Cristina Esteban Alberdi
Universidad Autónoma de Madrid

Luis Rubio Andrada
Universidad Autónoma de Madrid

José Miguel Rodríguez Antón
Universidad Autónoma de Madrid

Resumen La adaptación de las agencias de viajes a los patrones de la Nueva Economía nos conduce a una tendencia hacia la reestructuración del sector, centrada en la búsqueda de nuevas estrategias competitivas. La creciente importancia de los activos intangibles para la consecución de dichas estrategias ha motivado este estudio, que trata de identificar aquellos factores intangibles, presentes en la prestación de este servicio, y que están directamente relacionados con la percepción, por parte del cliente, de una diferenciación en el servicio recibido. Abstract Adaptation of travel agencies to the standards of the New Economy is leading to a trend towards restructuring the sector aimed at finding new competitive strategies. The growing importance of intangible assets in achieving such strategies has motivated this study, which has tried to identify those intangible factors, present in the provision of the service, that have a direct relationship with the client´s perception of a differentiation in the service received.

1.- Introduction   The new trends in tourist markets, the fact that consumers are now more demanding, technological innovations and recent studies concerning service quality lead the companies in the sector a continuous effort to present new approaches. Intermediary companies in the tourist industry are a clear example of this need for the continuous development of new products and services focused on providing the client with added value.   In this context, those intangible aspects related to service become more important and can be seen as key elements in providing the client with added value (Martinez y Salanova, 1999).   The increased complexity of the situation and the necessity to deal with the subsequent change involves a clear commitment to innovation as the sole way of finding and achieving competitive and lasting strategic approaches. The pillars supporting this innovation are ever more intertwined with those elements that really contribute to its development. These, include information, knowledge, creativity and learning, being the real reference parameters of what is known as the "knowledge Society", supported by the potential of technologies that permit higher levels of coordination, communication and collaboration.   The development of the Information Society and its evolution towards the Knowledge-based and Learning Society is leading to new scientific approaches, social models, systems and technical platforms to transmit and manage the knowledge of the agents, both in the scientific area and in the productive and social ones.   These situation demands a change in the very definition of the work done in present-day travel agencies, which must, in some cases, start and in others go into greater depth with the adaptation process. Travel agencies must focus their efforts on offering new, more personalised services on line according to the trends that can be seen in the markets. This study aims to investigate in which direction efforts to improve service quality in the sector should head, analysing some of the intangible aspects present in the delivery of the service and their relationship with providing the end client with greater value. 2.- Intangible factors analysed   Perceived quality of service is a complex and multidimensional concept, and is made up of various factors, based on different experiences with the agents who provide the service and on different characteristics related to the delivery of the said service (Esteban-Talaya y Millán, 2002). In the specific case of travel agencies, there are some previous studies that support, on theoretical and empirical grounds, the multidimensional nature of the assessment of the service made by the consumer, like Leblanc (1992), Albrecht (1996), Bigné & Andreu (1997), and Esteban-Talaya & Millán (2002) .   Considering the above mentioned researchs, this study will test whether certain intangible attributes present in service delivery in a travel agency are related to the perception of differentiation in the service received.   The measurement scale used is based on the quality dimensions verified on the SERVPERF scale, for measuring service quality, (Cronin & Taylor, 1992).   A previous questionnaire (Esteban, 2003)1 was designed to adapt it to the services offered by travel agencies. As a result of the adaptation we obtained the final questionnaire, deleting certain items and modifing the wording of others, focusing on the intangible aspects and adding aspects related to the perception of differentiation in the service received. Finally, the questionnaire consisted of four quality dimensions, these being: perceived differentiation, reliability, response capacity or empathy and professionalism of the agent. Each of these dimensions is divided into several items for in-depth analysis seen in Table 1. Table 1: Questionnaire Dimensions and Items   3.- Research goals and hypotheses   The goals of this study were: on first, to analyse the relationship between intangible aspects present in the delivery of the service and the perception of a differentiation in the service received by the client and, on second, to show to what extent the clients´ quality assessments depended on other related aspects, such as the type of client, the type of agency, the type of trip planned and use of Internet´s.  