Exploring Winery Visitors in the Emerging Wine Regions of the North Central United States

: Most of the pioneering studies related to wineries were conducted in famous wine regions, the results of which were not especially applicable to the small and emerging wineries in the United States. These wineries rely heavily on visitors’ on-site purchases; this highlights the importance of understanding winery visitors’ behaviors. This study explored the sequential relationships among winery visitors’ pre-trip characteristics (motivation, wine involvement), on-site experiences (perceived value), and post-trip evaluations (satisfaction, place attachment, winery loyalty). Intercepted winery visitors provided their e-mail addresses and were sent an online survey. Partial-least-squares structural equation modeling results showed that leisure-motivated and low wine-involved visitors perceive more diverse values in their winery experiences. Visitors’ perceived services, social beneﬁts, and wine-quality values contributed to the positive post-trip evaluations. Winery owners stand to beneﬁt from this research by being able to design winery experiences that are highly valued. Limitations are discussed and future research is suggested.


Introduction
The number of wineries across the United States has increased significantly in recent years. As of 2019, there are 10,185 wineries in the US, an almost 300% increase since 2000 [1,2]. The majority of these new wineries are located in regions not previously known for wine production, tend to produce small amounts of wine, and sell most of their wine directly to consumers from the tasting room [1]. Although wineries in emerging areas are mostly small producers, they have become vital to the health of rural communities that have seen a decrease in economic activity from family farming, manufacturing, and/or extraction [2][3][4]. Although winery visits are often the primary attraction, visitors combine wine tours with restaurant meals, shop purchases, gallery and museum visits, event attendance, and overnight stays [5,6].
Most of the wine produced throughout the world is made with (European indigenous V. vinifera) grapes, such as Chardonnay, Merlot, Riesling, and Cabernet Sauvignon. Wine production has not historically been feasible in cold climates because V. vinifera vines often do not survive harsh winters, and their grapes require longer growing seasons than others in northern climates. In recent years, however, grape breeding programs have developed dozens of new wine grape cultivars by breeding V. vinifera wine grape cultivars with nonwine grapes native to cold regions. These grapes are well-suited to cold climates and can be used to produce quality wines. Acceptance of these new hybrid grapes by growers, winemakers, and consumers has sparked the emergence of a vibrant wine industry in northern states of the US [7].
Generating visitors through wine tourism is critical for small wineries in emerging areas due to their high per bottle production costs and limited access to markets. In Wisconsin and Minnesota, 75% of produced wines are sold directly from tasting rooms, allowing for greater capture of margins [8]. Interestingly, the amount of wine sold in emerging regions appears to have more to do with visitor satisfaction with the tasting room experience than with the wine itself. In a study of visitors to small wineries in New York state, Shapiro and Gómez [9] found that satisfaction with the tasting room experience was the main driver of wine sales, and that factors such as customer service and atmosphere were more important to experience satisfaction than wine quality.
Similar to all businesses, wineries in emerging regions require information about consumers. However, most of the available market research on wine consumers is not appropriate for wineries in emerging regions. In the US, approximately 78% of wine is purchased off-premise (retail). Although some retailers occasionally offer samples of wines, most off-premise wine purchases are influenced by traditional retail methods (e.g., labels, shelf position, etc.) [10]. The concept of a vinothek, where wine is sampled off-premise before being sold is uncommon in the US. Approximately 20% of wine in the US is purchased on-premise (restaurants and bars), and the remaining 2% is sold directly to the consumer. In emerging wine regions, however, where the wineries tend to be small producers, direct to consumer distribution through tasting room sales is the most common method of distribution [3]. Significant market research exists on wine consumers, however, that research is based on consumer behavior in retail and restaurant contexts. Very little research exists on US wine tourism consumers, especially in emerging regions. This information is important to the success of small wineries because unlike retail wine consumers, most winery visitors are seeking wine-based experiences rather than simply bottles of wine. Moreover, there is evidence that visitors to winery tasting rooms in emerging regions have different characteristics than consumers, who purchase wine from retail and at restaurants [3,6]. In other words, the existing market research about wine consumers may not adequately address the needs of small wineries in emerging areas whose consumers are different, and behave differently, than retail and restaurant wine consumers.
One approach to understanding winery visitors is to investigate pre-, during-, and post-travel behaviors. Understanding the pre-trip stage of tourism behavior can indicate subsequent behaviors during and after the trip. Travel motivation is one of the most well-defined factors in the pre-trip stage. It is a contextual factor in which the construct is determined by the type of tourism destination [11,12]. In the wine tourism context, the most common visitor motivations are purchasing wine, tasting wine, viewing vineyards, and visiting wine trails [13]. Since people may have different reasons for visiting emerging wine regions, there is a need to understand these different tourists' motivations [14]. Another possible pre-travel factor is tourists' level of involvement with wine, which is known as a key factor to explaining their wine purchasing and consumption behaviors [15]. Since a winery is a distribution point for wines, evaluating visitors' levels of wine involvement can help predict their behavior.
In addition to the pre-trip approach, investigating on-site tourism experiences has become more important as the tourism suppliers increasingly focus on designing experiences [16]. To measure on-site experiences, researchers have used tourists' perceived values [17][18][19]. According to Vargo and Lusch [20], in a tourism context, perceived value indicates tourists' cocreation of values through their consumption of services provided by tourism suppliers. Wine tourism provides a variety of services, such as wine tasting, learning about winemaking, food pairings, and vineyard tours [21]. Too small to medium-sized wineries, however, running tourism business will be challenging due to a lack of information about winery visitors' preferences of tourism activities [22]. It is, therefore, important to understand how winery visitors value each of these services because they also directly and indirectly impact their post-trip attitudes and behaviors, such as satisfaction, attachment to wine regions, destination loyalty, and wine purchases [9,[23][24][25][26].
In summary, this study's objective is to investigate current winery visitors' behaviors through their pre-trip stage (motivation, level of wine involvement), during-trip state (perceived value of winery attributes), and post-visit evaluation (place attachment, satisfaction, destination loyalty). This holistic approach will help identify key winery attributes that influence post-visit evaluations and contribute toward experience economy research in the context of winery tourism. Moreover, this study will inform the experiences winery leaders design for visitors, resulting in repeat visitation and improved reputation of the wine brand.

