Abstract
This study extends SERVQUAL and ECOSERV models to include interaction and experience attributes in an ecotourism context and to identify how ecotourism service attributes can affect Vietnamese ecotourist satisfaction and dissatisfaction using Tetraclass approach. This study uses a combination of both qualitative and quantitative methods. The results show 47 ecotourism service quality attributes belonging to 10 dimensions and four categories as Basic, Key, Plus, and Secondary. The findings also indicate the 10 most important attributes contributing to ecotourist satisfaction/dissatisfaction and confirm the complex relationships of ecotourism service attributes on ecotourist satisfaction/dissatisfaction. While ecotourism has become one of the most important subsectors in the tourism industry, understanding ecotourists’ evaluations and improving their satisfaction with ecotourism is still mostly underexplored, in particular, in the context of a developing country like Vietnam. Therefore, this study helps managers understand effective ways to improve customers’ satisfaction and eliminate their dissatisfaction. However, it ignores customers’ affective moods. Thus, future studies could benefit to include these aspects to gain a better understanding about the phenomenon.
Introduction
There is a noticeable shift from mass tourism to sustainable tourism services, in which ecotourism has received considerable attention by both tourism researchers and managers (Cini et al., 2017; Pornprasit and Rurkkhum, 2019; Shi et al., 2019). Most previous studies about ecotourism focus its impact on the environment, local economy and community life quality (e.g. Jones, 2005; Shi et al., 2019; Tran and Walter, 2014), ecotourist motivation and experience (e.g. Chan and Baum, 2007a, 2007b; Chiu et al., 2014; Zhang and Lei, 2012), or exploring ecotourist segmentations (e.g. Hvenegaard, 2002). Only a few studies focus on ecotourism service attributes (e.g. Khan, 2003; Lu and Stepchenkova, 2012) as well as their effects on ecotourist satisfaction (Abou-Shouk et al., 2018; Lu and Stepchenkova, 2012) and are mainly based on the SERVQUAL model (Parasuraman et al., 1985) or its modified and extended version, ECOSERV model (Khan, 2003). A lack of unanimous understanding of the nature of ecotourism and its activities has caused confusion and difficulties in designing and organizing tours to attract customers and create customer satisfaction for ecotourists. These obstacles have negatively affected tourists’ experiences as a result of poorly designed and performed services. This shortcoming calls for a new research effort to identify factors that contribute to tourist satisfaction and eliminate their dissatisfaction.
ECOSERV has become the most basic measurement instrument for ecotourism attributes (Ladhari, 2008) and has been applied in most studies in the field (Bagstad et al., 2013). However, the ECOSERV model has some limitations. First, it places more emphasis on physical facilities and equipment rather than endemic characteristics, local recreational activities, or especially ecotourist interactions with a destination or with other ecotourists (Devesa et al., 2010; Kastenholz et al., 2012). The presence, behaviors, and the quality of interactions with others could positively and/or negatively influence the quality of trips (Camelis et al., 2014). Therefore, ecotourism interaction attributes, such as sharing views with people having the same interests (Reichenberger, 2017), must be important dimensions or aspects for creating social values related to personal orientations (Richins, 1994). Moreover, because tourists are likely to perceive tourism service quality as an expressive experience rather than merely as a utilitarian transaction, tourist relaxation and sensual feelings play a vital role in the process of such experience (Chan and Baum, 2007b; Kim, 2010; Pan and Ryan, 2010). Considering these limitations of ECOSERV, this study extends the present literature by including ecotourism experiential attributes, such as sensory feelings and social experiences. In addition, while price–quality inference attributes have proven as good indicators of product/service quality in previous research (Völckner and Hofmann, 2007), they are rarely applied in the tourism service domain, especially with ecotourism. Therefore, this study also contributes to the literature by including price–quality inference attributes and explores its role in the relationships with ecotourist satisfaction and dissatisfaction.
In addition, both SERVQUAL and ECOSERV models implicitly assume a linear relationship between service quality attributes and satisfaction. This means that satisfaction is proportionate to the level of performance of service attributes and that customers feel a high level of satisfaction when all service attributes are desirably fulfilled (Robinot and Giannelloni, 2010). However, this linear nature has been questioned when several studies suggest the presence of a zone of indifference where there is no satisfaction nor dissatisfaction irrespective of the perceived performance of service attributes (Oliver, 2010). This implies that the quality attributes of a service could have asymmetric impacts on satisfaction and dissatisfaction (Lichtlé et al., 2002; Materla et al., 2019). In this line, the Tetraclass model provides deep insight into the distinct characteristics of different groups of service attributes (e.g. Basic, Plus, Key, and Secondary) categorized according to their relationships with customer satisfaction (Llosa, 1996). It also can generate the most effective approach which allows to consider the relationships between service quality attributes and satisfaction/dissatisfaction from different service quality models like SERVQUAL or ECOSERV (Poubanne et al., 2006; Ray and Gotteland, 2005; Robinot and Giannelloni, 2010). This present research is exploratory in nature, as the Tetraclass model has not yet been thoroughly tested in the literature, and this is the first attempt to apply this model in the context of ecotourism. Therefore, this study also hopes to contribute to the literature by applying the Tetraclass model to explore the complex nature in the relationships between ecotourism service quality attributes and ecotourist satisfaction. In addition, the determination of the most important attributes contributing to ecotourist satisfaction and dissatisfaction is also important for practical issues (Robinot and Giannelloni, 2010).
