of the station/stop 4 Route coverage 5 Comfort in the station/stop 5 Ease of getting the mode 6 Cleanliness in the vehicle 6 Punctuality 7 Security in the station/stop 7 Driver s skill 8 Safety and security in the vehicle When the loadings of attributes in Table 5 are combined with the Equations (1), (2), and (3), the weight of each service quality attribute for each city can be produced as presented in Equations (4)-(6). These formulations can be used to explain the needed variables (attributes) to obtain high score of service quality. In other words, to obtain high service quality in each city (Fj), which is expressed by high level of importance, the operator would need to have high scores on the related attributes (which is represented by Qi). In rank order, they are displayed in Table 9. Level of importance for Jakarta = Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q12 (4) Level of importance for Bandung = Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q7 (5) Level of importance for Yogyakarta = Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q12 (6) 1076

14 Table 9. Rank order of the attributes for each city Rank by Service Quality Attributes Weight Jakarta Bandung Yogyakarta 1 Cleanliness in the station/stop Cleanliness in the station/stop Cleanliness in the station/stop 2 Comfort in the station/stop Condition of the station/stop Condition of the vehicle 3 Condition of the station/stop Cleanliness in the vehicle Comfort in the vehicle 4 Comfort in the vehicle Comfort in the station/stop Condition of the station/stop 5 Access to the vehicle Safety and security in the vehicle Comfort in the station/stop 6 Ease of payment Condition of the vehicle Ease of payment 7 Route coverage Ease of getting the mode Access to the vehicle 8 Cleanliness in the vehicle Access to the vehicle Attitude of the staff 9 Security in the station/stop Security in the station/stop Cleanliness in the vehicle 10 Condition of the vehicle Punctuality Security in the station/stop 11 Ease of getting the mode Ease of payment Route coverage 12 Safety and security in the vehicle Route coverage Ease of getting the mode 13 Attitude of the staff Comfort in the vehicle Punctuality 14 Punctuality Driver s skill Safety and security in the vehicle 15 Driver s skill Attitude of the staff Driver s skill As shown in Table 9, in general, it can be observed that the rank of attributes in Jakarta is more similar to Yogyakarta than to Bandung. Considering that Bandung is the third largest city in Indonesia after Jakarta and Surabaya, this finding is intriguing. Exploring and examining the similarity and difference in the rank of attributes among the three cities is considered necessary and interesting. The first four attributes for all cities are dominated by attributes related to physical conditions of vehicles and stations/stops, such as cleanliness of stations/stops and inside vehicles, well-maintained facilities, and comfort. This category of attributes was evaluated at the top priority partially because of the poor conditions of the presently available public transportation. An observation from the data shows that the level of satisfaction (data is not presented in this paper) for these attributes is relatively lower compared to other attributes. Attribute Route coverage received higher importance in Jakarta than in the other two cities. The possible reason is that the users of public transport in Jakarta come from wider area, not just the residents of Jakarta. The other two cities have lower importance for this attribute since the two cities are smaller in size. Attribute of safety and security in vehicle has the lowest importance in Yogyakarta, while in Bandung this attribute has the highest. The attribute of ease of payment has the same level of importance for Jakarta and Yogyakarta, while in Bandung has the lowest. The possible reasons is the variety of mode of transport in these two cities which requires more standard method of payment. In Bandung, the majority of public transport is paratransit which means the payment method is the same. Similar reasons can be applied for the attribute of comfort in vehicle. Higher variety of mode imply higher variety in comfort, which means higher requirement from the users regarding the comfort in the vehicle. In Bandung, it has lower importance since there is no comparison of comfort, i.e. dominance of paratransit. Attributes related to personnel: Driver s skill and Attitude of the staff are weighted as not so important, particularly for Jakarta and Bandung and partly for Yogyakarta as displayed in Table 9. For Yogyakarta, the position of Driver s skill is at the bottom of the rank (number 15) and not different from the other two cities. On the other hand, the Attitude 1077

