Following the methodology described in the previous section, we collected attributes and tested the scale. The scale was tested by collecting data on a WTD in Georgia. This study resulted in a WTDI scale that can be used in future to measure the WTDI of any wine region. Below we describe the research in detail as well as its results.
3.2. Qualitative Data Collection to Gather More Attributes
The next step in the scale design was to collect the data with qualitative research. The research instrument was focus group interviews with 47 respondents. At this stage, it was decided that the study subject would be wine tourists who had traveled to wine regions and/or participated in wine tourism activities at least once in the past three years. Interviews were held in April and May of 2022. The respondents were found online through social media, and interviews were held via Zoom. We posted in Facebook groups related to wine tourism and travel, as well as general travel- and tourism-related groups. The post asked people to volunteer as participants in focus group interviews if they had traveled to wine regions and/or participated in wine tourism activities at least once in the past three years. This question was once again asked before starting the interviews; anyone who had not participated in wine tourism activities in the past 3 years was not recruited for the interviews. In each focus group, there was an average of 3–4 participants. We recorded the interviews to use the script later during data analysis. We asked the respondents two questions. Overall, 567 words and short phrases were collected after we manually scripted the interviews. The nationalities of the sample varied (from all the continents).
The questions that the respondents were asked to gather the characteristics of the regions were adapted from Echtner and Ritchie [
10], as follows:
By asking these questions, we were able to collect data about functional and psychological holistic elements of the destination image perceived by wine tourists. This information helped us to collect a list of attributes for the WTDI scale that we aimed to develop. This step was necessary as using only a literature review does not ensure a full list of the attributes.
In this section, wine tourists were asked to provide their images of five wine regions as travel destinations. A total of 62% of the respondents answered “no” when we asked whether they had visited the wine region in question. Our objective was to develop a scale that can measure perceptions of the people who have visited the WTD before and perceptions of the people who have not. This is why we included responses from both groups of wine tourists. We chose diverse wine regions to ensure that the final scale would be relevant to different kinds of wine tourism destinations globally. The wine regions were Mendoza (in Argentina), Napa Valley (in USA), Barossa Valley (in Australia), Marlborough (in New Zealand), Kakheti (in Georgia), Colchagua Valley (in Chile), Tokaj (Hungary), Peloponnese (in Greece), Chianti (in Italy), and Stellenbosch (in South Africa). A different group of five regions from the ten was used in the interviews.
3.5. Quantitative Data Collection and Analysis
We then used the quantitative method with a survey as an instrument. The goal of this questionnaire was to validate the scale. The online survey had closed-ended questions. The survey was designed in Google Forms. It was posted in different social media groups to collect responses between December 2022 and January 2023. Similar to the qualitative data collection, we posted on Facebook groups related to wine tourism and travel, as well as general travel- and tourism-related groups. In the post, we asked people to fill in our survey, which would take approximately 7–8 min and would be about WTDI. We also mentioned that Georgian nationals could not participate. As a result, the nationalities of the sample were varied but excluded Georgians. The questions of the survey were grouped into different sections. The questionnaire was tested on 20 respondents to eliminate any bias. We slightly corrected the survey after the test.
The questions of the interview were grouped into different sections. The first section asked whether respondents had ever visited Georgia.
The second section gathered demographic information such as nationality, age, gender, education, marital status, and occupation. In this section, we also asked an extra question to determine the ratio of wine tourists to the total number of respondents:
Have the respondents visited any region in the past 2 years with the purpose of visiting vineyards, wineries, wine tasting, consuming and/or purchasing wine or attending wine events?
The third section asked respondents about characteristics of Georgia as a WTD. We used a 7-point Likert answer format, from the answer “strongly disagree” to the answer “strongly agree”. As we expected that most of our sample had not visited Georgia, we decided to add an additional answer, i.e., “no opinion”. This answer allowed non-visitors to skip a question if they had no information from secondary sources regarding a specific attribute we were asking about. Later, during the analysis, we only included responses on the 7-point Likert scale and disregarded “no opinion” entries. The statements that respondents rated were worded in the following format:
I think that as a wine tourism destination, Georgia has interesting history/customs/culture.
I think that as a wine tourism destination, Georgia is easily accessible.
I think that in a wine tourism destination Georgia it’s easy to communicate with locals.