Perceived Value
According to the theory of consumption values, consumer choice is a function of multiple consumption values, including functional, social, emotional, epistemic, and conditional values [27]. Each value is defined as the perceived utility acquired from an alternative's (1) ability to perform certain physical functions, (2) connection with specific social groups, (3) function that arouses feelings or affective states, (4) capacity to intrigue one's curiosity by providing novelty and knowledge, and (5) capacity that fits in certain situations and circumstances, respectively [27]. The theory also proposes that each of these values make differential contributions in any given situation and the consumption values are independent [28].
Sweeney and Soutar [29] developed a sophisticated scale that measures a broader range of customers' perceived values called the PERVAL scale that is based on the theory of consumption values by Sheth et al. [27]. Numerous researchers have applied the theory of consumption values and the PERVAL scale in tourism studies [21,[30][31][32]. Since many tourism products are not tangible but experiential, the PERVAL scale is applicable, especially when tourists experience a destination by consuming tourism products and receiving services [32]. Perceived value can explain what tourists' value about particular attributes they experience at a tourist destination through on-site evaluations rather than post-visit evaluations, such as satisfaction [20,26].
In a winery tourism study, Carlsen and Boksberger [23] identified key wine tourism attributes based on meta-analysis. The attributes were classified based primarily on Roberts and Sparks's [33] enhancement factors, which include authenticity, value for money, service quality, setting, indulgence, convenience, entertainment, and lifestyle. The results of the meta-analysis indicated that the most important attributes in wine tourism settings are service quality and winery setting. The least important attributes are value for money and authenticity. Carlsen and Boksberger [23] also emphasized the importance of understanding emotional and epistemic values for future research.
Relatively few studies have empirically utilized perceived value in a winery tourism context. Gill, Byslma and Ouschan [21] adopted diverse perceived value items from Sheth, Newman and Gross [27], Sweeney and Soutar [29], Petrick [32], and Zeithaml [34] and identified five factors including service quality, quality of wine, price, and social and epistemic values: These were all positively related with winery visitors' overall satisfaction. Chen et al. [35] included only three types of perceived value-value for money, hedonic value, and utilitarian value (e.g., how well the cellar door meets the visitors' needs), which were positively related to destination loyalty.
Since each winery tourist visits the winery with a different motivation and level of wine involvement, they will perceive different types of values. Accordingly, in the case of emerging wineries, it would be critical to investigate winery visitors' values toward winery attributes.

