Journal Archive

Destination image can be defined as the psychological understanding of a destination through recognition or the impressions of individual visitors (Alhemoud & Armstrong, 1996; Seaton & Benett, 1996; Fakeye & Crompton, 1991). Destination image often influences tour satisfaction and post-visit behaviors, such as intention to recommend and intention to revisit (Bigné et al., 2001; Castro et al., 2007; Chi & Qu, 2008; Ashworth & Goodall, 1988; Mansfeld, 1992; Chris et al, 1993). Recommendation is a type of word-of-mouth advertising in which an individual shares with others their specific positive experience; thus recommendations may affect the expectations of others as well as the expectations of the travelers themselves (Bendall-Lyon & Power, 2004).

Bigné et al. (2001) found that destination image had a direct effect on the quality of individuals’ visits, their overall satisfaction, their intention to revisit, and their intention to recommend. In addition, these researchers found that the overall quality of the visit had an effect on visitors’ satisfaction and their intention to revisit; satisfaction had an effect on visitors’ intention to recommend; and furthermore, intention to revisit and intention to recommend were interrelated. Castro et al. (2007) conducted a survey of 1,526 visitors to metropolitan areas in southern Spain and found that destination image had an effect on service quality, tourists’ overall satisfaction, and their intention to recommend. Chi and Qu (2008) conducted a survey of 345 visitors to major sightseeing areas in Arkansas and other areas in the U.S. and found that destination image directly influenced attribute satisfaction, both destination image and attribute satisfaction had an effect on overall satisfaction, and both attribute satisfaction and overall satisfaction directly affected destination loyalty. Žabkar et al. (2010) conducted a survey of 1,056 respondents who visited four tourist destinations in Slovakia. They found that destination attributes influenced the perceived quality of a destination’s offerings; in turn, the perceived quality of the destination’s offerings had an effect on both satisfaction and behavioral intentions (i.e., loyalty, intention to revisit, and intention to recommend). Veasna et al (2013) surveyed 398 visitors about the World Heritage area of Taiwan. They found that destination source credibility had an effect on destination satisfaction through both destination image and destination attachment; and furthermore, destination source credibility had an effect on destination attachment through the mediator of destination image.

Some studies have found that not only destination image but also personal characteristics influenced tourists’ future behavior. Assaker et al. (2011) conducted a survey of 450 visitors to Europe and used a latent growth model to analyze their immediate intent to return. They found that both novelty seeking and low satisfaction decreased the tourists’ immediate intent to return; however, a good destination image increased both future and immediate intent to return. Chen & Phou (2013) conducted a survey of 428 foreign respondents who visited the Angkor Temple region in Cambodia. They found that both destination image and personality had an effect on satisfaction, trust, and attachment to the destination; and that satisfaction, trust, and attachment influenced destination loyalty.

A couple of studies examined the relation between ecotourism and revisit intention. Wuling et al. (2011) tried to build a model that explains the relations between visitors’ revisit intention and coastal ecotourism by combining the theory of planned behavior(TPB) and the theory of customer satisfaction. They found that tourist satisfaction and destination image had direct effects on the revisit intention, and the cognition of coastal ecotourism indirectly influence revisit intention. Huang et al. (2012) investigated the influence of perception of ecotourism and recreation experience over willingness to revisit. From the survey of 342 respondents who experienced recreational farm, they obtained the results that perception of ecotourism had direct and positive effects on recreation experience and revisit intention, and recreation experience also influence tourists’ revisit intention.

In sum, destination image performs an important role in influencing satisfaction and future behaviors. In addition, degree of satisfaction can influence various types of future behaviors, such as intention to recommend and intention to revisit.

