The influence of extrinsic cues on wine purchase in Hong Kong
2020
TABLE OF CONTENTS
Abstract
Declaration
Acknowledgements
Table of contents
List of abbreviations
List of tables
List of figures
CHAPTER ONE – INTRODUCTION
1.1 Background
CHAPTER TWO – LITERATURE REVIEW (Working title*)
2.1 Introduction
2.2 Brand Recognition
2.3 Country of Origin
2.4 Critic Scores and Wine Ratings
2.5 Label/Packaging design (LPD)
2.6 Pricing
CHAPTER THREE – RESEARCH GAP, DESIGN AND METHODOLOGY
3.1 Conceptualisation
3.1.1 Aims and Objectives
3.1.2 Research Philosophy
3.1.3 Research Design
3.2 Research Methodology
3.2.1 Preamble
3.2.2 Research Design
3.2.3 Qualitative Method – Phase 1
3.2.3.1 Sampling Approach
3.2.3.2 Data Collection
3.2.3.3 Qualitative Data Analysis
3.2.4 Quantitative Methodology
3.2.4.1 Design
3.2.4.2 Sampling
3.2.4.3 Data Collection Instrument
3.3 Ethical Considerations
3.4 Data Analysis
3.4.1 Measures used for Consumer Behaviour
3.4.1.1 Subjective Knowledge
3.4.1.2 Objective knowledge
3.4.1.3 Self-confidence
3.4.1.4 Survey Data Analysis
3.5 Attribute Importance Determination
3.6 Validity of Research Instruments
3.7 Pilot Study
3.7.1 Validation of the research Instruments
3.7.1.1 Subjective knowledge and self confidence
3.7.1.2 Objective Knowledge Items (pilot)
3.7.1.3 Sensory Experiment-Focus Groups
3.7.1.4 Sample Demographics
3.8 Summary
CHAPTER FOUR – RESULTS AND ANALYSES
4.1 Introduction and Overview
4.2 Response Rates and Distribution of Surveys
4.2.1 Distribution of Surveys
4.2.2 Response Trends
4.3 Screening and analysis of data from Focus Group
4.3.1 Descriptive statistics on critic scores derived from Focus Group
4.3.2 Correlation test results for Critic Scores
4.3.3 Demographic characteristics of the Focus Group
4.3.4 Type of Wine
4.3.5 Different wines for different occasions
4.3.6 Factors about the kind of wine to drink or buy
4.3.7 Source of best wines
4.4 Data screening and analysis on Part I-III Findings
4.5 Data screening and analysis on Survey Findings
4.6 Test of Significance
4.7 Review of Key Findings
4.8 Summary of Chapter
CHAPTER FIVE – DISCUSSION AND CONCLUSION
CHAPTER 1 – INTRODUCTION
In 2007, HKSAR Financial Secretary Henry Tang, halved the excise duty on wine and liquor (Alcohol By Volume ABV<30%) to 40 per cent duty rate. The next year when he became Chief Secretary, his successor as Financial Secretary Mr. John Tsang went even further and abolished the duty. The wine market in Hong Kong presents tremendous opportunities for growth and expansion of business for wineries and marketers ever since. Hong Kong has grown as a top wine import/export and auction hub in the world, taking over from London and New York, while auction houses such as Sotheby and Christie have been pushing Asian buyers for new business as prices drop in the wake of global economic downturn and as luxury spending dwindles.
An extensive research is critical in understanding the characteristics of the Hong Kong market after the ten years of duty abolishment. The primary purpose of the proposed study is to examine the perceptions of wine purchase among the non-wine educated consumers. The proposed research is to investigate the consumers’ attitudes towards 1) Brand Recognition, 2) Country of Origin, 3) Critic Scores and Wine Ratings, 4) Label/Packaging Design and 5) Pricing. Fundamentally, the study will aim to establish the impacts of aforementioned factors on the purchase behaviours of the consumers in Hong Kong and rank the importance of the five factors. It is vital to note that the proposed study will make immense contribution to the research on wine consumer behaviours, thereby helping marketers and wineries to position their brands in the Hong Kong market.
CHAPTER 2 – LITERATURE REVIEW (Working title)
2.1 Introduction
Wine is one of the products that are consumed almost across all cultures globally. Its existence in many different markets with consumers possessing different characteristics and perceptions implies that conducting the wine business on a global scale is a complex process. This is because producers need to consider the different preferences of consumers in different markets so as to ensure sustainable sales for their products. According to Kelley et al. (2015) and Brochado et al. (2015), the decision-making process, buying behaviour, and client perception in the wine market have a robust relationship with Brand Recognition, Country of Origin, Critic Scores and Wine Ratings, Label/Packaging Design and Pricing. However, Gartner et al. (2017) reveal that these external cues have different levels of impact as far as influencing the purchase decisions of consumers is concerned.