Our working hypotheses can be stated as follows:   H1: The assessment of service quality made by the client looking for a leisure trip depends on a wide variety of factors related to the perception of differentiation in the service received.  H2: In the perception of differentiation, of the client looking for a leisure trip, some factors have a greater weight than others.   H3: The assessments made by the clients are related to individual differences and the type of agency where the purchase was made. 4.- Methodology   The data for this study is part of a much broader research within the framework of service quality (Esteban, 2003). In order to carry out this study we used the data obtained through surveys carried out in the Autonomous Community of Madrid. The population from which the sample was taken was defined as: individuals who had used the services of a travel agency in the Autonomous Community of Madrid in the three months prior to carrying out the survey, and whose reason for travelling was leisure.   In order to determine the required sample, the only data available concerning the use of travel agencies was the Familitur Statistic (IET, 2000). It showed the percentage of trips made by Spaniards that were organised through travel agencies. These figures show that 15% of the trips made by Spaniards with destination in Spain and 57% of the trips made abroad were organised through agencies and 67% of all trips were for leisure purposes. The total number of trips, made by Spaniards, was 123 millions of trips (IET, 2000) and therefore the required sample was very high.   Due to the economics limitation of our study, a sample consisting of 150 individuals was chosen it adequate, following the criteria of similar studies in this field (Millán, 2000). The sample was allocated taking into account the percentages used for each age stratum used by the IET (1999)2 in its studies of the make-up of the population of Spanish travellers. After the processing and sifting of the questionnaires, the sample was reduced to 133 valid questionnaires, all completed between September and October 2001. A 7-point Likert scale was used for the assessment. The sample description appears in Table 2. Table 2: Sample Data   Regarding the statistical analyses, we first studied the construct validity of the questionnaire and considering it convergent and discriminant validity. To study the convergent validity of the scale, we obtained the linear associations between the different variables of the same by using Pearson´s correlation coefficient, which allows the study of the degree of linear association between each pair of variables. We also carried out an exploratory factor analysis, which served to ratify that the scale displayed convergent validity.   To check the discriminant validity of the scale, we decided to carry out Pearson´s chi-square test and check whether the quality assessments were related to the use of the Internet, the type of purchase made or the type of agency.   The casual model was studied using Bentler´s covariance structure model (EQS). It seems to be the most appropriate one for studying models with latent variables. This model, created by Bentler (1995), intends to offer a general approach to linear structural equation systems and is the basis of the EQS program.   After these descriptive analyses of the questionnaire questions, other analyses were carried out based on drawing up a synthetic index for them as a whole. This was created using the principal component method and considering:   - The weight of each question in the general opinion of those interviewed in the study.   - The existence of significant differences between the index values for the different groups or categories of each one of the classification variables to be compared by using of variance analysis. This allows a profile of those interviewed to be drawn up if differences do exist.   A summary of our data analyses and the statistical packages used appear in Table 3. Table 3: Data Analyses and Statistical Packages used 5.- Factors related to the perception of differentiation   After carrying out the correlation analysis, we saw that all the variables were highly and significantly related to the perception of differentiation, with the exception of variable 7. Table 4 shows the correlations between the different components of the scale. Table 4: Correlation Matrix   We confirmed our first hypothesis: quality assessments depend on multiple factors related to the perception of differentiation. Specifically, in our study of the response capacity, the professionalism of the agent and their reliability are factors that are directly related to the perception of differentiation of the client who comes looking for a leisure trip.   To further analyze the results obtained, we decided to carry out a causal analysis using Bentler´s covariance structure model for studying relationships in models that accept latent variables. We used the EQS program and obtained statistically significant relationships for all the individual coefficients.   The goodness of fit tests were also significant, obtaining a CFI (Bentler´s chi-square statistics for the Independence) of 0.980. Only six iterations were needed for the convergence of the model. The model obtained is that shown in Figure 1.   In our case, the dependent factor is the differentiation perceived in the service (F4). The independent factors are factors 1, 2 and 3. Factor 1 corresponds to the adaptation of the service provided to the client´s needs, Factor 2 is the professionalism of the agent and Factor 3 is the reliability in relation to the quality of the products offered.   