15 of the staff is ranked number 8, relatively higher than the other two cities. The specific characteristic owned by Yogyakarta as a special administration region where the sultanate regime is still exist and well respected may contribute to this. Yogyakarta people still strongly uphold their traditional and cultural values and a good manner is considered very important in the society. 5.3 Discussions In Indonesia, the Government published a law to regulate road transport, i.e. the Law of Republic of Indonesia No. 14 / According to the law, it is an obligation that the government provide the services of public transportation. The government and local government has to foster the operator of public transportation in order to increase the quality of service, safety, security, order, and connectivity of road traffic (article 254). This obligation also covers a task for training, certification, and licensing the operators of public transport. In the article 141 of this law, the operator of public transport should comply with the minimum standard of services, i.e. security, safety, comfort, affordability, equity, and regularity. In the article 215, the operator has several obligations, namely: a. To conduct an environmentally friendly development program for road traffic; b. To provide an environmentally friendly mode of transport; c. To provide a clear, correct, and transparent about the condition of the services of public transport; d. To provide an explanation about the usage, repairing, and maintenance of the mode; and e. To comply with the environmental quality standard. Moreover, in the article 242, the government, local government, and operator of public transport are requested to provide special treatment in road traffic for disable people, senior citizen, children, pregnant women, and sick people. The special treatment means an accessibility, priority service, and facility. In Indonesia, the operation of public transport is regulated by the Ministry of Transportation. Directorate General of Land Transportation at 2002 issued a technical guidance for the operation of public transportation for fixed and frequent route in urban area (Decree No. 687/2002). This guidance contained the principles of managing public transportation in Indonesia, which covers the determination of network and service area, fleet size, tariff, infrastructure and vehicles, bus equipment, and aspects of business. Several parameters are employed to measure route performance and operation, namely load factor, number of passenger traveled, headway, waiting time, travel speed, reasons for delay, mode availability, and fuel consumption. The performance of the infrastructure is measured by terminal and bus stop facilities, traffic management and public transport prioritization, information system, and road network inventory. Specifically, this guidance explained that there are general and specific requirements for the services. General requirements consist of waiting time, distance to reach stop, route change and mode, travel time, and travel cost. Meanwhile, specific requirements cover aspects of service, security, easiness to get bus, and route lanes. Based on these factors, the services are classified into two types, namely economy class and business class. Meanwhile, results of this study shows a different attributes (variables) in explaining service quality of public transport. The gap between the government s point of view and the users point of view reveals a need to adjust the required standard of service quality with the need of users. Since public transport relies mainly on the loyalty of passengers, the government and operators are requested to understand the need of community. This study 1078

16 provides several aspects which is needed to be incorporated in the standard of service quality. Moreover, specific characteristic of users in each city is also needed to be understood by operators and local government. It is not suggested to have a general approach for each city, i.e. only comply with the national standard. Adjustment to local characteristics from national standard can be expected to increase the satisfaction of users in each city. 6. CONCLUSIONS This study has explored the attributes of service quality of public transportation in three cities in Indonesia. The motive is to find out whether there are differences among these cities regarding the attributes of service quality. It wants to answer a simple question: is there any similarity of attribute to be prioritized in each city? If the answer is yes, then it is possible to apply the same approach (policy) to any city in Indonesia. If the answer is no, pollicy must be tailored to the uniqueness of each city. The initial part of this study explores differences in user perception regarding the importance of service quality. The results show that each attribute is perceived as having different levels of importance. Comparisons among cities for each attribute is also conducted, showing that six out of 15 attributes are significantly perceived as different. Factor analysis is also employed as the second part of analysis. Fifteen attributes of service quality were grouped according to their relationships. By comparing the results, it is found that the three cities have similarity as well as disimilarity. Some attributes have high priority in one city but lower weight in another city. Thus, each city s users are unique. Each city has different numbers of group, i.e. two and four groups. This has policy implications for both government and operators. As resources are very limited in Indonesian cities, selecting an appropriate approach to solve problems is essential. Selecting which group of service qualities to target provides an opportunity for the government or operator to provide more focused action, thus best employing its limited resources for maximum effect. It can also be concluded that the existing generalized approach for all cities in a big country like Indonesia is doubtful. A unique approach which considers the city s specific characteristics seems to be more rational. This study shows each city s uniqueness, which implies a tailored-made and city-specific approach. Further study is needed in order to get richer information regarding the real condition of the user and the operation. Multi-discipline exploration seems to be beneficial. ACKNOWLEDGMENT The authors are grateful to the Ministry of Research and Technology, Republic of Indonesia, for providing research fund through the Incentive Program for Applied Research (Program Insentif Riset Terapan). This article is a part of the research project, i.e. The Development of Transportation Policies based on Travel Behavior to Increase the Usage of Public Transportation in Indonesia. The authors also express their appreciation for the help of Muhamad Andi Syafrizal Indrawan and Rd. Nandyarini Prillianti in preparing the article in the initial version. REFERENCES 1079