The prominence of most wine companies has mainly been pinned on strong brand image that resonates well in the consumers’ mind. It has however emerged in previous research in different markets that brand image alone cannot account for all the purchases by consumers. It has for example, been found in research that some customers choose specific wines over others as a result of critic grading (Mueller et al., 2011), while others are more concerned about the country of origin (Vesela and Zich, 2015). This literature review will focus on investigating the what previous research has found in relation to the influence of eternal cues; Brand Recognition, Country of Origin, Critic Scores and Wine Ratings, Label/Packaging Design and Pricing – and their influence on wine buyer intentions. In this literature review, the specific focus will be to present findings from research studies conducted in the global market and compare them with the general observations in the Hong Kong market. This comparison will be essential in revealing any gaps in the literature that might in relation to the external cues that influence the wine purchase decisions of consumers in Hong Kong.
2.2 Brand Recognition
A study conducted in the wine sector of Portugal by Brochado et al. (2015) to determine the link between consumer perception and brand recognition found that the latter greatly depends on the knowledge of a customer regarding a particular product. The scholars described brand recognition as the ability to identify or spot a certain product from a variety of commodities based on its design components and features. Brochado et al. (2015) further realised that developing an awareness system and robust brand recognition have become increasingly important for corporate bodies in their bid to sustain the public brand image. Beside brand recognition, Heslop, Cray, and Armenakyan (2010) found personality trait as another factor that influence the perceptions of the consumers as regards to their readiness to buy the product.
While brand recognition is strongly related to the perceptions of customers towards a particular wine product, Vlachevei et al. (2012) found that it does more than that. The authors in their review of the literature found that brand recognition is important in securing the reputation of the product thus reducing the likelihood of the existence of imitations in the market. An important question is how this protection from imitations leads to sustained win purchase in the market. In response to this, Vlachevei et al. (2012) go ahead to point out that having no imitations means that consumers will be able to trust the quality of a particular brand more than the quality of a brand that has multiple imitations in the market. This confidence with the authenticity of a product translates to positive brand perceptions in the market which leads to the likelihood of purchase. From the argument presented by Brochado et al. (2015), it emerges that the customer’s knowledge of a product positively influences brand recognition. Study of Vlachevei et al. (2012) reveals that customers can only be confident about their knowledge of a brand that has not been linked with any imitations. In that regard, it can be argued that the guarantee of authenticity of a brand leads to a high brand recognition and a good brand reputation which translate to higher purchase intentions by customers in markets.
With the link between the brand recognition and wine purchase intentions among customer established, it is important to investigate why brands are important in wine sales. In a research study by Viot and Passebois-Ducros (2010), it was found that one of the primary roles of branding is that it helps in the identification of the origin of different wines. This finding strongly concurs with the findings presented by Vlachevei et al. (2012) showing that brand recognition is influenced by the place of origin of the wine. This implies that brand names that identify a highly trusted wine production region or producer are likely to positively influence the purchase intentions of people in different markets. The second important aspect of brand recognition that is found to stimulate purchase as found by Viot and Passebois-Ducros (2010) is the prestige and image associated with particular brands. According to the authors, most well recognised wine brands are associated with class and prestige. Individuals that yearn for such distinction are thus influenced by the brands when making purchase decisions. The finding of the strong link between specific wine brands and prestige is corroborated by Sjostrom et al. (2014) and Heine et al. (2016) who all point to more sales for brands recognised as prestigious in markets with individuals seeking distinction.
Since this research will specifically focus on wine sales in a particular region of the world, it is essential to investigate how brand recognition influences purchase intentions in different regions and then narrow down to the area of interest. In their review of the literature, Forbes and Dean (2013) found that brand names play a very important role in the decision to purchase wine by different customers in the German market. For Chinese consumers, on the other hand, different naming strategies were observed to lead to different levels of sales for different wine products. In both countries, therefore, the selection of brand names is found to be one of the most important considerations in the success of wine companies. In their specific research study, Forbes and Dean (2013) found that international brand names in most markets were associated with high quality, thus attracting high-class customers. Geographical names, on the other hand, elicited perceptions of trustworthiness of the product among the study participants, thus meaning that geographically linked brand names would be the most appropriate for new products. Much more generally, Robertson et al. (2018), in their research study, observed that most wine consumers in different countries make their purchase decisions depending on how well a particular brand is known in the market. This implies that, regardless of the market, wine companies can stimulate brand recognition through the choice of appropriate brand names (Forbes and Dean, 2013) and investing in marketing approaches that are effective in the respective countries (Robertson et al., 2018).