Wine Involvement
Involvement is defined as "the perceived importance or relevance of a person to an object/stimulus, which is based on the person's personal needs, values, and interests" [36]. Consumers' level of involvement encompasses their purchases, perceived values, and visit intentions, as well as what types of products they prefer [30,37]. Researchers, therefore, have used different types of involvement scales depending on their field of study [15,30,37,38]. For instance, Laurent and Kapferer's [38] consumer involvement profile (CIP) consists of four distinct facets: Perceived importance of product and consequences of a mispurchase (i.e., wrong choice of product), subjective probability of a mispurchase, hedonic value of the product class, and perceived sign value of the product class. Beatty et al. [39] identified two involvement items, one is enduring involvement-which is more related to self, hedonic pleasure and personal importance-and the other is purchase involvement, which is defined as "the outcome of an individual's interaction with the product and the purchase situation" [39].
In wine market research, Bruwer, Burrows, Chaumont, Li and Saliba [15] found that a wine consumer's consumption level, information search behavior, and other behaviors varied with their level of wine involvement. Hollebeek et al. [40] also found that consumers put more importance on the origin of wine and less on its price when they are high in product and purchase involvement. Lockshin, Spawton and Macintosh [37] segmented wine shoppers based on three involvement scales: Product, brand, and purchase involvement. Their study also revealed different types and levels of purchasing behaviors and demographic variables. Brown et al. [41] utilized CIP, though they focused solely on the facet of enduring involvement. The authors segmented winery visitors' behavior based on three extracted factors of involvement: Expertise, enjoyment, and symbolic centrality. This is known as the wine involvement scale (WIS). Demographically, the three identified groups were the same, though they were significantly different regarding their wine tourism behaviors.
In summary, the level of wine involvement is an important factor not only to predict and profile general consumers in retail settings but also to understand both wine shoppers and tourist behaviors together. This is particularly important since small wineries generally rely on revenue from wine sales to their visitors. Investigating wine involvement will help wineries satisfy both tourist and wine shoppers' needs.

Motivation
Motivation is defined as "psychological/biological needs and wants that arouse, direct, and integrate a person's behavior and activity" [42]. Traditionally, wine-related motivation is investigated in relation to wine consumption behaviors [15,37,43,44]. However, with the increasing popularity of winery tourism, numerous studies have measured winery tourists' motivations to visit wineries [11,[45][46][47]. Some studies have reported that the primary reasons for visiting wineries are to buy wine and to have a tasting experience [12]; others found motivations that are related to leisure and tourism, such as socializing, relaxation, and enjoying the views and scenery of wine regions [46]. Grybovych et al.'s [11] study of Indiana wineries also identified three wine tourism motivation factors: Learning about wine and the winemaking process (hereinafter termed epistemic motivation), having an authentic experience, and having a good time.
In general, the motivations of winery visitors differ widely from those of wine consumers. The more famous the wine, the more likely winery visitors are to visit to taste and buy wine [14,48]. Conversely, winery visitors venture to the smaller, newer, and less famous wines and wineries for leisure and tourism purposes [47,49].
While there are numerous factors that motivation can explain, one possible approach to understanding the needs of tourists with various motivations is perceived value. Prebensen, Woo and Uysal [30] stated that "value creation is driven by needs or motives and governed by cognitive processes and dispositions" [30]. Accordingly, Prebensen, Woo, Chen and Uysal [31] and Prebensen, Woo and Uysal [30] conducted research on whether motivation impacts perceived value-in both studies, a positive and linear causal relationship was found between motivation and perceived value of trip experience, which ultimately leads to satisfaction and loyalty. Since the motivation factor is itself a multidimensional scale, indicating both tourists' internal motivations and the external attributes of the tourism destination, identifying its relationship with different types of perceived value can give a better understanding of what kinds of motivated tourists put more value on specific types of destination attributes.