In summary, the purpose of this study is to expand service quality models to explore how service attributes can affect satisfaction and dissatisfaction in an ecotourism context using the Tetraclass model. Therefore, this study makes important contributions to the literature by identifying ecotourism service attributes, in particular, extensive interaction and experience attributes, which affect ecotourist satisfaction and dissatisfaction. In addition, this study also develops a better understanding of the structure of ecotourism service quality by classifying ecotourism service attributes depending on the relationship nature of their performance on ecotourist satisfaction and dissatisfaction. Furthermore, in spite of an enormous potential to develop ecotourism, most studies have focused on how to develop this service in Vietnam (e.g. Lam, 2004), and no study we know explores how Vietnamese ecotourists (Vietnamese ecotourists/or/ecotourists to Vietnam) evaluate ecotourism service attributes and their effects on their satisfaction. However, evidence from other developing countries has demonstrated that understanding how ecotourists evaluate ecotourism service attributes in relation to satisfaction is important to develop effective business and marketing strategies (e.g. Lu and Stepchenkova, 2012; Pornprasit and Rurkkhum, 2019; Shi et al., 2019). Therefore, considering the pivotal role of ecotourism in economic development as well as the growth potential of this sector in Vietnam (e.g. Lam, 2004), it is of great benefit for travel agents, tour operators, ecotourism sites, and even the government to understand its structure and identify service features that offer the best experience to ecotourists. This study uses quantitative methods (e.g. correspondence analysis) to fulfill the purposes.
Conceptual framework
Ecotourism definition
There is no consensus on understanding the nature and characteristics of ecotourism for different definitions of ecotourism existing in the literature. Ziffer (1989) pointed out that ecotourism is “a complex notion which ambitiously attempts to describe an activity, set forth a philosophy and espouse a model of development” (p. 5). Ecotourism involves traveling to relatively undisturbed or uncontaminated areas with the specific objective of studying, admiring, and enjoying the scenery and its wild plants and animals, as well as any existing cultural manifestations (both past and present) found in these areas (Ceballos-Lascuráin, 1991: 25). Ecotourism is also conceptualized as an emphasis on the preservation of natural resources of ecotourism sites and environmental knowledge of tourists, or as community or nature-based tourism activities which have strong links with sustainable development (Dolnicar and Leisch, 2008) and the life quality of poor local people (Lu and Stepchenkova, 2012). These perspectives emphasize the appreciation of the natural environment, conservation of nature, interactive and learning experiences, and local community responsibility. However, most researchers in tourism service marketing (e.g. Bagstad et al., 2013; Ban and Ramsaran, 2016; Ceballos-Lascuráin, 1991; Wurzinger and Johansson, 2006) accepted a definition by Khan (2003) that ecotourism is as “purposeful time spent in natural environment to interact, learn, and experience other cultures, and to economically help local communities that work toward preservation of the ecosystem” (p. 111). However, ecotourism also provides opportunities for experiencing and appreciating nature, local customs, and culture (Bagstad et al., 2013; Wurzinger and Johansson, 2006) and allows tourists to learn about responsible traveling (Ban and Ramsaran, 2016). Therefore, this study extends the definition by Khan (2003) and defines ecotourism as ecotourist motivations for natural, cultural, soft, discoverable, adventurous, and responsible tourism activities and experience. Not only is this definition consistent with the multi-dimensional construct of ECOSERV, but it also allows integration of new ecotourism attributes to expand the ECOSERV model (Bagstad et al., 2013).