The purpose of this research is to clarify the interrelationships among Image of Ecological Resources, Tour Satisfaction, Intention to Revisit, Intention to Recommend, and Intention to Participate in Ecotourism. Definitions of main terms in this research are as follows. Image of Ecological Resources (IER) means subjective score given by tourists regarding the natural and environmental resources and landscape of Ulleung-do. Tour Satisfaction (TS) means score given by tourists regarding the degree of overall satisfaction with Ulleung-do’s infrastructure and programs. Intention to Revisit (IRV) means response to whether tourists intend to revisit Ulleung-do in the future. Intention to Recommend (IRC) means scores regarding whether tourists have the intention of recommending Ulleung-do to other people. Intention to Participate in Ecotourism (IPE) means response to whether tourists intend to experience any ecotourism programs with an ecotourism interpreter, which is a touring style differed from the current touring style in the Ulleung-do.

To accomplish this main objective, two conceptual models were constructed, based on the related literature regarding tourism and Ulleung-do (see Figure 1) and the set of hypotheses was tested jointly. In this study, the basic hypothesis was that IER would have an effect on TS, IRV, IRC, and IPE. Model 1 was constructed under the hypothesis that IER would have an effect on IRV, IRC, and IPE through the mediator of TS, which is affected by tourists’ current experiences.

Figure 1. 
Two Models with Nested Relationships

Model 2 was structured based on the hypothesis that IER had direct effects on IRV, IRC, and IPE, as well as indirect effects through TS. For this analysis, Structural Equation Modeling (SEM), a useful statistical tool for identifying both direct and indirect effects among several variables, was used.

The following hypotheses were tested through the SEM models and are listed below:

H1. IER has a positive influence on TS.
H1-1. IER has a positive direct effect on IRV.
H1-2. IER has a positive direct effect on IRC.
H1-3. IER has a positive direct effect on IPE.
H2. TS has a positive influence on IRV.
H3. TS has a positive influence on IRC.
H4. TS has a positive influence on IPE.
H5. IRV and IRC are interrelated.
H6. IRC and IPE are interrelated.
H7. IRV and IPE are interrelated.

Ulleung-do is approximately 160 km off the eastern coast of the Korean Peninsula; as a result, its topography, geological features, ecosystem, and cultural and historical resources are different from those of the mainland. Ulleung-do is an island that was formed by volcanic activity approximately 2.5 million years ago. Its size is approximately 72.82 ㎢. Ulleung-do is a double volcano. A small summit (Al-bong) exists in a caldera (Nari Basin), which is highly valued in the field of geology because of its rarity. The Nari Basin is the only flatland on Ulleung-do; hence, the oldest villages on the island were formed there many years ago. In addition, because of the heavy coastal and fluvial erosion, Ulleung-do has interesting coastal features, such as coastal cliffs, a marine terrace, a sea stack, and abrasion caves.

Because Ulleung-do is relatively far from the mainland, the Island is able to accommodate both subtropical forests and temperate forests in close proximity to one another. As a result, a large percentage of the island is designated and managed as a Forest Gene Preservation District, Natural Environment Preservation District, first grade National Land Environment Assessment or first grade Natural Ecology (preservation) Resource. Nearly 75% of the total area of Ulleung-do is covered with forests that are home to rare plant species, such as beech and the Japanese hemlock, as well as mature Acer okamotoanum, so Ulleung-do is an ecologically worthwhile tourist destination with a high diversity of plant species, especially for its size. Ulleung-do has its own indigenous plants, such as aromatic trees, Seomgaeyagwang, Seomdaenggangnamu; the Nari flower of Ulleung, Seombaekrihyang flower; the Seonginbong Peak primeval forest and wood pigeon habitat. These plants were designated and protected as a Natural Monument in South Korea in order to certify their uniqueness.

Four species of fish, 81 species of birds, 822 species of insects, and six species of mammals inhabit Ulleung-do and the area around it. The number of mammals in Ulleung-do is much smaller than in the mainland, and larger mammals cannot live on the island. Representative cultural resources in Ulleung-do include the Tumakjip and Neowajip, which are dwellings built in the 1880s.