17 Eboli, L. and Mazzulla, G. (2007) Service quality attributes affecting customer satisfaction for bus transit, Journal of Public Transportation, 10(3), Hair, J. F., Black, W. C., Babin, B. J., Anderson, R. E.,Tatham, R. L. (2006). Multivariate data analysis, 6th ed., Pearson, Prentice Hall, New York, NY. Hensher, D.A., and Brewer, A.M. (2001), Transport: an economics and management perspective, Oxford University Press, New York. Hensher, D.A., and. Prioni, P.A. (2002) Service quality index for area-wide contract performance assessment. Journal of Transport Economics and Policy 36(1), Hensher, D.A., Stopher, P., and Bullock, P. (2003) Service quality - developing a service quality index in the provision of commercial bus contracts. Transportation Research Part A, 37(6), Joewono, T.B., and Kubota. H. (2007a) User satisfaction with paratransit in competition with motorization in Indonesia: anticipation of future implications. Transportation. 34(3), Joewono, T.B., and Kubota, H. (2007b) Exploring negative experience and user loyalty in Paratransit. Transportation Research Record, Journal of Transportation Research Board Issue: 2034, Joewono, T.B., and Kubota, H. (2007c) User perception of private paratransit operation in Indonesia. Journal of Public Transportation, 10(4), Joewono, T.B., and Kubota. H. (2008) Paratransit service in Indonesia: user satisfaction and future choice. Transportation Planning and Technology, 31(3), Joewono, T.B., Susilo, Y.O., and Santosa, W. (2009) Understanding travel behavior of public transportation user in Indonesia metropolitan area, 11th International Conference on Advanced Systems for Public Transport, Hong Kong, July 20-22, Kaiser, H.F. (1974) An index of factorial simplicity. Psychometrika, 39(1), Kittelson & Associates, Inc., Urbitran, Inc., LKC Consulting Services, Inc., MORPACE International, Inc., Queensland University of Technology, and Y. Nakanishi. (2003) TCRP Report 88: A Guidebook for Developing a Transit Performance-Measurement System. Transportation Research Board of the National Academies, Washington, D.C. Latu, T. and Everett, A. (2000) Review of satisfaction research and measurement approaches. New Zealand. MORPACE International, Inc., and Cambridge Systematics, Inc. (1999) TCRP Report 47: A Handbook for measuring customer satisfaction and service quality. TRB, National Research Council, Washington, D.C. QUATTRO. (1998) Quality approach in tendering/contracting urban public transport operations. Final report, EU, RTD Programme. Santosa, W., Joewono, T.B., and Susilo, Y.O. (2008) Pengembangan kebijakan transportasi berdasarkan perilaku pelaku perjalanan dalam rangka meningkatkan penggunaan transportasi publik di Indonesia. Kementrian Riset dan Teknologi, Jakarta, Republik Indonesia. (in Indonesian). Susilo, Y.O., Joewono, T.B., Santosa, W., and Parikesit. D. (2007) Reflection of motorization and public transport in Indonesia: lessons learned from Jakarta Metropolitan Area and future implications towards better transportation development in developing countries. Journal of Eastern Asia Society for Transportation Studies (EASTS), Vol. 7, Sustainable Urban Transport Project (SUTP) (2009), Sustainable urban transport project. Online, Accessed on 15 January Victoria Transport Policy Institute (VTPI). (2009) TDM encyclopedia. Online, Accessed on 15 January

18 Walpole, R.E., Myers, R.H., Myers, S.L., Ye, K., (2012) Probability & Statistics for Engineers & Scientists, 9 th ed., Pearson Education, Inc., Boston. Warner, R.M. (2008) Applied Statistics: From Bivariate Through Multivariate Techniques, International Edition, Sage Publications, Singapore. World Bank (1987) Bus services: reducing costs, raising standards, World Bank Technical Paper Number 68, Urban Transport Series, Washington, DC. 1081