Satisfaction and Destination Loyalty
One of the most widely used post-trip variables in consumer and tourist research is satisfaction. An increasing number of tourism studies have sophisticatedly identified the relationship between perceived value and satisfaction [50,51]. The primary difference between perceived value and satisfaction is that perceived value can happen during various stages of the purchase process (before, during, and after), while satisfaction is considered part of the post-purchase stage [29]. Perceived values have thus been used as a different concept than satisfaction, usually as an antecedent of satisfaction [30,31,52]. Generally, the relationship of these two factors allows for the identification of destination attributes that have the strongest relationship with satisfaction. That is, if perceived performance of certain winery attributes is better than one's expectation or perceived importance of the same attributes, visitors are going to put values on the attributes which lead to satisfaction [53].
According to the service-profit chain model [54], an industry's revenue growth and profitability are determined by customer loyalty, which is mainly achieved by ensuring customer satisfaction. Several winery studies have demonstrated the positive relationship between satisfaction and the profitability of the winery [55], preference for a particular wine brand [56], returning visitors [57], and sales performance and retail retention [9].
In summary, it can be seen that perceived value, satisfaction, and destination loyalty have a well-defined relationship worthy of application to current wine tourism research. In particular, since wineries are generally equipped with diverse tourism attributes, it is important to identify the perceived values that make the most significant contributions to destination loyalty via winery visitors' satisfaction.

Place Attachment
Shumaker and Taylor [58] defined place attachment as "a system of interlocked attitudes and behaviors that refer to the home and the household and reflect the intimacy of strength of the individual's tie to the locale" (p. 220). Due to its emotional effects, place attachment has been used as both an antecedent and outcome variable of perceived values, satisfaction, and destination loyalty [59][60][61][62][63].
In Su et al.'s [61] study, the perceived service quality of Taiwanese hot-spring tourism was the main factor that explained satisfaction and predicted place attachment and destination loyalty. Tsai [62] also found a positive relationship between the perceived quality of a seaside holiday resort in southern Taiwan and tourists' satisfaction, which was used as a mediator between emotional attachment and destination loyalty. In addition, Xu and Zhang's [63] study found that tourists' involvement in visiting cities positively predicted attachment to the city of Hanzhou, China, and that led to destination loyalty. Lastly, this relationship was also found in Brown et al.'s [59] study, in which involvement in sport and attachment to the Olympic venue in London were positively related and led to visitation intentions.
In summary, it can be seen that, in addition to satisfaction, place attachment could be a good indicator of tourists' destination loyalty. Given that previous studies have shown mixed results for the relationships between the aforementioned variables, further research is required to define the relationship, particularly in the winery tourist context.

Hypotheses
Based on the literature, hypotheses were developed as listed below: Hypotheses 1 (H1). Leisure-based motivation is more positively related to perceived value dimensions than is epistemic motivation.

Hypotheses 2 (H2).
Low wine involvement is more positively related to perceived value dimensions than is high wine involvement.

Hypotheses 4 (H4).
Perceived values are positively related to place attachment.

Study Background
Wisconsin and Minnesota have developed a new and rapidly growing industry of small vineyards and wine enterprises. These wineries rely on new cultivars of cold-hardy grapes that can withstand cold weather and ripen in a shorter growing season. There are approximately 118 wineries in Wisconsin and 49 in Minnesota (see Figure 1). The wine produced in Wisconsin and Minnesota and that of rapidly expanding wine regions has less exposure than the wine of more famous regions (e.g., California) where a primary reason to visit is the reputation of the wine itself. However, the increasing number of visitors to these emerging wine regions (estimated to be 1.78 million in 2015; 8) demonstrates the potential of becoming a new type of winery tourism destination with types of visitors that are different from those who visit the more famous regions [1].
Sustainability 2020, 12, x FOR PEER REVIEW 6 of 20 Wisconsin and Minnesota have developed a new and rapidly growing industry of small vineyards and wine enterprises. These wineries rely on new cultivars of cold-hardy grapes that can withstand cold weather and ripen in a shorter growing season. There are approximately 118 wineries in Wisconsin and 49 in Minnesota (see Figure 1). The wine produced in Wisconsin and Minnesota and that of rapidly expanding wine regions has less exposure than the wine of more famous regions (e.g., California) where a primary reason to visit is the reputation of the wine itself. However, the increasing number of visitors to these emerging wine regions (estimated to be 1.78 million in 2015; 8) demonstrates the potential of becoming a new type of winery tourism destination with types of visitors that are different from those who visit the more famous regions [1]. This study is part of a larger suite of studies that explore this growing industry with integrated research on viticulture, winemaking, marketing of new wine grapes, and winery tourism [7,[64][65][66]. As part of the project, available winery services in 60 selected wineries were explored and illustrated in this study to provide general background of emerging wineries in North Central regions. It found that the most common service provided was wine tasting, followed by winery tours and live music performances. Half of the wineries served Hors d'oeuvres/food sampling services and two-thirds of the wineries offered a vineyard tour and barrel sampling. Some of the wineries provided a winemaker dinner, art show, and picnic/BBQ. A few of the wineries provided a winemaking/wine related class, grape stomp, and cooking class. The least common attribute was an animal viewing/petting. Other services included art classes, U-pick fruits, festival/concerts, and customer events (see Table 1).  This study is part of a larger suite of studies that explore this growing industry with integrated research on viticulture, winemaking, marketing of new wine grapes, and winery tourism [7,[64][65][66]. As part of the project, available winery services in 60 selected wineries were explored and illustrated in this study to provide general background of emerging wineries in North Central regions. It found that the most common service provided was wine tasting, followed by winery tours and live music performances. Half of the wineries served Hors d'oeuvres/food sampling services and two-thirds of the wineries offered a vineyard tour and barrel sampling. Some of the wineries provided a winemaker dinner, art show, and picnic/BBQ. A few of the wineries provided a winemaking/wine related class, grape stomp, and cooking class. The least common attribute was an animal viewing/petting. Other services included art classes, U-pick fruits, festival/concerts, and customer events (see Table 1).