Ecotourism satisfaction and dissatisfaction
Customer satisfaction has attracted enormous and continuous attention from both scholars and practitioners because of its role in generating the company’s competitive advantage (Bal and Boehm, 2017; Martinaityte et al., 2019; Mittal et al., 2005; Zhao et al., 2019). Satisfaction has been defined and measured in different ways over the years (Oliver, 2010). There has been a growing tendency to consider customer satisfaction as an aggregate of satisfaction and dissatisfaction at an attribute level (i.e. satisfaction and dissatisfaction toward different components of a product or service) (Slevitch and Oh, 2010). For example, Gardial et al. (1994) demonstrate that customers are more likely to evaluate their post-purchase experiences and describe their consumption outcomes based on attribute-level quality than on the overall offering, acknowledging the multi-faceted nature of service consumption. This approach concedes that different aspects or stages of a service/product production involve different levels of customer satisfaction. Thus, it provides better diagnostic benefits for business managers by identifying areas (attributes) for improvement (Thi and Llosa, 2015; Verhoef et al., 2004).
Most scholars implicitly assume a linear relationship between each constituent attribute and overall satisfaction (Verhoef et al., 2004). In other words, adequate quality at an attribute level leads to a certain level of satisfaction, while the same level of dissatisfaction is engendered by equally inadequate quality. This means that each constituent attribute has an identical weighting in the formation of satisfaction and no intermediate state exists between satisfaction and dissatisfaction (Oliver, 1981). Satisfaction in this case is considered a bipolar concept, implying that the opposite of satisfaction is dissatisfaction (Oliver, 2010). However, previous studies (Kano et al., 1984; Lichtlé et al., 2002; Llosa, 1996; Lu and Stepchenkova, 2012) suggest that the attributes of a service might have an asymmetric impact on overall satisfaction. This means that satisfaction is dimensional with the factors that create satisfaction being independent from the factors that create dissatisfaction. In this line, the Tetraclass model (Llosa, 1996) is the best because it uses a relative approach which considers the relationships between service attributes and satisfaction/dissatisfaction based on customers’ real experiences (Ray and Gotteland, 2005; Robinot and Giannelloni, 2010) (Figure 1).

Two approaches to the asymmetry of satisfaction and dissatisfaction (Llosa, 1996).
Ecotourism service attributes
Because ecotourism is defined as ecotourist motivations for natural, cultural, soft, discoverable, adventurous, and responsible tourism activities and experience, ECOSERV measures ecotourism service attributes based on ecotourist experience including six experiential dimensions (Khan, 2003). Besides the five dimensions from SERVQUAL, namely, reliability, responsiveness, assurance, empathy, and tangibles (Parasuraman et al., 1985), the dimension of eco-tangibles is added (i.e. service equipment and facilities and elements related to the environment) (Khan, 2003; Robinot and Giannelloni, 2010). As mentioned in the Introduction, considering the limitations of ECOSERV, this study includes new tourism service attributes to ecotourism to gain a more comprehensive view about ecotourism service quality such as experience (e.g. sensory, feelings and social experiences) (Chan and Baum, 2007b; Kim, 2010; Pan and Ryan, 2010), interaction (Reichenberger, 2017), and price–quality inference (Camelis et al., 2014; Devesa et al., 2010; Kastenholz et al., 2012; Völckner and Hofmann, 2007). Our extensions are based on the following reasons. First, because experience as a new value attribute is designed, intentionally produced, organized, foreseen, calculated, priced, and charged for, experience attributes add a somehow comprehensive living adventure to the short time the ecotourist spends in his destination (Pine and Gilmore, 1998). Second, experience emerges from the interaction between destinations (e.g. local residents or communities) and ecotourists, and both local communities and ecotourists have to play their own roles depending on the extent of their immersion (Stamboulis and Skayannis, 2003). Third, like every other type of tourism, ecotourism not only concerns learning, but also a process of hedonistic experience (Ryan et al., 2000); ecotourists have to enter into a multifaceted interaction with staff, other ecotourists, and local communities during the visit (Stamboulis and Skayannis, 2003). For example, Chan and Baum (2007b) found that the customer service-related attributes (e.g. the staff were very warm and helpful) were classified as critical for the potential to affect satisfaction/dissatisfaction. Their research also revealed that ecotourists perceived ecotourism as participation in eco-activities (e.g. wildlife viewing, nature-based tours, and local culture). Blamey and Hatch (1998) reported that friendliness and helpfulness of the staff were the elements that gave ecotourists the most satisfaction. Previous studies (e.g. Holden and Sparrowhawk, 2002) also indicated that ecotourists were satisfied when meeting people with similar interests. Finally, empirical evidence showing the relationship between price–quality inference attributes and ecotourist satisfaction/dissatisfaction was also provided (Lu and Stepchenkova, 2012; McKercher, 2002; Weaver, 2011). The 10 dimensions of ecotourism service attributes are described in Figure 2.

Ecotourism service attributes.