When it comes to the evaluation of the wine market, specifically in Hong Kong, very few recent sources investigate brand recognition in the region. The few available sources presenting information in this direction investigate other issues relating to the purchase of wine other than brand recognition. However, sources such as Tang and Cohen (2014) can be used in making some important inferences about the nature of the Hong Kong market as far as wine brands are concerned. According to the authors, people in Hong Kong are strongly influenced by western culture when it comes to their wine preferences. Tang and Cohen (2014) went ahead to point out that brand evaluations play an important role in the purchase of wine by consumers in the Hong Kong market. This implies that well-known western brands whose evaluations are presented on different platforms, such as mass and social media, would be successful in the Hong Kong market. The appreciation of foreign wine brands in the Hong Kong market was also highlighted by Lau et al. (2014), who revealed that the city removed import tax on wine, thus making it an important market for most producers all over the world. In order to boost brand recognition for wine products in the region, Lau et al. (2014) pointed out that the Hong Kong government has made efforts by organizing events such as the Hong Kong International Wine Expo and encourage participation in events such as the China International Wines and Spirits Exhibition. While there are no explicit mentions about how these efforts lead to wine brand recognition in Hong Kong, it can be inferred that such marketing efforts make wine products known in the market, thus boosting brand knowledge among consumers. This is in line with the argument by Robertson et al., (2018) that investment in different marketing approaches leads to improved brand recognition.
The sources above show that indeed brand recognition is an important element in the process of making purchase decisions by customers. Brand recognition is also an important determinant of purchases in the specific context of Hong Kong. A closer investigation of the factors that lead to brand recognition is thus essential if practical recommendations about branding are to be made for the Hong Kong market. In their research study, Sjostrom et al. (2014) found that brand recognition in the wine industry results from factors such as status associated with the specific brand and brand reputation. It is the brand reputation, for example, that enables individuals to share details about a specific brand of wine, thus increasing its popularity in different markets. In that regard, a wine brand is able to gain recognition based on how many existing customers stimulate other customers to make purchases. Other important elements of brand recognition include the origin of the brand (Vlachevei et al., 2012) and the level of promotional activities (Sjostrom et al., 2014). Since the place of origin is highlighted by Vlachevei et al. (2012), it is plausible to assume that wines from specific regions in the world are more reputable than those in others, meaning that different levels of purchases should be expected depending on the origin. While the origin is a very important element influencing brand recognition, it will be discussed more exhaustively in the next section.
2.3 Country of Origin
From the previous section, it has emerged that the country of origin of wine is an important factor in determining how consumers view different brands. This means that the success of any given brand is, to some extent, dependent on the country of origin of the specific brand. The relevance of product origin in brand recognition, as presented by Vlachevei et al. (2012) is an indication that brands from different countries are perceived differently by consumers in different markets. In corroboration of this argument, Garcia-Gallego et al. (2015), in their review of the literature, presented numerous sources indicating that the place of origin has emerged as one of the most important differentiating factors for wine in most markets throughout the world. While in their research studies, no explicit comparison of the perception towards the wine from different countries was made, Garcia-Gallego et al. (2015) found that a region’s image as far as wine production is concerned was significant in leading to positive perceptions towards specific win brands. These perceptions lead to higher purchase intentions for wines from particular regions as opposed to others. The findings by Garcia-Gallego et al. (2015), therefore, confirm the argument by Vlachevei et al. (2012) that brand recognition for wine is dependent on the country of origin of the particular wine. This explains why it is common for wine producers to indicate the origin of wine on labels, as pointed out by Garcia-Gallego et al. (2015).
While Garcia-Gallego et al. (2015) took a general approach in investigating the relationship between country of origin and wine purchase intentions, a number of other researchers were found to take a more in-depth stance. Vesela and Zich (2015) investigated purchase intentions in the light of ethnocentrism. From their evaluation of responses by interviewees, it emerges that factors such as patriotism, history, and culture have a profound impact on a person’s likelihood to purchase wine from a particular region or country. The authors, for example, revealed that the Europeans were opposed to purchasing products from South Africa as a result of the Apartheid history. This is an indication that wine brands that are in any way linked to South Africa would not appeal to individuals that consider the country’s history as important in their purchase decisions. With regard to the patriotism argument, on the other hand, Vesela and Zich (2015) indicate that foreign wine brands would find it difficult to penetrate markets that have a strong wine production culture since most people would more readily support local brands.