In this study, the considered target population was the group of individuals who have visited Ulleung-do. However, an accurate size of that population was not ascertained, thus, the sample was selected by both convenience sampling and purposive sampling. First, convenience sampling was conducted by the interviewers’ judgments. In the process of convenience sampling, this study excluded the tourists who only stopped over at Ulleung-do on their way to visit Dokdo, another island further to the east. To collect the data, four trained interviewers conducted a survey via face-to-face interviews for three days (Sept. 12, 2012–Sept. 14, 2012), querying tourists who had just completed an Ulleung-do tour. A total of 410 surveys were collected, but only 363 were used in this analysis after excluding insincere answers. A frequency distribution, factor analysis, reliability analysis, and structural equation model (SEM) were all conducted using SPSS (PASW statistics Ver. 17.0) and AMOS (Ver. 18.0)

The questionnaire consisted of 34 items. The questionnaire was composed of three sections: a) Characteristics of Respondents; b) Actual Conditions of the Ulleung-do Tour, overall satisfaction with the tour, IRV, and IRC; and c) IER and IPE in the future.

Correlation analysis was conducted to verify the relationships IER, TS, IRC, IRV, and IPE. Correlations were also explored among Clean Environment(CE), Historic Resources(HR), Natural Landscape(NL), Animals and Plants(A&P), Volcanic Topography(VT), TS, IRC, and IRV. The results showed that most variables were positively correlated at the 0.01 significance level, except there was no correlation (r=0.077) between VT and TS. The correlation coefficients among independent variables TS, IRC, IRV and IPE did not exceed 0.8, which means multicollinearity was not expected. IPE is positively correlated with A&P, VT, and TS at the 0.05 level, but with IRV at the 0.01 significance level. Based on these results, it can be expected that a good IER would have a positive effect on TS, IRV, and IPE.

To test whether five items were sufficient to represent IER, an exploratory factor analysis was conducted(See Figure 2). The values used for the selected fit indices were χ²=5.097, p=0.165, CFI=0.99, and RMSEA=0.04. These results clearly indicated that the single factor model for IER explained the data well without any autocorrelation, thus, the reliability was reasonable (α= 0.804).

Figure 2. 
SEM Measurement Model for Image of Ecological Resources

In this study, two structural equation models (SEMs) were tested to identify the effects of IER on TS, IRC, IRV, and IPE in the future. As shown in Figure 3, both SEMs could be useful because both models satisfied a guiding principle (CFI > 0.90) that is considered to indicate a reasonably good fit (Hu & Bentler, 1999); and both had RMSEA values between 0.05 and 0.80, which suggests a reasonable error of approximation (Browne & Cudeck, 1993). Models 1 and 2 were situated in a nested relation and both models could be compared directly with a chi-square value comparison. The chi-square value of Model 2 was 20.665, which was higher than that of Model 1, and Model 2 had two degrees of freedom, which was lower than that of Model 1. Because the criterion chi-square value for two degrees of freedom at the 0.05 level is only 5.99, Model 2 was seen to be excellent, compared with Model 1. In addition, Model 2 had a higher CFI and a lower RMSEA than Model 1; therefore, in view of this stability, Model 2 was considered to be better than Model 1. In the two models, there was no direct relationship between TS and IPE. In Model 1, IRC had a positive relationship at the 0.1 significance level with the intention to participate, but in Model 2 this relationship disappeared because of the direct influence of IER on IPE(see Figure 3).

Figure 3. 
Model Fit and Hypothesis Verification for Model1 and Model2

From the parameter estimates in Model 2, eight hypotheses could be verified. First, IER was evaluated higher, correspondingly, so was TS(H1), IRV(H1-1), and IRC(H1-2). IPE(H1-3) also received a high score(see Table 10). Second, the higher TS score, the higher the scores of IRV(H2) and IRC(H3). Finally, IRV and IRC were interrelated(H5), and IRV and IPE were interrelated(H7).

In brief summary, IER had a positive effect on the current level of TS as well as on future behaviors, including IRV, IRC, and IPE in the future. TS mediated the effect of IER on IRV and IRC.

IER influenced IRC, but this was mediated by TS; however, the image of ecotourism had a more direct effect on IRV (See Table 7). In addition, IPE was only directly influenced by IER, without any indirect effects.