The Tetraclass model
Several studies have been conducted to examine the asymmetry between product/service attributes and customer satisfaction and concluded that customers evaluate an offering according to a set of attributes, some of which play crucial roles in determining satisfaction, while others do not contribute to satisfaction if present, but unsatisfactory performance may trigger dissatisfaction (e.g. Kano, 1984; Maddox, 1981). Extending those findings and the unipolar view of satisfaction and dissatisfaction, Llosa (1996) proposed the Tetraclass model with four categories of service attributes according to their contribution to customer satisfaction or dissatisfaction as in Figure 3(a) and (b). A description of the characteristics of each category is provided below:

(a) Tetraclass model (Llosa, 1996, 1997, 1999) and (b) Llosa’s categorization.
The Tetraclass model has been applied to a number of service industries such as automobile finance, catering, health, and pharmacies, all of which have confirmed the existence of the four categories of attributes (Ray and Gotteland, 2005). Compared to other methods, such as Flanagan’s (1954) critical incident method, Kano et al.’s (1984) simulation method, and Brandt’s (1988) method, the Tetraclass model has some distinctive advantages (Bartikowski and Llosa, 2004; Bodet, 2006; Llosa, 1996; Poubanne et al., 2006; Ray and Gotteland, 2005). First, the model is easy to implement as no additional or specific questionnaire is needed. Second, unlike many other models, the Tetraclass model is not based on a simulation but on customer’s experiences to evaluate satisfaction. Third, it respects the ordinal nature of data collected, which are sufficient to obtain expected results. Fourth, scientific categorization comes with visual contribution charts to allow better communication of the results. Finally, the model has the possibility to deal with a large number of service attributes.
In summary, for exploring ecotourism service attributes and their asymmetric effects on ecotourist satisfaction/dissatisfaction, this study first accepts ECOSERV-related service attributes (Khan, 2003) as an effective model for ecotourism (Bagstad et al., 2013; Ladhari, 2008) and also extends it to include the experiential and interactional dimensions as suggested by previous studies (Camelis et al., 2014; Chan and Baum, 2007b; Devesa et al., 2010; Kastenholz et al., 2012; Kim, 2010; Lu and Stepchenkova, 2012; Pan and Ryan, 2010; Stamboulis and Skayannis, 2003; Völckner and Hofmann, 2007). Finally, this study applies the Tetraclass model to explore the complex nature in the relationships between ecotourism service attributes and ecotourist satisfaction/dissatisfaction.
Methodology and results
Study context
While ecotourism development has a quite long history in developed countries (e.g. Whelan, 1991), it is in the first stages in most developing countries, such as Vietnam (Doan, 2000; Magio et al., 2013). Weaver (2011) pointed out Southeast Asian countries have shown slow and unimpressive efforts in terms of their achievement of sustainable tourism despite their abundant natural resources. Vietnam’s natural and cultural potential for ecotourism was well established with 13,000 floral species and over 15,000 faunal species in a transitional area of organisms from the biota of the North (Himalaya–south China), the South (Malaysia–Indonesia), and the West (India–Myanmar) (Khanh, 1999; Luong, 1999). With a 3260-km coastline, Vietnam hosts a variety of coastal ecosystems with 125 benthic species and 158 seaweed species (Tien, 2000). In addition, Vietnam’s forests and mountains are very rich with about 11 million hectares in a tropical region where 54 ethnic minorities live. These communities have a deep understanding of traditional festivals, cultivation and land-use customs, culinary culture, traditional lifestyle and handicrafts, and historical places (De, 2002).
Despite the efforts of the Vietnamese government to promote sustainable tourism as well as the nation’s cultural and environmental values (Vietnamese Government, 2011), Vietnam has not been able to take advantage of the natural beauty and rich culture it has to reach its full potential for ecotourism (Hong et al., 2002; Trang, 2017). The inadequate and uncoordinated management and organization of ecotourism activities is problematic. Ecotourism sites and resources are under the management of various departments and organizations, such as the provincial people’s committees, provincial tourism departments, the Ministry of Agriculture and Rural Development, the Ministry of Culture, Sports and Tourism, and tourism site’s management boards (Nguyen, 2018). The result of this tangled web of oversight is a lack of national strategy for ecotourism in Vietnam, consequently hindering the development of this sector (Hong et al., 2002). The lack of a comprehensive development strategy is accompanied by confusing definitions of ecotourism and unorganized activities. Ecotourism involves education of responsible traveling in which tourists visit and learn about the sites and appreciate the beauty of the natural environment (Bagstad et al., 2013; Wurzinger and Johansson, 2006). Environmental information and education are supposedly provided for inquisitive tourists through tour guides, site managers, and staff. However, in Vietnam, most of these people have insufficient environmental training and knowledge to perform the task (Luong, 1999). A part of the real value of ecotourism lies in the distinctive experience with local people and culture. However, such exposure of tourists to Vietnam’s diverse cultural identities is limited due to the inactive involvement of local people and lack of knowledge of the indigenous culture by most ecotourist guides (Hong, 2016).