Although Tamas (2016) did not use the ethnocentric lens in evaluating how people in different countries would prefer to purchase wines from particular countries as opposed to others, his findings concurred with those presented by Vesela and Zich (2015) by showing that specific attitudes towards the wine from different nations influenced purchase intentions. France was found to be the country with the most widely known wine tradition and also with the most famous wine brands. In addition, the respondents in the study by Tamas (2016) indicated that France had the best tasting wine followed by Italy and then Germany. From these results and the argument by Vlachevei et al. (2012) on the relationship between brand recognition and country of origin, it is plausible to argue that wine brands from France would be more readily accepted in different global markets compared to wine brands from other countries. However, according to the ethnocentric approach employed by Vesela and Zich (2015), this effect cannot be observed in all countries since other factors such as perceptions of the target countries towards France and patriotism would play a moderating role in purchase intentions. This means that it is important to consider ethnocentric elements before making the decision to launch wine brands from specific countries in a particular market.
With the general influence of place of origin found to be an important predictor of customer purchase intentions, it is imperative to investigate how the factor has been observed to influence different markets globally. In the specific context of Lithuanian consumers, Casas and Makauskiene (2013) observed that a country’s perceived competence in wine production is one of the most important predictors of wine sales in the country. This implies that wine from countries that have a more elaborate wine production culture would penetrate the Lithuanian market much faster than wine from countries that are not famous for wine production. Bringing in the finding by Casas and Makauskiene (2013) in the context of the finding Tamas (2016) that France is perceived to be the most famous wine producer globally, it is plausible to argue that French wine brands would be very successful in Lithuania. In a much different research study, Defrancesco et al. (2012) investigated the performance of wine brands labelled as being from Argentina in markets such as the US, Germany, Netherlands, and the UK. The authors found that geographical names were more likely to succeed in the US and UK compared to Germany and the Netherlands. However, the authors also found that a blend Malbec wine that had no geographical indication linking it to Argentina would be more likely to be sold at a premium in Germany and the Netherlands compared to the same wine but with a geographical indicator.
From the evaluation of different wine-producing countries in the world by Vesela and Zich (2015), many European countries such as Germany have an elaborate wine production history compared to the US and the UK. What this means is that Germans would be inclined to purchasing wine produced in Germany or at least in Europe as opposed to countries such as Argentina. The findings by Defrancesco et al. (2012) that wines from Argentina with geographical indications would be more successful in countries like the UK, and the US clearly confirms the ethnocentric aspect of wine purchase presented by Vesela and Zich (2015). Ethnocentrism, especially the patriotism aspect, implies that consumers, as a sign of solidarity with their local wine industry, would, to some extent, shun exotic brands that explicitly indicate their origins as being other countries. Following this argument, it would be expected that wines with Argentine geographical names would not perform very well in the French market since there is already a very strong wine production culture in the country. In such a market, ethnocentrism in most consumers, as presented by Vesela and Zich (2015), would lead them to interpret the existence of exotic brands as a threat to their wine industry. Any patriotic individual in France would, therefore, be unlikely to purchase Argentine wine with geographical indicators as a sign of loyalty to their country. However, this is not to say that it is impossible to sell Argentine wine in countries with very rich wine production histories. According to Defrancesco et al. (2012), wine exports from Argentina to Europe, and other regions have experienced accelerated growth in recent times. This can be attributed to the increased globalisation where people from different parts of the world interact in different countries, thus leading to cultural exchange. In the specific context of globalisation, therefore, it is likely to have many people in European countries that have positive perceptions towards Argentine wine, thus leading to high purchase rates in traditional wine-producing countries despite the existence of patriotic locals.
It has emerged that the issue of country of origin in relation to wine purchase is of profound importance. At this point, a more specific investigation of sources looking at the impact of country of origin in wine purchase in Hong Kong is necessary. However, most research studies that have been conducted on the subject matter look at Chinese market in its entirety rather than Hong Kong only. Li et al. (2011), for example, found that country of origin is a very important factor in the purchase of wine among Chinese consumers. More specifically, the authors found that the country of origin is a stronger guarantee of the quality of wine in the Chinese market compared to other extrinsic factors such as brands. In corroboration of the findings by Li et al. (2011), Bobik (2014) revealed that the aspect of the country of origin is so important in Chinese wine sales such that wine shops organise products in terms of country of origin as opposed to the specific type of wine. From these sources, therefore, it is evident that similar to other markets all over the world, the country of origin is an important factor in wine purchase among Chinese consumers. However, there are hardly any sources investigating the perception of Chinese consumers towards wines from different countries. This implies that it is difficult to ascertain which geographical indicators would ensure the success of wines in the market. Also, there are hardly any sources looking specifically at the level of interest of consumers in Hong Kong to show the country of origin in making their decisions to purchase wine. This implies that there is a potential gap in this area that needs to be addressed by researchers, especially through the ethnocentric lens presented by Vesela and Zich (2015).
