Wine choices are not always fully understood by academic researchers or the industry. This paper aims to outline and test a theoretical model proposing that wine consumption may be dependent on differences in consumer expertise, the hospitality situation, characteristics of the wine itself and an interaction of these variables.
Three empirical studies (total sample size = 356) tested these theoretical propositions. Consumers with varying levels of wine knowledge were presented with experimental vignettes showing videos of wine opening and pouring and were asked to pair wines with hospitality situations.
Study 1 found that consumers with low product knowledge were more sensitive to hospitality situations and extrinsic product attributes (closures) than were the experts. Study 2 found that wine hospitality situations fall into three predicted categories, namely, food, friends and formality, although contrary to prediction, the presence of food was the weakest predictors. Study 3 demonstrated the robustness of the three-dimensional structure of wine hospitality situations.
These studies provided important practical information because targeting various market segments requires the industry to know what product attributes are favored by different groups of consumers different situations.
Previous researchers have discussed the difficulty of measuring consumption situations. By limiting these studies to wine consumption within hospitality situations, the authors learned much about how consumers’ characteristics, product attributes and the situations interact to influence not only product assessments but also choices.
Duhan, D.F., Rinaldo, S.B., Velikova, N., Dodd, T. and Trela, B. (2019), "Hospitality situations, consumer expertise, and perceptions of wine attributes: three empirical studies", International Journal of Wine Business Research, Vol. 31 No. 1, pp. 68-88. https://doi.org/10.1108/IJWBR-07-2018-0035
Emerald Publishing Limited
Copyright © 2019, Emerald Publishing Limited
There are many traditions regarding the consumption of wine. These traditions often have powerful influences on the perceived appropriateness of a particular type of wine in a particular situation. As an example, for a beef entrée, serving red wine is usually considered appropriate, whereas white wine is often a suitable match for fish and seafood dishes. In addition to food and wine pairing, certain consumption situations may call for a specific type of wine. Champagne, for example, tends to be seen as a drink of celebration and is often served symbolically to emphasize accomplishments. On the contrary, consumers may perceive certain wine types unsuitable for certain situations. For instance, Charters et al. (2011) found that respondents in their study unanimously thought that serving champagne or sparkling wine at a funeral would be very inappropriate.
The importance of such traditions varies among consumers and situations. In this study, we selected some hospitality situations and investigated their influences on consumer assessment of wine. We proposed and tested three dimensions of wine-related consumption situations and explored differences in information used by experts and novices when making assessments of these situations.
The research provided further support for the people-product-situation framework in the wine consumption context. The model proposed in this study could be a useful tool for a variety of situational assessments of wine consumers. In addition, the study highlighted various dimensions of hospitality and the role it plays concerning wine. Although there have been numerous studies integrating hospitality and wine consumption, this study reinforces the link between the two.
Attributes: people, products and situations
There is a rich history of research concerning the interactions among people, products and situations. Research into decision-making has led to the development of the elaboration likelihood model (ELM) of consumer information processing. ELM focuses on consumers’ inclinations to make decisions by processing information through either “central” or “peripheral” processing “routes” (Petty and Brinol, 2011). It posits that consumers with higher levels of involvement in and knowledge of a product category are more likely to use a central (more elaborate) processing route to make judgments about products. Those who are less involved and/or less knowledgeable are likely to use a peripheral processing route. In a classic study of consumer information processing, Alba and Hutchinson (1987) found that experts can process more information, in more detail, and do so more quickly than novices. Also, experts tend to use information that is inherent (intrinsic) to products, such as features and functions, whereas novices tend to use peripheral (extrinsic) information, such as prices, packaging and brands. Their evaluation approaches differ as well – novices are more holistic, whereas experts are more analytic. For these reasons, we have chosen the level of expertise (experts versus novices) with the product category (wine) to serve as peoples’ attributes in our study.
The distinction between the types of information used by experts and novices led us to focus on extrinsic versus intrinsic product attributes. Extrinsic attributes of wine can strongly influence the evaluations of intrinsic sensory characteristics (Lange et al., 2000; Mueller et al., 2010). Therefore, under our framework, we suggest that intrinsic product attributes are sources of influence on the assessments of extrinsic attributes of wine.
The last component of the paradigm is the “situation” – a notion commonly known as a combination of circumstances that occur concerning one another. The term by itself is too broad to have a clear operational meaning, so researchers have typically used compound terms to narrow the domain of appropriate circumstances (e.g. purchase situations, disposal situations, consumption situations, etc.). We focused on wine consumption situations.
Wine and hospitality
Previously, frequently examined wine consumption situations included classifications by occasion (e.g. birthday party/celebration), place (e.g. at home/restaurant/bar), people (e.g. with friends/family/self), formality (e.g. casual/fine dining/picnic), ceremonial context (e.g. sacramental), time (e.g. lunch/dinner), etc. The domain of wine consumption situations is multifaceted. The innovative angle that we take in this research is an investigation of wine consumption situations through the overarching theme of hospitality. We believe that a link between wine and hospitality is vital to the understanding of wine markets.