Data Collection
For participants, the study sought out wineries in Wisconsin and Minnesota and access to their visitors. Seventeen Wisconsin and four Minnesota wineries cooperated with the current research. For data collection, visitors to winery tasting rooms were intercepted at wineries. Visitors were asked to provide their e-mail. A post-visit online survey was conducted with visitors to winery tasting rooms. A week after the e-mail address was collected, online surveys were sent out with links to the survey. The data collection was conducted between late August and October of 2015.  [11] winery tourism motivation scale was modified based on the literature and discussions with wineries. Place attachment was measured based on Williams and Jerry's [67] place attachment scale. The factors of satisfaction were friendliness of staff, knowledge of staff, tasting room facility, quality of wine, and overall experience. Loyalty was measured by intention to recommend the winery to others, to revisit the winery, and to repurchase the winery's wine. All variables were measured on a five-point Likert-type scale. Lastly, sociodemographic characteristics, such as education, age, gender, and income, were included.

Data Analysis
The current study used partial least squares structural equation modeling (PLS-SEM) to explore the causal relationships among variables. PLS-SEM is useful when the primary focus of a study is predicting and exploring rather than confirming theory [68]. Since covariance-based SEM focuses on confirming theories, its main technique for estimating model parameters is to minimize theoretical covariance and the estimated covariance matrix. PLS-SEM, on the other hand, is similar to multiple regression in that it needs to maximize the explained variance of the dependent variables (endogenous variables in PLS-SEM) while allowing for the assessment of data quality based on measurement model characteristics [68]. In addition, the bootstrapping process, which aimed to obtain standard errors through repeated random sampling for hypothesis testing, allows for the use of smaller sample sizes and non-normal data [68]. The skewness and kurtosis values of the study data showed that not all the variables were normal; therefore, it is more appropriate to use PLS-SEM than covariance-based SEM. Moreover, PLS-SEM was used because the primary focus of the current study is to understand the predictive relationship among variables and develop the theoretical structure needed for further exploration. PLS-SEM was conducted using the Smart PLS statistic program. Furthermore, for the frequency analysis of demographic, SPSS 22.0 was used.

Characteristics of Respondents
Out of 866 Wisconsin and Minnesota winery visitors sampled, 340 surveys (raw response rate 39.3%, effective response rate 45.3%) were collected. After cleaning the data, 319 surveys were used for the analysis.
Of the winery visitors who responded, almost two-thirds were female and 50% of them were over 50 years old. Approximately 50% of respondents were from Wisconsin, 17.9% were from Minnesota, and 12.5% were from Illinois. The majority were Caucasian (91.8%). Approximately 73% of respondents were married. Almost 80% of respondents were full-time, part-time, or self-employed. Around 80% of the respondents had college degrees. Nearly half of the respondents had an annual household income of more than $70,000.