2.4 Critic Scores and Wine Ratings
Apart from brand recognition and country of origin, ratings, and scores of wine by different wine experts have been observed to influence purchase decisions in different markets throughout the world. An important question at this point would be why such opinions would be so strong in influencing purchase decisions. In response, Ellis (2015) pointed out that unlike other products, it is difficult to assess the attributes of wine simply by looking at it. For that reason, consumers need as much information as possible before they make decisions to purchase a particular wine brand. In an evaluation of why consumers choose specific wines over others, Mueller et al. (2011) noted that most wine consumers always find themselves overwhelmed by the different wine varieties that are stocked in different outlets. Selecting from these different types, especially where one does not know most of the brands can be a very challenging process for consumers. In that regard, information such as critic scores and star ratings become very important in guiding consumers in the purchase process. Ellis (2015) and Mueller et al. (2011), therefore, reveal that scores by critics and ratings are important for customers since they assist them in choosing specific wines out of a large number of options. With the importance of ratings and critic scores thus established, it is important to investigate how such cues specifically influence purchase decisions.
The study by Chocarro and Cortinas (2013) investigated the effect of reviews by experts on the customers’ buying decisions for red wines. The researchers discovered that the information from the experts eases the judgment process of the customers based on the existing brands. The findings of Chocarro and Cortiñas (2013) have been supported by Carsana and Jolibert (2017), who argued that critics and expert opinions greatly inform the decision-making process of the shoppers. Some of the aspects highlighted by experts and critics and considered significant by the customers include sensory definitions, expert grading, star quality ratings, and consistent awards. By extension, therefore, Chocarro and Cortiñas (2013) and Carsana and Jolibert (2017) confirm arguments by Ellis (2015) and Mueller et al. (2011) that expert ratings are important in helping consumers make decisions on which wines to purchase especially when they are faced with many different options. In this regard, assuming two wine brands are sold at the same price, consumers will generally prefer to purchase the one with the best expert rating, thus showing that favorable ratings by trusted critics are likely to be a source of competitive advantage in the wine market. In their research study, Mueller et al. (2011) found that a high level of agreement between wine critics of the high quality of different wine brands led to a 9.8% increase in the number of people choosing these brands over others. This clearly shows that critic scores have a profound impact on consumer choices. It can, therefore, be argued that good ratings by well-respected wine critics would increase the chances of a wine brand succeeding in any market throughout the world.
In a more specific research study, the Global Wine Score team in a report presented in GWS (2018) investigated the impact wine scores had on the prices of wines from Napa Valley. The findings indicated that the higher the wine rating, the higher the price at which a particular wine could be sold. This finding shows that people would be willing to spend more money on wines with very high scores as opposed to similar wines with low scores. For that reason, in addition to boosting sales, good critic scores have the effect of improving profit margins for wines with good ratings. The findings are highly consistent with those by Mueller et al. (2011) that high scores by critics lead to a significant gain in the purchase of specific brands. Even more interesting results were presented by Snow and Weckman (2018) with regard to wine scores in the entire US. Their findings revealed that while critic’s scores were not important in the purchase decisions of cheaper wine brands, critic scores were very important with regard to how people purchase more expensive wines. More specifically, the authors found that the consumers buying wines costing $20 and less did not pay attention to critic scores while those purchasing wines costing $50 and above, almost always referred to critic scores before making purchase decisions. These findings have serious implications for wine brands seeking to venture into new markets. Producers making premium wines for high-end markets will need to always seek critic scores so as to make their products more trusted in the targeted market segments.
Sources such as Mueller et al. (2011), Chocarro and Cortiñas (2013), Ellis (2015), Carsana and Jolibert (2017), and Snow and Weckman (2018) all agree that ratings by critics and wine experts influence the purchasing behavior of wine consumers. Ellis (2015) and Mueller et al. (2011), also went ahead to highlight that the lack of information by consumers about the many wine brands available in the market is one of the reasons why most people rely on assessments by critics when making purchase decisions. However, all of these sources fail to highlight what specifically these ratings tell consumers so that they can purchase specific wines as opposed to others. According to Lee (2012), one of the main intrinsic factors that are strongly linked to extrinsic cues is wine quality. For example, Lee (2012) revealed that just by raising the price of a wine brand, potential consumers would see it as being of high quality, thus leading to increased purchases. This situation, however, makes it possible for unscrupulous winemakers to set high prices for low-quality wines. Wine critics thus help in avoiding such situations by offering objective wine ratings against predetermined industry standards (Lee, 2012). For that reason, instead of relying on the price as a means of determining the quality of wine, consumers are able to get accurate assessments of wine quality from assessments by experts and critics.