We can see in Figure 1 that the highest coefficient is that relating factor 2 to factor 4 (0.877). We can, therefore, state that the advice and help offered by the agent (F2) is specially relevant for service delivery in travel agencies and clearly contributes towards increasing the degree of differentiation of the service received that is perceived by the client in the leisure travel segment (F4).   The coefficient between Factor 1 and Factor 4 (0.814) is also very significant. The capacity of the agency to respond to the tastes and needs of the client is related to the perception of differentiation. Agencies should be really interested in getting to know their clients in order to offer the product best suited to each of them. At the same time, the coefficient relating reliability in relation to the product with the perception of differentiation is also significant (0.818). Finally, the model shows the fact that the capacity for adaptation to the client´s needs, the professionalism of the agent and reliability are interrelated. 6.- Overall satisfaction index   The overall satisfaction index is related the principal components method. It is appropriate for the goal sought and has the advantage of not requiring a priori conditions regarding normality, independence, and homoscedasticity as in other statistical techniques like factor analysis, canonical correlation and variance analysis, among others.   After applying principal component method, we observed that the first component had an eigenvalue of 7.774, making up for 64.785% of the total variance, with the second component falling to an eigenvalue of 0.884, making up for 7.369% of the total variability. Thus, only considering the first component, we explained more than half of the total variance. This represents a relevant satisfaction index with the service of the travel agencies (ISAV). As we know, the first component has all the coordinates with same sign and is interpreted as a weighted average of all the items or as an overall size factor .   Table 5 shows the correlations between the stated variable and the first component and also the coefficients for calculating standard ISAV scores. In the second column of this table we can see that all the correlations are statistically significant, and that almost all of them show high values, except for question 7 ("On other trips, quality specifications coincided with those found at the destination"), which was under 0.5. If we look at the last column, we can see that the coefficients are above 0.1 in all cases, except for the questions in the third group referring to reliability relating to the quality of the products. This shows that the group mentioned has a lower relative weight in configuring the overall opinion on the subject. Table 5: Importance of the questions in the ISAV   These results lead us to confirm our second hypothesis: "Certain aspects have a relatively greater weight than others in the perception of differentiation of the client". We can conclude that those aspects related to the capacity to adapt to the needs of the client and the professionalism of the agent have a relatively greater weight in the perception of differentiation of the service when compared with the other aspects studied. 7.- Incidence of the classification variables   Standard ISAV scores were calculated using the coefficients in Table 5. These scores allowed us, considering them as a dependent variable, to study the existence of statistically significant differences between the different categories groups- contained in each of the classification variables -factors- under consideration.   This was done using consecutive variance analyses for a factor. However, contrary to the principal components method, this analysis demands compliance with the starting statistical hypotheses. The assumption of normality, homoscedasticity and independence have to be fulfilled. Infringement of one or more of these, means that the interpretation of the F statistic -or t in the case of a factor with two groups- is seriously limited.   In our case, the assumptions were contrasted using the Kolmogorov-Smirnov test for one sample for the study of normality, Levene´s test to check equality of variance, and the residual graphs for independence. Significant results were obtained in all cases, meaning that all assumptions were accepted or statistically met .   Once our assumptions had been confirmed, variance analyses were carried out, one for each of the factors. The last two columns of Table 6 show the values of the F statistic and its significance. Once more we can see that, when the significances are clearly greater than 0.05, as in sex, Internet and number of trips, we cannot reject the equal average assumption for each of the groups or levels. This implies that there are no significant differences between the different groups that make up each of these three classification variables. Age and type of purchase are close to the limit, meaning that the mentioned classification variables may show certain differences. More specifically, age, using the Bonferroni test, shows differences between the first two age groups and the third. At the other extreme we have the type of travel agency, where the null hypothesis is clearly rejected, meaning that the type of travel agency chosen plays a relevant role in configuring the overall opinion. Table 6: Variance analysis for the different factors According to these results, we cannot fully confirm our third hypothesis, as we have not found significant differences between the different types of clients. However, we did find them for the type of agency the person interviewed had visited, with the specialised agencies obtaining higher scores in all aspects when compared with other types of agency. 