The term “hospitality” is commonly associated with both the provision and consumption of food/beverage or accommodation in a commercial sense (Williams, 2002). Lashley and Morrison (2000, p. 3) note “current preoccupation with commercial provision”, in that hospitality is widely perceived, researched and taught as an economic activity (sets of consumers and suppliers). However, hospitality is a series of behaviors (primarily involving mutuality and exchange) which originate with the very foundations of society (Lashley and Morrison, 2000). The traditional (archaic) understanding of hospitality is rooted in the building of bonds among individuals and the turning of strangers into friends; thus, the act of hospitality becomes a bedrock upon which ancient cultures developed (Scott, 2006).
Implicit in this understanding is a dichotomy between hospitality as a private act (the socialization of the stranger) and commercial transaction (the economic value added to the provision of food/beverage or accommodation) (Scott, 2006). Regardless of the context, a central aspect of hospitality is that people, both as recipients of physical aspects of hospitality (e.g. food/beverages) and as players in the socialization processes. Hospitality encompasses “symbolic performances, marked by affective and expressive acts” (Bugge, 2005, p. 5). Therefore, hospitality includes ritual performances.
Wine consumption is part of the ritualized acts of hospitality. Just as hospitality is much more than the provision of food/beverage or accommodation, wine consumption encompasses much more than the mere intake of wine. It embraces all three domains of hospitality suggested by Lashley and Morrison (2000) – social, private and commercial. The social domain of hospitality considers social settings in which acts of hospitableness take place. Consider a group of friends sharing a bottle of wine over a casual get-together. In such a social situation, wine (and food) becomes a means of binding people together, thus helping create a collective identity of the group. The private domain of hospitality reflects the impact of host-guest relationships. When offering wine to guests at a party, the hosts provide acts of hospitality as a means of communicating welcome, goodwill, generosity and friendliness. Finally, the commercial domain of hospitality concerns economic activities in both the private and public sectors. Wine consumed at a dinner party at a restaurant exemplifies the commercial provision of hospitality (Bruwer et al., 2017).
To conclude, the fundamental notion of hospitality is sharing and provision with others. We take this idea as the key approach to our examination of wine consumption situations.
General theoretical proposition
Pulling together the core aspects of the people-product-situations paradigm, we propose an application of this framework to consumer assessment of wine in various hospitality situations. Specifically, we suggest that perceptions of hospitality situations directly influence consumer evaluations of extrinsic wine attributes. Both intrinsic product attributes and consumers’ expertise with wine moderate this relationship. Additionally, both intrinsic product attributes and consumer expertise have direct influences on evaluations of extrinsic product attributes. Figure 1 illustrates a general person-product-situation framework (on the left) and our theoretical proposition for this framework’s application to consumer evaluations of wine attributes (on the right).
Three studies are presented herewith to investigate this theoretical proposition. Study 1 tests specific research hypotheses derived from the theory. Study 2 focuses on three proposed dimensions of hospitality situations. Study 3 uses results from the first two studies to propose a structure of consumer information for assessing wine products in various situations.
Study 1: wine, hospitality situations and expertise
The objective of this study was to produce an empirical test of the people-product-situations framework in the context of wine markets.
The first of the three broad sources of influence in the framework is the consumption situation for the product. In the research literature, we found a variety of wine consumption situations, many being related to consumer characteristics, such as product involvement (Quester and Smart, 1998), values and consequences (Hall et al., 2001), social identity and susceptibility to normative influences (Orth, 2006) and demographics/consumption behavior (Thach, 2011). For our study, we have selected a broad range of situations to capture a variety of hospitality circumstances. The goal was to examine how influential these situations are on consumers’ assessment of specific wine attributes.
Extrinsic product attributes
For extrinsic product attributes, we decided to examine wine closures (corks versus screwcaps). Corks are a classic type of closure for wine packages that are traditionally associated with better quality and more expensive wine (Barber and Almanza, 2007; Marin and Durham, 2007; Marin et al., 2007; Wilson, 2009). Recently, however, screwcaps have become pervasive in some markets and are now making inroads, even where they were once considered negative indicators of wine quality (Halstead, 2011). Screwcap closures provide some significant practical advantages, including the avoidance of cork taint (Barber and Almanza, 2007). However, many consumers still resist non-traditional forms of packaging and continue to prefer the romance and drama of opening cork closures (Atkin et al., 2006).
Consumer perceptions of alternative closures receive attention in the wine marketing literature because those perceptions continue to evolve within and among markets. Nevertheless, to the best of our knowledge, no previous study has specifically examined consumers’ perceptions of the appropriateness of closures for various consumption situations. Therefore, we propose our basic research hypothesis:
Hospitality situations will influence the appropriateness of wine closures.
Intrinsic product attributes
Intrinsic wine attributes may include a multitude of features, such as grape varieties, levels of alcohol and residual sugar, tannins, blends and winemaking styles. However, including multiple attributes would have increased the level of complexity for respondents to the point of being counter-productive because it would likely cause information overload. Therefore, we limited intrinsic product attributes to the basic types of wine. We examine three most often consumed wine types (red, white and sparkling). The reason for including sparkling wine was its strong association with certain hospitality situations, such as celebrations (Charters et al., 2011). Thus, we propose the following research hypotheses:
The type of wine will moderate consumer assessment of wine closures in hospitality situations.