Analytical methods
This study combines both qualitative and quantitative methods (e.g. Llosa, 1996; Robinot and Giannelloni, 2010). The qualitative methods aim to explore and to identify ecotourism service quality dimensions and their attributes. The quantitative methods are used to classify the contributions of those service quality dimensions/attributes to ecotourist satisfaction and/or dissatisfaction. In regard to the analytical technique, the Tetraclass model uses a combination of quantitative methods, such as factor analysis, correspondence analysis, and others. With the integrated analyses and aggregating several approaches to survey data analysis, the Tetraclass model is expected to be more effective in increasing the information quality derived from a customer satisfaction survey than most traditional approaches using only one method (Kenett and Salini, 2011 for discussion).
Qualitative research
As discussed above, while ECOSERV is used in this study as a basic framework for assessing ecotourism service quality, further dimensions, including ecotourism experience, interaction, and price–quality inference, are the subjects to be explored. Therefore, an in-depth interview approach was implemented during the summer of 2017 by the coordinators of the present research project with tourists visiting an ecotourism point in Mekong Delta, Vietnam. The participants were clearly informed that the interview aimed to explore tourists’ evaluations about their actual experiences with the ecotourism point as well as their feelings about that ecotourism point. They were then asked to participate as volunteers. Richardson et al. (1999) suggests that, to understand subjective experience, customers need to be asked to talk about what has happened and what has been explored, perceived, and felt. This method helps to elicit firsthand and personal information from those who have actually experienced the service (Richardson et al., 1999). The rationale for using a qualitative approach in this case is that it provides an insight into the complex behavior of ecotourists (Walsh, 2005). The in-depth interview approach has also been used to explore a destination’s service quality attributes in the literature (e.g. Hudson and Shephard, 1998). In-depth interviews were conducted with 22 eco-tourists who have travelled in a national forest or park area in the last two years. The list of eco-tourists was provided by two travel agencies that organized these tours regularly. The sample size of 22 was adequate for qualitative research, as 20–40 in-depth interviews are recommended as sample size for qualitative research (Travers, 2009). An interview guide with a combination of open-ended and semi-structured questions was used, which ensures consistency of the interviews and systematic collection of data. Interview lasted between 35 and 40 min, with seven being conducted over the phone and the remaining 15 being face-to-face. The description of the main characteristics of the informants is presented in Table 1.
Demographic profiles of the respondents of the qualitative research.
Each pair of questions and answers took between 3 and 5 min. After the tourist agreed to voluntarily participate in the review, the interviewers introduced three sets of interview questions. First, the informants were asked about their recent trip at one ecotourism place regarding: (1) what did they see? (2) what did they do? (3) what did they explore in term of ecotourism service quality including the tangibles, environment, and culture of the place? (4) what did they perceive and feel about the staff? (5) what did they feel about the local people? (6) what were their feelings and emotions during their trip? (7) what would they say about the social interactions on their trip? and how did they perceive about the overall service quality of the trip? Next, the participants were asked about critical service quality attributes during their experience. This second step included open-ended a sub-question based on service expectation attributes on ECOSERV and the ecotourism service quality attributes explored in the first step. Finally, the informants were asked about the most positive and negative experiences in order to explore the ecotourism service quality attributes that they may not have mentioned before.
All responses were recorded, transcribed, and analyzed by inductive reasoning techniques (Lincoln and Guba, 1985). For enhanced understanding of the intended concepts, the content analysis recommended by Lauri and Kyngas (2005) was conducted by manual procedures. This method focuses on the content or contextual meaning of the text (Lauri and Kyngas, 2005; Potter and Levine-Donnerstein, 1999). Basically, it works by dividing the many contents of the text into much smaller content units, then classifying the units that share similar meaning into the same category (Bardin et al., 2005). The relevant units contribute to the explanation for each category. After the different categories had been established, we counted the frequencies of the units (i.e. ecotourism attributes) to identify the presence of dominant and non-dominant units. Some units were similar across most of the responses, while others were unique. We noted that convergence in the responses or repetition and saturation started to occur at the 15 interview mark. Seven more interviews were conducted to ensure no new information was being provided.