While research evidence shows that critic scores and ratings are of profound importance in wine purchases in different countries globally, the number of sources specifically investigating this cur in the specific context of China is very low. In contrast to the findings of the importance of wine scores as the sole indicator of quality among consumers from different countries, as presented by Lee (2012), Xiong and Li (2017) found that people in China relied on a wide range of sources for information about different wine brands. Some of these sources are the internet, professional books, and expert evaluations. Findings by the author also revealed that people in China are more guided to purchase different wine types by the degree to which they associate the wines with different situations as opposed to purely evaluations by experts. This implies that the first main determinant of purchase intentions among the Chinese is their particular situation rather than expert ratings. Other sources such as Muhammad et al. (2013) completely ignored the role of expert and critic ratings in driving the demand for different wine brands in China. While this could suggest that these ratings are not important in the Chinese market, it is safe to assume that specifically focusing on customer ratings in research would reveal important findings in relation to how critics can influence purchase intentions among consumers.
A search of sources specifically investigating the relationship between consumer purchase intentions and critic scores in Hong Kong shows that there has been hardly any such research for this market. A number of inferences can, however, be made from findings presented in relation to other markets. From the findings by Lee (2012), it is plausible to assume that the general link between high critic scores and perceptions of quality will lead to highly-rated international brands selling at premium prices in the Hong Kong market. This is due to the fact that Hong Kong is a cosmopolitan city and so people from other parts of the world in the region still show a strong trust of critic assessments when purchasing wine. Following the findings by Snow and Weckman (2018), on how critic scores are more important for higher-priced wines and less important for low-priced wines, it is plausible to assume that expert evaluations of cheap wines in Hong Kong will not lead to any significant changes in sales. However, evaluations of expensive wines will lead to profound changes in the level of sales. Despite the fact that sources looking at the impact of critic evaluations generally might provide information that can be used to accurately determine what should be expected in specific markets, it is important to address the issue within the specific target market. For that reason, the inexistence of sources relating wine purchase intentions to critic ratings in Hong Kong presents an important gap in the literature that needs to be bridged, research in this specific market should be very important in providing insights on the strategies new entrants need to implement so as to succeed in Hong Kong.
2.5 Label/Packaging design (LPD)
According to Hirche and Bruwer (2014), the colour and the appearance of the packaging materials are some of the brand awareness cues that cannot be ignored as being influential in changing consumer perceptions. The cues have been given a general term – the label/packaging design (LPD). Wine manufacturers tend to modify the appearance of wine bottles to attract shoppers. Captivating label/packaging designs increase the buyers’ probability of purchasing certain commodities. Such designs are commonly applied to communicate or reveal the uniqueness of the product to the customers. Some of the LPD features that attract the attention of the shoppers include the manufacture date, ingredient variety, techniques of consumption, and the winery’s history. The study conducted by Hirche and Bruwer (2014) revealed that the packaging design (PD) was one of the critical determinant factors that impacted on the buying patterns of wine shoppers. The researchers narrowed their study to specific elements of PD – the bottle design as well as the back and front label. These features have been integrated into the marketing strategies of firms as they facilitate the differentiation of products. According to Heslop et al. (2010), the use of wine among the millennials is mainly triggered by LPD components, such as flavour, alcohol volume, vintage, available ingredients, and the history of the winery. Generally, therefore, Hirche and Bruwer (2014) and Heslop et al. (2010) in their respective research studies, present the notion that LPD components are of great importance in influencing the purchase intentions of consumers in different markets.
The superficial features such as bottle shape and colour and label design are the first elements of wine that a customer observes while making a purchase decision. This means that the appearance of a wine bottle should be the first thing that attracts a person towards a particular product. In support of this argument, Elliot and Barth (2012) found that the millennial population is showing an increased interest in design features in the selection of wine and so the importance of LPD will only increase with time as one of the important factors in wine purchase. Henley et al. (2011), on the other hand, found that the design of wine bottles and labels are correlated to perceptions of product quality among millennials. This implies that this population would readily pay premium prices for wines in very stylish and trendy bottles as opposed to ones that are not perceived to be stylish and trendy. As a result of this observation, winemakers have invested heavily in the development of uniquely designed bottles and labels in different markets. The findings by Henley et al. (2011) and Elliot and Barth (2012) also go ahead to indicate that very attractive external designs would lead to better brand visibility in markets made up predominantly of millennials. For that reason, brand marketing strategies would significantly benefit from exceptional bottle and label designs, especially in markets with young populations. However, it is important to note that the findings by Henley et al. (2011) and Elliot and Barth (2012) only speak for the millennial market segment. For that reason, it is impossible to generalise such findings to all populations in all markets. This means that sources presenting general information on how LPD influences purchase intentions in the general population are necessary.