The type of wine will directly influence assessments of wine closures appropriateness.
For our study, the ELM underlies the logic of the relationship between consumer expertise and perceptions of appropriateness for wine closures. The significance of wine consumer expertise has been studied extensively (Aurier and Ngobo, 1999; Dodd et al., 2005; Mueller et al., 2008; Viot, 2012). Such studies have confirmed that expert and novice consumers differ in the amount, content and organization of their wine knowledge. This is important because targeting various market segments effectively requires the industry to know what product attributes are favored by different groups of consumers. As wine experts and novices are known to differ in their use of different product attributes, we expected that their attitudes regarding wine bottle closures might also differ for different hospitality situations. Thus, we propose the following research hypotheses (Figure 2):
Levels of consumer expertise will moderate the relationship between hospitality situations and the appropriateness of wine closures.
Levels of consumer expertise will directly influence the appropriateness of wine closures.
An experimental vignette study was designed to test the model. Vignettes are brief descriptions of scenarios that systematically include combinations of conditions are under investigation. The experimental vignette method (EVM) provides much of the ability to control some variables while manipulating others, which is necessary for the investigation of causality, thus strengthening internal validity. It also provides greater realism, which enhances the generalizability of results by strengthening external validity (Atzmüller and Steiner, 2010).
Our design used a full factorial set of conditions, including 12 hospitality situations, 3 types of wine and 2 levels of consumer expertise. Each respondent assessed 36 vignettes (a within-subjects or repeated measures design), and they were grouped into two levels of consumer expertise (between or among subjects measures). We collected additional data regarding consumer demographics, behavior, product category involvement and knowledge.
Measures and manipulations
Appropriateness: The dependent variable (DV) was the perceived appropriateness of extrinsic product attributes – package closures. Measurements for corks and screwcaps were collected separately by using semantic differential scales ranging from “very inappropriate” (1) to “very appropriate” (5) for each of the 36 combinations of hospitality situations and types of wine. The value for appropriateness in each combination is the difference between the score for corks and screwcaps. The resulting differences indicate, at one extreme, that corks are much more appropriate than caps (a cork score of 5 minus a screwcap score of 1 equals +4); and at the other extreme that screwcaps are more appropriate than corks (a cork score of 1 minus a screwcap score of 5 equals −4). The middle of the scale indicates that there is no difference in the levels of perceived appropriateness (e.g. the difference between a cork value of 5 and a screwcap value of 5 is 0, similarly 4 − 4 = 0).
Expertise groups: Alba and Hutchinson (1987) found that consumer knowledge comprises two parts, namely, expertise and product familiarity. Familiarity is the number of product-category-related experiences that a consumer has accumulated. It was found to positively influence expertise by improving:
consumers’ ability to process product-related information; and
the speed with which they can process it.
We argue that consumers’ levels of involvement with a product category positively influence the likelihood of having experiences within that category, which in turn leads to greater familiarity, knowledge, and therefore, expertise. For this study, five indicators were used to assess consumer expertise: wine involvement, purchase frequency, consumption frequency, years of experience with wine consumption and wine knowledge. The items for assessment of wine involvement were adapted from Slama and Tashchian (1987), and wine knowledge items were adapted from Smith and Park (1992).
Hospitality situations: Based on an extensive review of wine research literature, we selected 12 hospitality situations:
Going out with friends
Dinner party at home
Dinner party with friends
Casual everyday use
Attending a dinner party
Dinner for two
Casual get together
Restaurant with clients
Wine type: The most important intrinsic attribute of wine – its type – was used, with three different wine types tested – red, white and sparkling wine.
Data from both the high and low expertise groups were collected through an online presentation of the vignettes and questions. To enhance the realism of the vignettes, we used six short videos averaging 35 s each. The videos showed the opening of traditional (750 mL) wine bottles with corks and screwcaps. They were presented in pairs by types of wine (e.g. a cork video for the white wine paired with a screwcap video for white wine) that were embedded at appropriate points in the questionnaire to provide a context for specific questions and responses about each type of wine. The videos provided a visual and audio experience of the differences between opening each type of closure. They also focused the respondent’s attention on the appropriate type of wine and closure for each stage of the data collection. As a control, labels were removed from the bottles shown in the videos. After viewing each pair of videos, the respondents assessed the appropriateness of each type of closure for that particular type of wine in each of the 12 hospitality situations. Please see Figure 3 for the screenshots of each video.
Sample and data hygiene
The subjects were selected to represent consumers with different levels of expertise with wine. We sought the high expertise sample among winemakers, growers and wine industry suppliers who were contacted through a grower’s association conference and an enology network membership list. Consumers who self-identified as being interested and involved in wine were also included. They were contacted through their affiliation with a wine specialty retail shop and its wine club membership list. The non-expert (novice) sample came from students attending a large university in southwest USA. The students were chosen because they typically have less experience with wine and they have not self-selected as interested in wine. There were 135 usable responses from the presumed low expertise group and 82 from the presumed high expertise group, for a total of 217. The novice and expert groups were not significantly different regarding gender, but as expected, they differed in age and preferences for wine. No anomalies in the data and no out-of-range values (outliers) were found. Fewer than 10 per cent of the values were missing for any of the variables. Because of this low incidence, mean-substitution was applied to missing values.