Following the procedure, similar units were clustered together to produce 10 dimensions (together consisting of 47 units or attributes) including: assurance, reliability, responsiveness, empathy, tangibles support, eco-tangibles, price–quality inference, interaction with other customers, interaction with the locals, and relaxation feelings. The list of 47 attributes was presented to a panel of experts in ecotourism, who were asked to evaluate the pertinence of each attribute in terms of its possible contribution to the formation of overall satisfaction. All those attributes were validated, therefore we included all in the final questionnaire. The first six are consistent with the service quality dimensions as SERVQUAL or ECOSERV, namely, assurance, reliability, responsiveness, empathy, tangibles, and eco-tangibles. However, there were additional service experience attributes found under each of these six dimensions. The corroboration of these findings with the existing service quality or service quality expectation in the literature supports the rigor and trustworthiness of the study. As expected, the four new dimensions are labeled price–quality inference, interaction with locals, interaction with other customers, and relaxation feelings. All attributes or items of those 10 dimensions are described in Figure 4, in which the attributes in italic were found and added by this study. The reliability for all dimensions meets the requirements with all Cronbach’s alpha above 0.7 (Nunnally, 1978).

Ecotourism service quality dimensions.
Quantitative research
Quantitative design
A questionnaire was constructed on the aforementioned 47 attributes. Face-to-face interviews were conducted by the research members with the help of tour operators in Ho Chi Minh City during the spring of 2018. Respondents included 367 ecotourists who had visited ecotourism destinations in Vietnam, such as Ho Tram Lake, Bach Ma garden, and Cao Minh garden, within the year. After eliminating questionnaires with missing or incomplete values, 324 valid cases were used. Table 2 presents the respondent characteristics.
Respondent’s characteristics of the quantitative research.
We explored ecotourism service quality attributes and their categorization for drawing up a contingency table by combining ecotourism evaluations of individual attributes with overall satisfaction. The columns in the contingency table show the modalities of satisfaction scores divided into two groups: satisfaction and dissatisfaction. Ecotourist satisfaction was measured using a satisfaction scale consisting of four items used in previous studies on a five-point Likert scale, such as “in general, how satisfied are you with the ecotour in recent time: 1 = completely dissatisfied, 2 = slightly dissatisfied, 3 = neither dissatisfied nor satisfied, 4 = satisfied, and 5 = completely satisfied”, and “when I think about my experience with a recent ecotour, my feeling is: 1 = very unpleasant, 2 = quite unpleasant, 3 = neither pleasant or unpleasant, 4 = quite pleasant, and 5 = very pleasant” (see Robinot and Giannelloni, 2010). The sum of those four items forms a satisfaction index that expresses the cognitive, affective, and conative facets of satisfaction. We then calculated the mean of this satisfaction score. Individuals with higher scores than the mean were considered as satisfied, while individuals with lower scores than the mean were considered as dissatisfied (Robinot and Giannelloni, 2010).
Quantitative analysis
A contingency table was drawn up. In this sense, we applied the analytical procedure similar to previous studies in the field (Clerfeuille et al., 2008; Llosa, 1997; Robinot and Giannelloni, 2010). The contingency table was completed by combining these overall satisfaction scores with each participant’s evaluations of satisfaction for each of the 47 attributes. Each attribute in the table is shown in two rows, one for positive evaluations and one for negative evaluations. The columns in the table show ecotourist satisfaction, with one column for positive evaluations and one for negative evaluations. Therefore, an ecotourist whose satisfaction index was positive but who evaluated an attribute negatively was recorded in the column for positive evaluation but in the row for negative evaluation of that specific attribute. A factor analysis of the connections in this contingency table was then carried out and a single factorial axis was obtained. This factorial axis was interpreted as a satisfaction axis, along which the different modalities of the overall satisfaction index were arranged. Hence, attributes positioned toward the two extremities of the satisfaction axis contribute more strongly to satisfaction or dissatisfaction, either positively or negatively, than attributes positioned nearer the middle of the axis. The results of the first two stages of the analysis allowed us to plot a graph of the attributes of ecotourism service that contribute to the formation of overall satisfaction (Figure 5) as the result of a correspondence analysis. The correspondence analysis was run with 47 attributes (94 modalities) and two groups (see Figure 5): satisfaction and dissatisfaction (Llosa, 1997). Each attribute was divided into two modalities based on its two levels of evaluation (positive and negative) of ecotourist satisfaction. As Llosa (1996) points out: “each attribute is represented by a point.” The abscissa of that point is the coordinate on the factorial axis of its negative evaluation; the ordinate is that of its positive evaluation. The weight of ecotourist satisfaction index marks the boundaries and, consequently, allows the attributes to be categorized into Key, Basic, Plus, and Secondary attributes. Plus attributes are in the northwest quadrant of the figure; Secondary attributes are in the southwest quadrant; Key attributes are in the northeast quadrant, and Basic attributes are in the southeast quadrant.

The graph of categorization of service attributes in ecotourism destinations.
The categorization result for ecotourism service quality attributes, given in Figures 5 and 6 (in the next page), include four categories: Plus (18 attributes), Key (9 attributes), Basic (15 attributes), and Secondary attributes (5 attributes).