According to Mueller and Szolnoki (2010), multiple past studies have shown that the design of wine packages and labels influence the taste expectations and quality perceptions of consumers. Generally, therefore, very well designed bottles and labels are most definitely associated with positive purchase intentions among customers in any market. In their specific research study in two US markets, Mueller and Szolnoki (2010) found that the design of wine packaging and labelling has a significant impact on the price at which wines are sold. This finding is consistent with the one presented by Henley et al. (2011), showing that unique design features on wine bottles and labels allowed specific wine brands to be sold at premium prices. The corroboration of the results between the two sources is an indication that LPD is just as important for the general population as it is for millennials in influencing purchase intentions. However, while LPD appears to be very important in the determination of pricing points for wines in markets, there are very few research studies specifically investigating the role of design attributes in influencing the purchase intentions of customers. Most sources, including Stolle (2019) and Boncinelli et al. (2018), only mention the role of information presented on labels and their impact on different aspects of consumption, including taste expectations. The link between specific design elements and purchase cannot, therefore, be well understood exclusively from these sources.
Unlike Stolle (2019) and Boncinelli et al. (2018), Arnold (2012) presented a discussion of how packaging design trends in the UK are influencing the contemporary consumer. In the evaluation of packaging formats, for example, Arnold (2012) revealed that the Burgundy and the Bordeaux are the two most common and best-selling bottle designs in the UK. In comparison to other packaging formats such as aluminium bottles and cans and the bag-in-box format, there is a clear preference of the two bottle designs meaning that the country’s consumers are very particular about the kind of bottles in which wines are packed. While Arnold (2012) does not make any mention of brand visibility of wines packed in different bottle designs, the findings by Henley et al. (2011) and Elliot and Barth (2012) indicate that wines packed in the Burgundy and the Bordeaux bottle designs achieve better visibility in the UK market. It is as a result of this visibility that Arnold (2012) revealed that these two designs are the most purchased by the consumers in the country. In addition to the bottle design, there is evidence that the caps used to close the bottles also influence consumers towards specific wines over others. According to Arnold (2012), around 70% of wine consumers show positive attitudes towards screw caps despite the fact that natural corks are still perceived as an important element of premium wines. From these observations, a new wine brand would be able to achieve a considerable level of sales in the UK if the correct bottle design and bottle closure are used to target consumers in a specific segment.
An evaluation of label design presents interesting results that would be of profound importance to any wine manufacturer. At face value, it would be plausible to assume that front and back labels on wine bottles play more or less the same role in attracting consumers. In their review of the literature, Kelley et al. (2015) found research studies conducted in Australia and in the UK showing that the attractiveness of the front label is not as influential in the purchase decisions of consumers as the back label. Most front labels contain information such as brand names and origin of the wine, which Kelley et al. (2015) point out as playing a secondary role in the purchase intentions of the contemporary consumer. Back labels, on the other hand, contain detailed product descriptions, including product history, taste, and potential food pairings (Kelley et al., 2015). Contemporary consumers in Europe and Australia show a very strong interest in such information rather than brand image. In corroboration of this finding, Arnold (2012), in his description of important cues in the UK, revealed that most wine producers in the market put important information on the back label in appreciation of the fact that modern consumers show more interest to back labels rather than front labels. In their specific research study in the US market, Kelley et al. (2015) found that information on food-wine-pairing at the back of wine labels was significantly correlated to purchase intentions. This is attributed to the fact that healthy living trend has made most people in all regions of the world be very particular about their feeding habits. The inclusion of health information and potential food pairings in wine labels should improve purchase intentions among consumers, as found by Kelley et al. (2015).