Confirmation of expertise groups
The presumption that the two groups were different in their levels of expertise was tested by using five indicators: purchase frequency, consumption frequency, years of experience with wine, involvement with wine and knowledge about wine. The means of these indicators for each group were significantly different in the directions expected. The correlations among the variables were all positive and significant (a mean of 0.64 and a median of 0.67) providing some indication of convergent validity of these measures (converging on the concept of expertise). A logistic regression analysis was performed with these indicators, and the resulting model was 96.8 per cent accurate in classifying individuals into the two expertise groups.
Examination of appropriateness
The next stage of analysis focused on the perceived levels of closures appropriateness. The means within each category for each of the exogenous constructs (expertise, wine type and situation) and their significance levels are listed from highest to lowest within each construct in Table I, along with two different types of means comparisons. The first comparison is the significance of the difference for each category mean against a null (test) hypothesis of zero (indicating that there is no difference between the appropriateness of corks and screwcaps). Significant differences are indicated with an asterisk. All, but three means, are significantly different from zero and have positive signs. The exceptions are:
“casual get together,” which is not significantly different from zero; and
“casual everyday use” and “picnics,” which are significantly different.
The second comparison was among the appropriateness scores within each of the constructs (wine types, expertise levels and situations). For instance, the means for types of wines show that the appropriateness of corks is greater for red wines than for white wines and fall between for sparkling wines.
Tests of propositions
Next, we tested the research hypotheses. The data were composed of a DV with interval level data, two independent variables (IVs) with categorical data and one IV with ordinal data. We controlled for the effects of individual respondent tendencies by including respondent IDs in the analysis as a covariate. Table II presents the ANOVA results. The main effects and interaction terms were all significant at p < 0.001, except for the interaction of hospitality situation and wine type. These results provide empirical support for four of the five research hypotheses.
Given the relatively large numbers of degrees of freedom (total df = 7,812), the levels of statistical significance tell only part of the story in these data. Therefore, we also examined the effect sizes using the partial eta squared (ƞ2) values. Partial ƞ2 values are presented in Table II to provide additional insight into the role of each source of influence. Various standards have been recommended for the interpretation of partial ƞ2. We used Cohen’s (1988) generally accepted standards for interpretation with multiple IVs (Table III):
In Table II under Source Effects, we can see that Hospitality Situation and the interaction of Hospitality Situation with Expertise Group standout as the greatest sources of influence. We also see that Respondent ID, the control for individual respondent tendencies, was significant but it had a trivial influence.
Summary of Study 1
The significant interaction between Expertise and Hospitality Situations and the lack of significance for the interaction between Wine Type and Hospitality Situation are shown in Figure 4.
The novice group with a much wider range of appropriateness scores is more sensitive to situations than the expert group. This is consistent with the notion that novice consumers are more responsive to non-intrinsic attributes when making product assessments. Clearly, this framework based on the ELM provides a useful lens through which to examine wine consumption attitudes. The hospitality situations and their interactions with the expertise levels of consumers are the strongest sources of influence regarding the appropriateness of wine closures.
Study 2: dimensions of hospitality situations
Hospitality situations were the dominant source of influence found in Study 1. However, consumption situations are multidimensional, which can make it difficult to generalize findings from one research context to another. The focus of this study was on identifying generalizable dimensions of the situations that can be useful for understanding wine consumer attitudes. We re-examined wine research literature seeking potential theoretical dimensions of hospitality situations associated with wine consumption. We found three dimensions that warranted further examination. We then developed an additional survey instrument to gather assessments of each of the three dimensions in the 12 hospitality situations. The resulting assessments of dimensions were substituted for the categorical indications of situations in a re-analysis of the data from Study 1. This was a semi-exploratory study intended to produce more generalizable results.
The first of the three identified dimensions is food consumption, which has been used as a basis for differentiating among wine situations (Hall et al., 2001; Quester and Smart, 1998; Thach, 2011). We labeled this dimension simply Food. The second dimension is the level of social interaction in a situation (Orth, 2006; Quester and Smart, 1998). We labeled that dimension Friends. Study 1 revealed a possible third dimension, where the three highest scores for appropriateness were for the “restaurant with a client,” “gift giving” and “celebration-wedding” situations; and the three lowest scores reported were for the “picnics,” “casual everyday use” and “casual get-together.” These differences indicated the third dimension, which we labeled Formality. Thus, drawing from both existing literature sources and Study 1 results, we propose the following three dimensions for the wine-hospitality context situations:
Food: Hospitality situations often involve eating and perhaps preparing meals. In these circumstances, the intrinsic characteristics of the wine (e.g. sweetness, acidity, carbonation and color) can influence the taste and the appearance of a meal; thus, these characteristics are important to food-related situations. The pairing of food and wine flavors may strongly influence the perceived appropriateness of a wine. This dimension indicates whether food is influential to the hospitality situation.
Friends: Hospitality situations may be prominent social occasions involving family, friends and/or acquaintances. The appropriateness of wine products is often determined by the presumed preferences and attitudes of the people gathered. It also raises the potential social risk of serving wine that some may perceive as inferior. This dimension indicates many or fewer individuals present, giving the social dimension of the situation more or less influence on assessments of appropriateness.