We collected 298 responses to our questionnaire. Most of our results are based on non-visitor perceptions, as 85% of the respondents answered that they had never visited Georgia before. In terms of age, most of our respondents were between 18 and 24 years old (41%), 38% of them were between 25 and 34 years old, and 13% were between 35 and 44 years old. We received the fewest answers from other age groups, with 4% being from people between 45 and 54 years old, 3% of people were between 55 and 64 years old, and only 1% of the respondents were 65 years or older. A total of 60% of our respondents were female and 38% were male; 2% did not wish to answer. Most of the respondents had Bachelor’s (39%) or Master’s degrees (30%). Most of our respondents were students (51%) or employees (39%). In terms of marital status, most of the respondents were either in a relationship (35%), single (43%), or married (17%). A total of 9% of the respondents were never involved in wine tourism, while the rest had been involved. The demographics of our sample is shown in
Table 3.
We used SPSS to analyze the data that we collected using the questionnaire. In the beginning of the process, we checked the convenience of factor analysis. We wanted to examine how suitable our data was for factor analysis. We found the value of the Kaiser–Meyer–Olkin (KMO) test for sampling adequacy was 0.967. As this value was close to 1, it means our data were suitable for factor analysis.
We also preformed Bartlett’s sphericity test to determine whether there was a strong enough correlation in our data to use factor analysis and principal components analysis (PCA) to reduce dimensionality. Bartlett’s sphericity test determines whether the correlation matrix of the variables is an identity matrix. The correlation matrix was not an identity matrix in our analysis, as shown by the Bartlett’s test result, which also reveals an approximate chi-square value of 24,721.075 with 2415 degrees of freedom and a
p-value of 0.000; this indicates that the data can be used for factor analysis. The results of KMO and Bartlett’s test are shown in
Table 4.
To analyze the data and reduce dimensionality, we used factor analysis (FA) including PCA. Promax with Kaiser normalization was used to standardize the data before factor analysis. This rotation technique gave us the cleanest results.
In our analysis, six components had eigenvalues greater than 1 and they were retained. Together, these six elements accounted for 71.66% of the variance, which indicates that they represented the most important variables in the dataset. Our minimum factor loading was set at 0.3.
We measured the internal consistency reliability of a collection of items or variables using the Cronbach’s Alpha coefficient. Examining the Cronbach’s Alpha values for each component is crucial, in addition to looking at the overall Cronbach’s Alpha value. We found that all the results of Cronbach’s Alpha coefficients were close to 1. As the reliability was high, we did not discard any item. We display the results of the factor analysis and reliability analysis in
Table 5.
The first component is linked to wine and the wine tourism experience. It includes factors such as wine quality, availability of wineries, opportunity to taste lots of wine, interesting wine styles, and tasting experiences. It is not surprising that a wine tourism destination image is strongly defined by wine and wine-related characteristics. The items included in the first component are mostly functional.
The second component explains the atmosphere and environment of the WTD. It includes factors such as sense of freedom, discovery, escapism, and happiness, as well as a pleasant, hospitable, and easily accessible environment. It seems like the affective characteristics of the destination are an important part of its image. The second component items can be interpreted as the psychological dimension of the WTDI.
The third component includes factors related to cleanliness, nice climate, price levels, and level of safety. As for any other type of destination, safety, cleanliness, and other social factors are crucial. The third component items are mostly functional.
The fourth components are all about tourism facilities, i.e., shopping facilities, nice beaches, availability of tourist information, crowdedness, urbanization levels, and quality of service. We can assign some of the items to the psychological dimension of the WTDI, for example, good quality of service. The others can be considered as functional dimensions, such as interesting fairs, exhibits, and festivals.
The fifth component is linked to cultural and natural attractions, such as rich wine culture, nice scenery, vineyard landscapes, winery settings, opportunity for adventure, and increasing knowledge. This factor is mostly connected to the functional destination image component.
The sixth factor explains the comfort and infrastructure in the WTD. For example, variables such as quality of accommodation and restaurants, interesting local products, gastronomy, nightlife, and entertainment seem to be an important part of the WTDI. We included all the attributes that are part of the six components in
Table 6 so that they are easily accessible for future research purposes. Some of the items can be categorized as the functional WTDI dimension, while others are psychological.