Partial Least Squares Structural Equation Modeling (PLS-SEM)
Regarding motivation, the results identified two dimensions: Leisure and epistemic motivation (motivation to learn about wine and the winemaking process). For perceived values, seven dimensions were identified as expected. Wine-involvement emerged with two dimensions: High and low wine involvement. Satisfaction, place attachment, and winery loyalty were unidimensional factors. The internal consistency of items (reliability) and convergent and discriminant validities were evaluated. Cronbach's alpha for the identified dimensions ranged from 0.81 to 0.96, exceeding the threshold of 0.70. Composite reliability scores were also over 0.70, establishing the reliability of the measures for all dimensions (see Appendix A) [69].
Convergent validity was evaluated. The average variance extracted (AVE) score across all dimensions exceeded 0.5, indicating that more than half of the variances were explained by indicators and thus demonstrating convergent validity [70]. In addition, all outer loadings were significantly loaded on corresponding factors and well above the threshold value of 0.70 (see Table 2).  Discriminant validity was confirmed by the Fornell-Larcker (F-L) criterion (Table 2) and heterotrait-monotrait (HTMT) ratio of correlations (Table 3). F-L allows one to determine whether a construct shares more variance with related indicators than with other constructs [70]. In Table 2, the correlation between the same constructs (diagonal scores) revealed a higher coefficient than the correlation with other constructs, thus satisfying the F-L criterion.  The HTMT ratio of correlations estimates the proportion of the mean of all indicator correlations across two different constructs to the mean of the average correlations of indicators for two corresponding constructs [70]. To satisfy discriminant validity, the HTMT statistic for the relationship between two constructs is required to be under 0.85 and the bootstrap confidence interval to not contain a value of 1 [70]. All correlation coefficients were under 0.85, thus satisfying discriminant validity (see Table 3). The lower (2.5%) and upper (97.5%) bounds of the 95% confidence interval across all pairs of correlations did not contain a value of 1, which also demonstrated discriminant validity.

Hypothesis Testing
Since the estimation of path coefficients is based on OLS regressions, the collinearity among predictor constructs was examined using the variance inflation factor (VIF). VIF is the reciprocal of the tolerance (TOL) value, which represents the amount of variance of one construct not explained by the other constructs in the model. VIF values indicated that all combinations of constructs were below the threshold of five, confirming that multicollinearity was not an issue [70].
Next, since PLS-SEM does not assume normal distribution, a nonparametric bootstrap procedure was conducted with the recommended 5000 bootstrap samples [70]. Based on 5000 subsamples and generated standard errors, each hypothesis was tested for whether path coefficients were, in fact, zero in the population.
First, it was hypothesized that perceived values would be impacted more positively by leisure-based motivation than by epistemic motivation. While epistemic motivation had a positive influence only on epistemic value (β = 0.33, p < 0.001) and social value (β = 0.22, p < 0.001), leisure motivation was positively related to all types of values. Emotional value (β = 0.54, p < 0.001) and service value (β = 0.42, p < 0.001) were specifically highly significant. Thus, the presence of more positive and diverse relationships among leisure-based motivation and seven different perceived values supports Hypothesis 1 (see Figure 2). The second hypothesis proposed that perceived values would be impacted more positively by low wine involvement than by high involvement. Highly involved wine drinkers were significantly and positively related only with social value ( = 0.22, p < 0.001). Meanwhile, emotional value ( = 0.23, p < 0.001), epistemic value ( = 0.21, p < 0.001), price value ( = .23, p < 0.001), and service value ( = 0.21, p < 0.001) were significantly and positively predicted by low involvement that confirm Hypothesis 2 (see Figure 2).
The third and fourth hypotheses proposed that the seven values would be positively related to satisfaction and place attachment. Among the seven values, only service value ( = 0.43, p < 0.001) positively influenced satisfaction. Place attachment was positively explained by social ( = 0.33, p < 0.001) and wine-quality values ( = 0.23, p < 0.001). That is, satisfaction and place attachment were partially predicted by three different value types, which indicates that Hypotheses 3 and 4 are partially supported (see Figure 2).
In Figure 2, the coefficients of determination (R 2 values) are illustrated inside each endogenous factor. The R 2 values indicate the combined effects of exogenous factors on an endogenous factor representing a measure of in-sample predictive power [70]. All exogenous factors (i.e., independent variables) did well in explaining the endogenous factors.