Formality: In some hospitality situations, bottles of wine are presented to hosts or guests of honor and opened for the occasion. So, the extrinsic characteristics of the wine (e.g. packaging or label) are more noticeable, and thus, are more important. Risk aversion tendencies in consumers may tilt preferences toward a more traditional presentation of wine in more formal situations. This dimension indicates whether the hospitality situation is low or high in its level of formality.
Interactions among dimensions
Situational attributes can interact with intra-individual and product attributes. When this occurs, it is referred to as a “person-within-situation interaction” (Quester and Smart, 1998). For instance, the relevance of a product’s origin identity may be dependent on the product category, the consumer’s knowledge of that category and the availability of the origin information.
A new questionnaire was developed, in which respondents were asked to assess each of the three dimensions for each of the 12 hospitality situations, using semantic differential scales (Table IV).
Sample and assessments of dimensions
Data were collected from a new sample and yielded 139 responses. An analysis for anomalies found no outliers. Fewer than 10 per cent of the item responses were missing, so mean substitution was applied.
As the purpose of Study 2 was to enhance the interpretability and generalizability of the results of Study 1, we chose to use ordinal summary assessments (high/low) on each dimension. These were determined by first calculating the averages for each of the combinations across the respondents. A median split was then used to identify either “high” or “low” on each dimension in each situation. Figure 5 presents the high-low values for each of the 12 hospitality situations.
Re-analysis with food-friends-formality assessments for hospitality situations
The assessments of the three dimensions were substituted for the single variable “hospitality situation” in the data from Study 1. The results of the re-analysis are presented in Table V. We see that nearly all of the variables and combinations of variables are statistically significant, primarily because of a large number of degrees of freedom available for the significance tests. Hence, we again use partial ƞ2 to assess the source effects. There is a similar pattern in the two studies; some aspect of hospitality situations (formality) and interactions with it are the dominant sources of influence. The predominant dimension in the re-analyzed model is formality. Indeed, formality and the interactions with both expertise group and friends comprise all of the nontrivial sources of influence in this model. The only compromise in using this approach is the loss of approximately 1.5 per cent points of explained variance.
Summary of Study 2
The positive result from Study 2 is that we can interpret wine-hospitality situations in terms of their underlying dimensions that are more generalizable and more easily understood. On the other hand, when hospitality situations are formal, there is generally a greater concern for tradition and appropriateness of the accouterments of the occasion.
A surprise in these results found the complete uselessness of food in explaining any variance. Food would seem to be a logical influence on the appropriateness of corks or screwcaps, especially as wine is often opened at the table and served from the bottle throughout the meal. Perhaps, this lack of influence is because food is arguably a part of most of the hospitality situations, and hence, it does not differentiate among situations well. It may also be that too little information regarding the nature of the food was presented in the study. Hospitality situations in which food is central to the occasion are typically focused on the flavors, textures, temperatures and other intrinsic characteristics of what is being consumed. The skillful blending of these attributes is key to the success of many occasions. It follows that the intrinsic characteristics of the wine might dominate what is important in such situations and that closure types would be less relevant. Indeed, in some food-centered situations, wine is often served in carafes or by the glass, and the nature of the wine package closure is mostly unknown. At gatherings, where many people are present, a variety of wines are often served, especially if the occasion is not focused on food. As a result, many guests may be unaware of the nature of the package closures and/or can choose one or the other.
Study 3: information structures for assessment of the wine-hospitality context
The objective of Study 3 was to gather what we have learned in the first two studies and to investigate the nature of information used for assessments in the wine-hospitality context. Information processing literature indicates that consumers with different levels of expertise process information differently. Additionally, we found that interactions between expertise levels and other factors have significant effect sizes. As a result, we separated our analysis of these data by expertise group.
Using the data in hand from Studies 1 and 2, we used factor analysis as an exploration tool. The first step was to examine the data for their suitability for factor analysis. The Keyser–Meyer–Olkin (KMO) measure of sample adequacy for the novice group was 0.795 and for the expert group, it was 0.831. The Bartlett’s test of sphericity was highly significant in both cases (p < 0.001), indicating that the samples for each group were adequate for factor analysis. The next consideration was the types of extraction and rotation processes. Oblique rotations are often recommended for exploratory analysis of social science phenomena because dimensions of such phenomena are often not independent of one another; thus, more realistic insights are likely to result from non-orthogonal analyses. Therefore, we used principal axis factoring for the extraction process and direct Oblimin with Kaiser Normalization for the rotations. For the novice data, eight factors with Eigenvalues greater than 1.0 were found; for the expert data, there were four factors.
Interpretation of factors
The interpretation of factors is often “an art” because many items and combinations of items are multidimensional, which can lead to ambiguity among the factors. To facilitate the interpretation of factors, we modified the labels for each variable to include abbreviations indicating the food-friends-formality (FFF) dimensions and the wine types. The first three letters indicate the level (H = high, L = low) for each of the hospitality situation dimensions, food, friends and formality, respectively. The fourth letter indicates the wine type (R = red, W = white, S = sparkling). The procedure for interpreting these factors was based on:
the situation dimensions described above; and
the correlations among the factors.