Categorization of service attributes in ecotourism destinations.
It is worth noticing that most of those attributes whose nature seems contradictory can be assigned to Plus and Basic. While the former elicits satisfaction if well performed, the latter causes dissatisfaction if it fails to meet the ecotourist’s expectations.
Regarding Plus attributes, four of them are associated with staff service (their competence and attitude), followed by nature-related elements (two attributes). Two attributes are linked with facilities (cleanness and comfortable) and two attributes “feeling safe and feeling like home” are also likely to generate satisfaction.
As for the Basic, it is found that environmental concerns can be a significant source of dissatisfaction (five attributes). Dissatisfaction is also associated with the quality of service/product and of information provided for tourists (five attributes), the feelings of tourists (three attributes), and price–quality inference (two attributes).
With regard to Key attributes, nine are mainly related to the interactions among tourists, between tourists and local people, local specialties, facilities, and the feelings of tourists (experience a new life and feel free to act).
Another noticeable feature of the result is that a significant number of attributes (13) are classified as Secondary. This means that there are many constituent elements of quality and experience at ecotourism destinations that are neutral in terms of defining satisfaction. Surprisingly, most of them are sympathy and empathy attributes such as courtesy, willingness to help, personal attention, never too busy to help, or sincere interest in solving a problem.
The categorized result shows that most ecotourism service quality attributes concerning the interactions between tourists and nature and between tourists and locals fall into the three groups of Plus, Basic, and Key. The attributes involving ecotourist feelings and experiences, experiencing new lifestyles, and escaping daily routines belong to Basic or Key, while feeling safe and feeling like home are Plus attributes. The attributes regarding staff members all belong to Plus, Basic, and Key. However, while most staff competence attributes are shown as Basic, staff attitude attributes are described as Plus and Key. On the other hand, the attributes of staff being polite and being responsive (e.g. willing to help and never too busy to help) are less of a determinant to ecotourist satisfaction. The attributes of support tangibles and eco-tangibles fall into Plus, Basic (mainly), and Key. Interactions with other ecotourists are considered as Plus and Key, whereas price–quality inference falls into Key and Basic.
Determining the most important attributes for satisfaction and dissatisfaction
The results also reveal 10 attributes with the strongest impact on ecotourist satisfaction and 10 with the strongest impact on ecotourist dissatisfaction as shown in Table 3.
The most important 10 attributes for satisfaction/dissatisfaction.
Notes: The most important 10 attributes for satisfaction and dissatisfaction were extracted on the basis of the attributes’ relative values on the axis of satisfaction in Figure 5.
The top 10 satisfiers involve different service aspects. Ecotourists often value nature and attractive scenery. They are also interested in services that allow easy access to and interaction with natural elements at the destination. Paying attention to visitor feelings (feeling like home and understanding specific needs) is also rewarding. The need for interaction among ecotourists, in particular, can be explained by the social nature of this type of travel. Moreover, cleanness and comfort of the tourism destinations also prove to have a significant effect on satisfaction. Finally, customers are also more likely to feel pleased with satisfactory levels of competence and attitude of staff at the tourism site (materials and facilities visually appealing, comfortable facilities, error-free service, having knowledge to answer questions, and prompt service).
The top 10 dissatisfiers, on the other hand, reveal the high expectations of visitors for some services of the ecotourism destination. People who take an ecotour, first, expect a peaceful environment, which allows an escape from work. Moreover, activities that offer authentic interactions with nature are also taken for important. Results also show tourists’ increased awareness of the appropriateness of the facilities to environmental safety. Discriminatory attitudes of local people also disappoint tourists (you might want to explain what you mean by “Discriminatory attitudes”). Finally, like other types of services, if the information quality and provided service/product quality does not meet expectations, this will displease customers.
Discussions
This study applied the Tetraclass model to classify ecotourism service attributes based on the complex relationship nature between them and ecotourist satisfaction and dissatisfaction in the Vietnamese context. It uses a combination of both qualitative and quantitative methods based on the data of Vietnamese ecotourists at some popular ecotourism destinations. The results show 47 ecotourism service quality attributes belonging to 10 dimensions and four categories as Basic, Key, Plus, and Secondary and indicate the 10 most important attributes contributing to ecotourist satisfaction/dissatisfaction. In particular, this study extends previous studies (e.g. Khan, 2003; Ladhari, 2008) by including six additional dimensions (price–quality inference, interaction with nature, interaction with local people, interaction with other ecotourists, and relaxation feelings) besides the original dimensions of ECOSERV model. Therefore, the findings of this study generate important contributions to both theoretical and practical perspectives. This study is among a few studies (e.g. Ray and Gotteland, 2005; Robinot and Giannelloni, 2010; Poubanne et al., 2006) confirming the suitability of the Tetraclass model (Llosa, 1996, 1997, 1999) for clarifying ecotourism service quality attributes and dimensions. The inclusion of new dimensions of ecotourism service attributes (e.g. ecotourism interaction and experience) helps to give a deeper insight into how ecotourists are satisfied or dissatisfied with ecotourism service. Understanding the nature of each dimension in relation with ecotourist satisfaction and dissatisfaction is important for managers and marketers in developing markets aiming at increasing ecotourist satisfaction and loyalty.