Discussion and Conclusions
As hypothesized, both leisured-based motivation and low-wine involvement were more positively related to perceived values than epistemic motivation and high-wine involvement. The second hypothesis proposed that perceived values would be impacted more positively by low wine involvement than by high involvement. Highly involved wine drinkers were significantly and positively related only with social value (β = 0.22, p < 0.001). Meanwhile, emotional value (β = 0.23, p < 0.001), epistemic value (β = 0.21, p < 0.001), price value (β = 0.23, p < 0.001), and service value (β = 0.21, p < 0.001) were significantly and positively predicted by low involvement that confirm Hypothesis 2 (see Figure 2).
The third and fourth hypotheses proposed that the seven values would be positively related to satisfaction and place attachment. Among the seven values, only service value (β = 0.43, p < 0.001) positively influenced satisfaction. Place attachment was positively explained by social (β = 0.33, p < 0.001) and wine-quality values (β = 0.23, p < 0.001). That is, satisfaction and place attachment were partially predicted by three different value types, which indicates that Hypotheses 3 and 4 are partially supported (see Figure 2).
In Figure 2, the coefficients of determination (R 2 values) are illustrated inside each endogenous factor. The R 2 values indicate the combined effects of exogenous factors on an endogenous factor representing a measure of in-sample predictive power [70]. All exogenous factors (i.e., independent variables) did well in explaining the endogenous factors.