First, we considered the attribute levels of the hospitality situations. Each factor was examined to determine if there were common attributes among the hospitality situations that constitute that factor. For instance, some factors may include situations that are all informal, and all of them involve sparkling wine. In such a case, we could easily label the factor “informal situations and sparkling wine.” This approach provides each factor with a name that is indicative of its predominant attributes. Second, we considered the correlations among factors. If factors correlated with related attributes (e.g. high/low formality), then their factor correlations may be an empirical basis for developing a taxonomy of wine hospitality situations.
Figure 6 presents the results for the expert group. The loadings of 0.5 or higher are shown, and the apparent grouping of product situation attributes (the four left-hand columns) are presented in bold type and outlined. In this case, the interpretation of the factors is clear and simple. Three of the four factors have only one consistent attribute among the situations, and in each case, the dominant attribute is the type of wine. Factor E1 (the first factor from the experts) is labeled “red wine,” factor E2 is “sparkling wine” and factor E4 is “white wine.” Factor E3 is a combination of “informal hospitality situations and red wine.” The experts’ perceptions of appropriateness are focused on intrinsic characteristics of the wine rather than on hospitality situations. This is consistent with results of Aurier and Ngobo (1999) and Viot (2012) that expert consumers attach more value to intrinsic characteristics of wine.
The factor correlations from the expert group indicate that red and white wine (factors 1 and 4) have a 0.57 correlation, while factor 3 (casual red) is not significantly correlated with any other factor. Additionally, the red and white wine factors have some strong cross-loadings. Factors 2 (sparkling) and 1 (red) are correlated at 0.29 and factors 2 and 4 (white) at 0.34. This pattern also suggests a two-level categorization of hospitality situations for expert consumers – differentiation between sparkling wines and still wines, as well as between red and white wine.
Figure 7 presents the results for the novice group. Seven of the eight factors clearly relate to two or more dimensions, so the names for them reflect the FFF dimensions and/or the wine type that reflects the predominate dimensions. The situations that comprise the first factor have a higher social content (many friends), generally have a high level of formality and food is generally peripheral to the situation. Factor N2 consists entirely of items related to sparkling wine and less formal situations. The factor N3 was also entirely sparkling wine-related, but casual situations (low formality). Factor N4 has situations involving many friends and generally low formality. The items loaded on factor N5 were informal and situations in which food is peripheral. Factor N6 items involve few friends. Factors N7 and N8 both involve white wine, while factor N7 items are informal and factor N8 items are high in formality. Clearly, the interpretation of these factors is more complicated than for the expert group.
Summary of Study 3
This exploration of the dimensions of hospitality situations found that the food-friends-formality paradigm was indeed present in these data. However, there is also a strong presence of the types of wines as dimensions, especially for the expert respondents. Consistent with previous research, we have found that relevant situational attributes often vary radically across individuals and products. This has led to some frustration among researchers as they strive to create taxonomies of situational attributes that can provide generalizable insights. Despite the appeal of such taxonomies, the “fickle nature of situation research” (Quester and Smart, 1998, p. 221) has made it difficult. Indeed, the endeavor has been labeled “futile” by Hornik (1982, p. 46). Therefore, understanding the influence of situations on consumer decision-making is more practical and realistic if it is limited to specific contingencies. In this research, we limited the scope to the wine category and hospitality situations (Figure 8).
General discussion and conclusions
This research investigated the influence of corks versus screwcaps on the perceived appropriateness of wines for various hospitality situations. This research found support for the person-product-situation framework as an approach to understand the wine consumers. An empirical test of the model found that all of the propositions from the theory are statistically significant. Among the person-product-situation influences on the appropriateness of closure types, hospitality situations have the most substantial effect. The second largest effect comes from the interaction of situations and expertise with the product. Corks were generally viewed as more appropriate for red wines. This is consistent with our expectation based on the association of corks with food consumption and wine aging traditions. Novice consumers have higher variability than experts in their assessments of appropriateness among hospitality situations. This may indicate that non-experts rely more on peripheral cues (e.g. corks or screwcaps) when judging the appropriateness of wines for various situations. These findings are consistent with the ELM of consumer decision-making, which argues that those who are less able to judge the intrinsic attributes of a product are more likely to rely on peripheral attributes that they can more easily understand. This research confirms the notion that hospitality situations are complex, and it also supports Halstead’s (2011) argument that neither corks nor screwcaps are likely to become the only type of wine closure. Instead, the use of corks and screwcaps is evolving to meet the needs of different types of wine, occasions and consumers.
Limitations, future research and industry implications
Certain limitations should be noted. First, this study was conducted in only one (USA) market. As wine consumption traditions are often deeply embedded in regional cultures, it may be risky to generalize these results to other cultures. We believe that expertise and intrinsic product characteristics probably influence judgments about wine consumption in all cultures, but the cultural meaning of corks and screwcaps may vary considerably across markets. That underscores the importance of further investigations of this issue in other cultures, as Atkin et al. (2006) noted, there are considerable differences in the acceptability of screwcaps in different countries. Additionally, the meaning associated with corks or screwcaps has changed over time, emphasizing the need to replicate similar studies across time. These differences can influence consumption behavior and should impact the marketing strategies of wineries and retailers.