8.- Conclusions   Throughout this study we have tried to construct and validate a scale to measure the service quality in travel agencies for the leisure travel sector and relate the quality dimensions studied to the perception of differentiation by the client.   As a result of the analyses of the relationships between the dimensions of the scale, we finally propose a model made up of four interrelated factors. In our model, the perception of differentiation of the service received in a travel agency is related to the capacity to adapt to the specific needs and tastes of the client, professionalism when giving advice and reliability.   Drawing up the overall ISAV index allows us to see which aspects have a greater relative weight in the perception of differentiation by the client. We can also use it to confirm that most of the classification variables, except for the type of agency, do not affect overall satisfaction.  The search for competitive strategic approaches has always been a constant organisational task, in large part represented by its capacity to adapt to the overall and specific trends of its context. Future lines of study might include the identification of other criteria that are relevant to other segments of the market and to our segment, by applying qualitative techniques that could provide complementary information. Notas a pié de página 1 The test prior to drawing up the final questionnaire can be found in the Doctoral Thesis: "La calidad en los servicios de intermediación turística. Una aplicación empírica a las AAVV de la CM para el segmento de los viajes de ocio" (Esteban, 2003). 2 In our sample we do not take into account underage travellers (IET also provides data for underage travellers). References ALBRECHT (1996): International HRM: Managing diversity in the workplace. New York. BENTLER, P.M. (1995): EQS Structural Equations Program Manual. Encino, C: Multivariate Software, Inc. BENTLER, P.M., & WEEKS, D.G. (1980): ´Linear Structural Equations with Latent Variables´, Psychometrika, 45, 289-308. BIGNÉ, J. & ANDREU, L. (1997): ´La cadena de comercialización del producto turístico´, in Planificación territorial y comercialización turística. Universidad Jaume I, 3, pp. 113-133. BIGNÉ. (1999): ´La calidad en los servicios en las agencias de viajes´, in Turismo: Organización y administración de calidad, servicios y competencia, I Congreso Universitario de Turismo. Valencia. CRONIN, J. & TAYLOR, S. (1992): ´Measuring Service Quality: A Reexamination and Extensión´, Journal of Marketing, 56, 55-68. ESTEBAN-TALAYA, A. & MILLÁN A. (2002): ´Nuevas tecnologías y comercio electrónico en turismo´, Economistas, 92, 46-52. ESTEBAN, C. (2003): ´La calidad en los servicios de intermediación turística: una aplicación empírica a las AAVV de la CAM para el segmento de los viajes de ocio´ (Doctoral Thesis), Madrid, 2003. ESTEBAN-TALAYA, A., MILLÁN, MOLINA & MARTÍN (2001): ´Factores determinantes de la satisfacción en los servicios de las agencias de viajes´ XIII Encuentro de profesores universitarios de Marketing. Málaga. IET (1999, 2000 and 2001): Statistical Database, www.iet.tourspain.es. INE (2002): Tempus Database, www.ine.es/tempus. LEBLANC, G. (1992): ´Factors Affecting Customer Evaluation of Service Quality in Travel Agencies´, Journal of Travel Research, 4, 10-20. LEWIS, I. & TALALAYEVSKY, A. (1998): ´The Impact of Information Technology on Travel Agents´, Transportation Journal. Summer, 26-30. MARTÍNEZ I. & SALANOVA M. (1999): ´Percepciones sobre calidad´, II Congreso de Universidad y Empresa. Benicasin, pp. 355-370. MILLÁN, A. (2000): Análisis de la satisfacción de los usuarios con los servicios con los servicios de las agencias de viajes (Doctoral Thesis). Universidad de Castilla La Mancha. SUÁREZ, L., VÁZQUEZ, R. & DÍAZ, A. (2002): ´Determinantes de las relaciones estables en el ámbito turístico: una aplicación a las agencias de viajes minoristas´, IX Congreso de ACEDE 2002. About the Authors Autor: Cristina Esteban Alberdi Dirección: Universidad Autónoma de Madrid. Carretera de Colmenar Viejo, Km. 15, 28049 Madrid, Spain Teléfono: 91 397 5243 Fax: 91 397 4218 Correo electrónico: cristina.esteban@uam Autor: Luis Rubio Andrada Dirección: Universidad Autónoma de Madrid. Carretera de Colmenar Viejo, Km. 15, 28049 Madrid, Spain Teléfono: 91 397 5243 Fax: 91 397 4218 Autor: José Miguel Rodríguez Antón Dirección: Universidad Autónoma de Madrid. Carretera de Colmenar Viejo, Km. 15, 28049 Madrid, Spain Teléfono: 91 397 5243 Fax: 91 397 4218

DOCUMENTOS DE TRABAJO EN ANÁLISIS ECONÓMICO (EAWP)
Derechos reservados 2002. El permiso para reproducir algún artículo está garantizado si Documentos de Trabajo en Análisis Económico lo acredita, las copias no son vendidas y es en acto de mayor difusión del documento.

Editor: Fernando González-Laxe. (Universidade de La Coruña)
Director: Venancio Salcines. (Universidade da Coruña)
Subdirector: Andrés Blancas. Instituto de Investigaciones Económicas (UNAM)
Editor Asociado para America Latina: Luis Miguel Galindo. Facultad de Ecomomía (UNAM)