Discussion and Conclusions
As hypothesized, both leisured-based motivation and low-wine involvement were more positively related to perceived values than epistemic motivation and high-wine involvement. Perceived values that contributed to the satisfaction and place attachment were service value, and wine quality and social values, respectively. Ultimately, these satisfied and attached winery visitors were more likely to become loyal customers.
First, the current study investigated whether winery visitors could perceive values differently depending on their type of motivation and the level of wine involvement. According to expectancy-disconfirmation theory, people evaluate products positively if the perceived quality exceeds their expectations and negatively if the perceived quality fails to meet expectations [71]. That is, people evaluate products differently depending on their personal needs, expectations, and previous experiences [72]. For instance, Dabestani et al. [73] found that customers who had higher expectations of hotel services before arrival were more likely to have lower satisfaction levels than customers with neutral expectations. Since perceived value originates in cognitive evaluations, it is more likely to be influenced by these pre-visit stage of winery visitors' characteristics [29].
In the case of motivation, positive relationships between leisure motivation with all other values could imply that visitors have relatively lower expectations toward winery attributes and be less strict about evaluating its features. This could lead them to be relatively generous about its features while focused more on their feelings of travel or social activities with their friends and families. The strongest relationship with the emotional value further supports the point that their winery visit was part of their leisure activities that originate in hedonic experiences [74]. Similar to this line of reasoning, those who were epistemically motivated might have had higher expectations toward wineries, especially related to social and epistemic features. According to Wolf et al. [75], those who have knowledge about wine are more eager to learn about wine, not only so they can choose better wines but also to improve their social status. Currently, wines are widely available for all types of consumers, yet they are still considered as luxury products [76]. Hence, wine consumers' confidence in their wine knowledge, and eagerness to learn more advanced information, can be a part of a luxury experience that can enhance their self-identify in a more desirable way. The relationships, therefore, could be due to their underlying motivation to achieve a luxury experience accrued from enhanced wine knowledge.
Regarding wine involvement, it was found that being highly involved in wine was only positively related with the social value. The relationship might be explained by Stebbins' serious and casual leisure perspective: Serious leisure indicates an amateur or hobbyist who considers an activity as highly substantial while causal leisure indicates a pleasurable activity that requires little or no training [77]. Ravenscroft and van Westering [78] asserted that one of the main motivations of serious leisure in the winery tourism context is to be part of a membership that signifies one's prestige. Accordingly, their special skills, knowledge, and experiences are not just for finding the best wines or to have hedonic experiences, rather they enable social communication with like-minded people or the expression of their social status. Those winery visitors who belong to the serious leisure group are less likely to be concerned about the price, a multi-sensory wine experience, or the wine brand [40,78]. Instead, they might prefer wines that are only recognizable to a committed group [79]. Researchers have found that this phenomenon is more dominant if the product is perceived as a luxury item [80][81][82][83]. Accordingly, those highly involved in wine may have only a perceived social value because they highly value their social image, which could be enhanced by visiting and consuming wines at wineries. By contrast, low involved winery visitors, as a group that pursue casual leisure, might be concerned more about their instant hedonic experiences as tourists. Ravenscroft and van Westering [78] differentiated winery tourists as "others" from the socially distinct group (serious leisure group) and considered themselves as those who come to experience a part of the wine world. That is, those low involved visitors might be closer to conventional tourists, while highly involved winery visitors might be considered as individuals who enjoy everyday leisure activities. Low involved visitors might purchase wines as souvenirs and gifts or for later consumption, but since that is not a part of their daily purchase routine, these visitors might care more about the price of wines. They are also likely to perceive winery services as having high value, since winery employees play an important role in choosing appropriate wines and creating a positive memorable winery experience [84,85]. This is particularly applicable to those who need more knowledge about wines. Lastly, positive relationships with emotional value indicates that these winery experiences are grounded in the hedonic nature of the tourism experience.
Perceived value factors that contributed to place attachment, and ultimately to destination loyalty, were social and wine quality values. While Carlsen and Boksberger [23] considered wine quality as less pronounced in a tourism context-instead being associated with wineries that only sell wines-they found that maintaining the good quality of the wine is still critical to determining winery visitors' emotional attachments. This result could be due to the fact that wineries in Wisconsin and Minnesota are still emerging and are therefore characterized by the original functions of wineries-serving and selling wines [86]. Moreover, the positive relationship with perceived social value implies that wineries in Wisconsin and Minnesota were well enough equipped with attributes to make visitors feel prestigious, leading to place attachment. Pratt and Sparks [82] also found that visitors whose self-identity matched well with a winery's image were more likely to have positive attitudes toward winery tourism. It seems that understanding visitors' social values is important when the product represents prestige and social status [27,87].
Lastly, consistent with the Carlsen and Boksberger's [23] meta-analysis, one of key wine tourism attributes that can satisfy winery visitors was service, which ultimately led to winery loyalty. In a winery setting, service values that reflect winery staff's expertise in wines, as well as good hospitality, is a key factor contributing to wine sales and satisfaction [21,61,88,89]. The role of the staff would be particularly important since those leisure-based visitors, who are not so familiar with wines, would more likely rely on winery staff services.
The current study investigates the sequential relationships of pre-trip characteristics (wine involvement and motivation), on-site evaluation (perceived value), and post-trip evaluation (satisfaction, place attachment, and destination loyalty). To the best of the author's knowledge, in winery tourism context, there is no study that has investigated the relationship between on-site cognitive evaluations and post-trip evaluations. Furthermore, while it has been known that perceived value is a multidimensional factor, most tourism studies have measured perceived value as unidimensional [30,72,90]. These studies generally investigated the extent of perceived value and its impacts, which makes it harder to identify which types of perceived value have the strongest impacts on tourists' behaviors. By exploring perceived value toward individual winery attributes, this study can find the factors that are most effective and thus identify areas that wineries need to improve.
Results also suggest that leisure-motivated and low wine-involved visitors appear to be the most valuable customers since they perceived values more extensively. For those visitors, wineries need to be prepared to have well-trained winery personnel, quality wine, and visually appealing winery decor, facilities, and landscape (e.g., vineyard, pastoral setting, etc.). They should also be able to provide appropriate information so that visitors can be entertained and get help tasting and purchasing wines. In addition, the winery personnel also have to have effective marketing skills that can enhance the value of wine quality. Several studies have demonstrated that to those who are not wine experts, the quality of wine can be determined by subjective factors, such as bottle appearance, winery service, consumption location, etc. [91][92][93]. While objective wine quality is the most important attribute, winery visitors' satisfaction might not just stem from the taste itself, but from the tasting experience, well-decorated winery facilities, an authentic atmosphere, and a beautiful surrounding landscape [92].
Even though the current study contributes to research on wine tourism, it is worth noting several limitations which further research should address. First, these results are applicable only to Wisconsin and Minnesota wineries. To be able to generalize the current study, further research is needed in other emerging wine regions. Additionally, most respondents were intercepted by winery personnel and asked to provide their e-mail information. Thus, there is a possibility that the sample does not fully reflect the general population of winery visitors. Future research might be improved by promoting winery visitors' volunteer participation or targeting general population.
In conclusion, evaluating visitors' on-site experiences based on perceived value helps in identifying the key winery attributes that influence post-visit evaluations. Moreover, depending on their level of wine involvement and type of motivation, winery visitors perceived different values of winery attributes. Better-informed plans and strategies may be designed by wineries with desired tourism experiences in mind. This should lead to increased wine sales, repeat visits, and an improved reputation of their wine brand.

Conflicts of Interest:
The authors declare no conflict of interest.