The hospitality dimensions provide a useful framework for future research. There are opportunities for future research to examine the interaction of closures and situations and intrinsic product attributes such as different types of packaging or product labels that could impact perceptions and behavior.
The findings regarding hospitality and formality are particularly relevant for marketing strategies. Hospitality often includes ceremonial consumption (e.g. different forks for salads, main courses and deserts), which are meaningful to customers and often provide signals of other attributes such as quality or formality. The nature of closures for wine products is another of those signals to customers. This is important for hospitality service providers as they decide on the particular combinations of food/beverage attributed to put before their customers. Also, distribution decisions made for wine products can benefit from these results. Some channels are more likely to reach the market for home consumption, which is more likely to be a venue for informal consumption. Some channels may be more likely to reach hospitality venues where formal consumption is typical. These results can also benefit producers who may make their product/market decisions (positioning) via extrinsic attributes such as the type of packages, the labels and the closures for their wines. These often provide signals to customers who have to rely on extrinsic cues to judge the products.
Perceived appropriateness of corks (+4) or screwcaps (−4)
|Test hypotheses for each category: mean = 0|
|Constructs||Categories||Category means||Difference within constructs|
|Expert||1.081*||F = 9.87|
|Novice||0.923*||p = 0.002|
|Wine types||Red||1.166*||F = 20.66
p < 0.001
|Hospitality situations||Restaurant with clients||2.075*|
|Dinner for two||1.580*|
|Attending a dinner party||1.560*|
|Holiday events||1.462*||F = 197.13|
|Dinner party with friends||1.053*||p < 0.001|
|Dinner party at home||0.924*|
|Going out with friends||0.442*|
|Casual get together||0.001|
|Casual everyday use||−0.426*|
*Difference from 0 sig. at p ≤ 0.05
Study 1 ANOVA results
|Dependent variable: appropriateness|
|Type III sum of squares||df||Mean square||F||Significance||Partial ƞ2||Source effects|
|RH1 Hospitality situation||5,797.4||11||527.0||156.3||0.000||0.181||Medium|
|RH2 Wine type * hospitality situation||130.6||22||5.9||1.8||0.015||0.005||Trivial|
|RH3 Wine type||191.8||2||95.9||28.4||0.000||0.007||Trivial|
|RH4 Hospitality situation * expertise group||1,863.8||11||169.4||50.2||0.000||0.066||Small|
|RH5 Expertise group||47.4||1||47.4||14.1||0.000||0.002||Trivial|
R2 = 0.281 (Adjusted R2 = 0.277)
|Value range||Effect size|
|0.26 ≤ ƞ2||Large|
|0.13 ≤ ƞ2 < 0.26||Medium|
|0.02 ≤ ƞ2 < 0.13||Small|
|ƞ2 < 0.02||Trivial|
|“Food” dimension||“To what degree does this activity involve food consumption?”||Never/Always|
|“Friends” dimension||“To what degree is this event social?”||Not at all social/Extremely social|
|“Formality” dimension||“To what degree is this activity formal?”||Not at all formal/Extremely formal|
Study 2 ANOVA results
|Dependent variable: appropriateness|
|Source||Type III sum of squares||df||Mean square||F||Sig.||Partial ƞ2||Source effects|
|Food * formality||45.408||1||45.408||13.224||0.000||0.002||Trivial|
|Food * friends||78.307||1||78.307||22.805||0.000||0.003||Trivial|
|Friends * formality||889.743||1||889.743||259.120||0.000||0.032||Small|
|Expert group * food||99.509||1||99.509||28.980||0.000||0.004||Trivial|
|Expert group * formality||1,530.318||1||1,530.318||445.675||0.000||0.054||Small|
|Expert group * friends||94.668||1||94.668||27.570||0.000||0.004||Trivial|
|Food * wine type||14.261||2||7.130||2.077||0.125||0.001||Trivial|
|Formality * wine type||30.077||2||15.038||4.380||0.013||0.001||Trivial|
|Friends * wine type||6.302||2||3.151||0.918||0.400||0.000||None|
aR2 = 0.266 (Adjusted R2 = 0.264)
A “route” is the sequence and type of information used in making an assessment or decision.
Partial ƞ2 is the proportion of variance associated with one or more main effects, errors or interactions that are included in an ANOVA.
Standardized means (to control for individual respondent tendencies) were also examined in this way; both approaches produce the same result.
Alba, J. and Hutchinson, J. (1987), “Dimensions of consumer expertise”, Journal of Consumer Research, Vol. 13 No. 4, pp. 411-454.
Atkin, T., Garcia, R. and Lockshin, L. (2006), “A multinational study of the diffusion of a discontinuous innovation”, Australasian Marketing Journal, Vol. 14 No. 2, pp. 17-33.
Atzmüller, C. and Steiner, P.M. (2010), “Experimental vignette studies in survey research”, Methodology: European Journal of Research Methods for Behavioral and Social Sciences, Vol. 6 No. 3, pp. 128-138.
Aurier, P. and Ngobo, P.V. (1999), “Assessment of consumer knowledge and its consequences: a multi-component approach”, Advances in Consumer Research, Vol. 26, pp. 569-575.