As most previous studies, this study applies and confirms the validity of the ECOSERV model for ecotourism in Vietnam (Ladhari, 2008). However, by including ecotourism’s price, experiential, and interaction attributes, this study extends the ECOSERV model to incorporate six dimensions including price–quality inference and interactions. While price–quality inference is confirmed as an important dimension for service quality (Völckner and Hofmann, 2007), it is the first explored in the context of ecotourism. As mentioned above, interactional dimensions (e.g. interactions with nature, local people, or with other ecotourists; Camelis et al., 2014; Chan and Baum, 2007b; Reichenberger, 2017; Robinot and Giannelloni, 2010) and sensual dimensions (e.g. feeling like at home or experiencing a peaceful atmosphere; Chan and Baum, 2007b; Kim, 2010; Pan and Ryan, 2010) are important for tourism including ecotourism. Thus, creating social values fulfilling ecotourist personal orientations (Richins, 1994) or the co-creation of value for ecotourism (Reichenberger, 2017) can provide a more comprehensive view of ecotourism service quality.
This study is also the first in ecotourism to challenge the assumption of a linear relationship between service quality attributes and satisfaction proposed by both SERVQUAL and ECOSERV models (Robinot and Giannelloni, 2010). This study goes beyond previous studies (Abou-Shouk et al., 2018; Khan, 2003; Ladhari, 2008; Lu and Stepchenkova, 2012) in this field by showing a dynamic role of various attributes of ecotourism service in generating ecotourists’ satisfaction and/or dissatisfaction. Therefore, the exploration and the addition of traditional and new dimensions according to the ECOSERV model and the categorization of those into Basic, Key, Plus, and Secondary according to their impacts on ecotourist satisfaction and/or dissatisfaction are important contributions.
This study has demonstrated that the Tetraclass model is appropriate to describe the complexity of ecotourism service quality attributes and dimensions for ecotourism destinations. Although the Tetraclass model has not yet been thoroughly validated in the literature (Robinot and Giannelloni, 2010) and only a minority of services showing Tetraclass class category variability nullifies the model (Poubanne et al., 2006), the findings from this study support the application of this model at least for ecotourism in the Vietnamese context. Therefore, the primarily exploratory results in this study are important for both academic and practical purposes. While classic regression methods cannot be applied, the use of correspondence analysis (Llosa, 1997) to explore the asymmetric effects of various attributes on ecotourist satisfaction and/or dissatisfaction is a unique approach in the tourism literature. For managerial purpose, the strength of this approach allows to map different service quality attributes according to their relationships with satisfaction/dissatisfaction. Thus, this approach can make ecotourism managers easily see and quickly understand what needs to be changed to improve ecotourism service quality.
This study provides important insights for service experience management. First, considering that all ecotourism managers work under budget constraints, our findings suggest that they should focus on first managing the key and basic elements to avoid dissatisfaction, instead of equally investing resources in ensuring the quality of all elements of experience. In particular, some companies may want to commit considerable funds to enhancing the Plus elements, a way to attract and delight customers and stand out from their competitors. Second, it has been shown that attributes which are destination-specific (such as authentic interaction with nature or escape from life) fall into the three groups of Key, Basic, and Plus, and, thus, need more attention from the management of these sites. Moreover, like many other tourism destinations, the service of staff and price–quality fit seem to have an important impact on customer satisfaction and dissatisfaction, with a majority of the attributes being categorized into Key and Basic. Accordingly, businesses need to assure that such attributes meet the visitors’ demand, especially those of the Basic group.
Some limitations should be noticed. First, this study ignores some control variables, such as ecotourist affective states, which might have biased their overall evaluations. For example, a positive mood is associated with a greater likelihood of evaluating an experience positively, and a negative mood is associated with a greater likelihood of evaluating an experience negatively. Second, we have not identified variations in the analyzed demographic variables of the ecotourists surveyed and their trip goals. It also will be important to test the analytical value of the model in the same sample during different periods to understand the changes of satisfaction of tourists over time.
Footnotes
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This research is funded by Vietnam National University Ho Chi Minh City (VNU-HCM) under grant number C2017-34–01.