Barber, N. and Almanza, B.A. (2007), “Influence of wine packaging on consumers’ decision to purchase”, Journal of Foodservice Business Research, Vol. 9 No. 4, pp. 83-98.
Bruwer, J., Perez, P.A.A. and Cohen, J. (2017), “Restaurants and the single-serve wine by-the-glass conundrum: risk perception and reduction effects”, International Journal of Hospitality Management, Vol. 62, pp. 43-52.
Bugge, A.B. (2005), “Domestic dinner – a result of social and cultural processes”, paper presented at the 7th Conference of European Sociological Association ‘Rethinking Inequalities’, Torun, 9-12 September.
Charters, S., Velikova, N., Ritchie, C., Fountain, J., Thach, L., Dodd, T.H., Fish, N., Herbst, F. and Terblanche, N. (2011), “Generation Y and sparkling wines: a cross-cultural perspective”, International Journal of Wine Business Research, Vol. 23 No. 2, pp. 161-175.
Cohen, J. (1988), Statistical Power Analysis for the Behavioral Sciences, 2nd ed., Lawrence Erlbaum Associates, Hillsdale, NJ.
Dodd, T.H., Laverie, D.A., Wilcox, J.F. and Duhan, D.F. (2005), “Differential effects of experience, subjective knowledge and objective knowledge on sources of information used in consumer wine purchasing”, Journal of Hospitality and Tourism Research, Vol. 29 No. 1, pp. 3-19.
Hall, J., Lockshin, L. and O’Mahony, G.B. (2001), “Exploring the links between wine choice and dining occasions: factors of influence”, International Journal of Wine Marketing, Vol. 13 No. 1, pp. 36-53.
Halstead, R. (2011), “Screwcap acclaimed but still finding its identity”, Wine and Viniculture Journal, Vol. 5, pp. 88-91.
Hornik, J. (1982), “Situational effects on the consumption of time”, Journal of Marketing, Vol. 46 No. 4, pp. 44-55.
Lange, C., Issanchou, S. and Combris, P. (2000), “Expected versus experienced quality: trade-off with price”, Food Quality and Preference, Vol. 11 No. 4, pp. 289-297.
Lashley, C. and Morrison, A. (2000), In Search of Hospitality: Theoretical Perspectives and Debates, Butterworth-Heinemann, Boston, MA.
Marin, A.B. and Durham, C.A. (2007), “Effects of wine bottle closure type on consumer purchase intent and price expectation”, American Journal of Enology and Viticulture, Vol. 58 No. 2, pp. 92-201.
Marin, A.B., Jorgensen, E.M., Kennedy, J.A. and Ferrier, J. (2007), “Effects of bottle closure type on consumer perceptions of wine quality”, American Journal of Enology and Viticulture, Vol. 58 No. 2, pp. 182-191.
Mueller, S., Francis, L. and Lockshin, L. (2008), “The relationship between wine liking, subjective and objective knowledge: does it matter who is in your ‘consumer’ sample?”, paper presented at the 4th Academy of Wine Business Research Conference, Siena, 16-19 July.
Mueller, S., Osidacz, P., Francis, I.L. and Lockshin, L. (2010), “Combining discrete choice and informed sensory testing to determine consumer response to extrinsic and intrinsic wine attributes”, Food Quality and Preference, Vol. 21 No. 7, pp. 741-754.
Orth, U.R. (2006), “Difference brands for different occasions – drivers of consumer public and private choices”, paper presented at the 3rd Academy of Wine Business Research Conference, Montpellier, 6-8 July.
Petty, R. and Brinol, P. (2011), “The elaboration likelihood model”, in Paul, A.M., Van Lange, A.W., Kruglanski, E. and Higgins, T. (Eds), Handbook of Theories of Social Psychology, Vols 1 and 2, Sage Publications, Singapore.
Quester, P.G. and Smart, J. (1998), “The influence of consumption situation and product involvement over consumers’ use of product attribute”, Journal of Consumer Marketing, Vol. 15 No. 3, pp. 220-238.
Scott, D.G. (2006), “Socialising the stranger: hospitality as a relational reality”, Master thesis, University of Otago, Dunedin.
Slama, M. and Tashchian, A. (1987), “Validating the S-O-R paradigm for consumer involvement with a convenience good”, Journal of the Academy of Marketing Science, Vol. 15 No. 1, pp. 36-45.
Smith, D.C. and Park, C.W. (1992), “The effects of brand extensions on market share and advertising efficiency”, Journal of Marketing Research, Vol. 29 No. 3, pp. 296-311.
Thach, L. (2011), “Wine for breakfast: exploring wine occasions for gen Y”, paper presented at the 6th Academy of Wine Business Research Conference, Bordeaux, 9-11 June.
Viot, C. (2012), “Subjective knowledge, product attributes and consideration set: a wine application”, International Journal of Wine Business Research, Vol. 24 No. 3, pp. 219-248.
Williams, A. (2002), Understanding the Hospitality Consumer, Butterworth-Heinemann, Boston, MA.
Wilson, D. (2009), “The silent debate – closure options in continental Europe”, Australian and New Zealand Wine Industry Journal, Vol. 24 No. 5, pp. 59-60.