In the specific context of the Chinese market, particularly Hong Kong, Tang and Tchetchik (2015) found that label designs play a very crucial role in wine purchases, especially among young consumers. The findings are found to be consistent with the ones presented by Henley et al. (2011) and Elliot and Barth (2012) with regard to millennials in other countries, meaning that consumer perceptions towards label designs in Hong Kong are comparable to perceptions by consumers in other markets. In the wider context of the entire Chinese market, Celhay et al. (2020) in their research study found that almost all brands imported into the country from Europe and America do not modify their label design to match the Chinese culture. The reason for this is that if modifications were made to the labels to include text in Chinese and designs that are found appealing to the general Chinese consumer, then the wines would lose their authenticity as exotic brands. By maintaining their original designs, therefore, Celhay et al. (2020) found that imported wines are able to be viewed favourably by consumers interested in brands from other countries. Through the label design, therefore, findings by Celhay et al. (2020) and Tang and Tchetchik (2015) indicate that LPD has a profound impact on purchase intentions in China and Hong Kong, respectively. Research findings by Tang and Cohen (2014) provided strong support for the important role played by label designs in the purchase intentions of Hong Kong Chinese consumers. Wine consumers in this market show a strong preference for wine packed in bottles with labels with ‘traditional with chateau’ claims. The lowest preference was recorded for ‘modern contemporary’ label designs. The consumers also showed a preference for yellow labels. From these results, it is evident that Hong Kong Chinese consumers are influenced by LPD in their wine purchases similar to consumers in the other countries researched by Arnold (2012) and Kelley et al. (2015). For that reason, the literature suggests that consideration of LPD elements such as colour, product claims, and wine-food-pairings would be as important in influencing purchase intentions in the Hong Kong market, similar to European and US markets.
2.6 Pricing
Generally, the law of demand asserts that for any ordinary commodity, demand would increase with a decline in price. This means that there would be an increase in purchases with each time the price of a commodity is lowered. While this might be plausible for most items sold in the market, it is not always the case, especially for most luxury products. In corroboration of this argument, Uzgoren and Guney (2012) pointed out that products that are associated with high-class consumers usually gain more demand as their prices increase up to a certain level. The reason for this is that the higher the price of the commodity, the more exclusive it becomes, thus giving the few people who can afford it some level of distinction from the rest of the population. Individuals wanting to be associated with the particular class, therefore, purchase such expensive items so as to boost their image. In other situations, price increments are interpreted as improvements in quality, thus leading to increased demands for specific products. Wine is one such product, as revealed by Lee (2012). According to Huang et al. (2017), wine is one of the products that have highly dispersed prices. There are wines costing as little as $10, while others cost well over $50 for the same quantity (Huang et al., 2017). Lee (2012) insisted that low-priced wines are seen as being of low quality compared to highly priced wines and also went ahead to show that increasing the price of cheap wines has in the past been associated with improvement in quality thus leading to increased demand. There is, therefore, a correlation between the price of wine and the purchase intentions of customers. A keener investigation of research studies evaluating the relationship should provide more insights into the pricing strategies that can be employed in Hong Kong.
Wine has been linked to social status by different researchers and authors. Ritchie et al. (2010), for example, in his literature review, found sources arguing that cheap wines are regarded as low class, thus discouraging consumption by a specific class of people. A similar link between social class and the price of wine is presented by Lee (2012). This, according to Ritchie et al. (2010), is significant as far as wine purchase is concerned, since historically, wine is an elitist product that was a preserve for the affluent. In the present-day, despite the fact that wine has become one of the fast-moving consumer goods, most of its purchases are still from elite groups of people striving to appear as elite in society. For that reason, more expensive wine brands are much more attractive to consumers than cheaper ones. This argument by Ritchie (2010) is found to be highly consistent with that presented by Lee (2012), insisting that purchase intentions towards wine increases with an increase in price. However, while Ritchie et al. (2010) linked his argument to social status Lee (2012) argued that the perceived quality of highly priced wines is the main reason for its high preference among consumers. It is, therefore, imperative at this point to look at how the prices of wine influence purchase decisions as a result of perceived quality and perceived status.
In their research study, on the perceived value of wine and the associated purchase intentions, Bizjak et al. (2020) revealed that perceived value and perceived quality of a wine are interrelated factors that are assessed together by consumers while making the decision to buy wine at any given price. This implies that low prices for a specific wine brand would indicate to a consumer that the wine is probably low quality and would, therefore, give them the status that they want. Such a consumer would, therefore, be more likely to purchase more expensive wines as a result of the perceived quality and the associated perceived status afforded by such brands. The argument by Bizjak et al. (2020), therefore, supports the assertions by both Ritchie et al. (2010) and Lee (2012). Despite this corroboration of different sources, it is actual research evidence that will help in reaching a conclusive verdict as to the actual relationship between the price of wine and the purchase intentions of consumers in the market.
Just like any other product, before a consumer makes the decision to purchase wine, they assess the potential value through external attributes, as has been discussed in the previous sections. Such assessments give consumers an idea of the value they get from the different brands in the market (Bisak et al., 2020). Outrevile and Le Fur (2017), in their research study investigating the price determinants of wine, found that in the assessment of these other external cues, consumers usually try to infer the quality which helps in the determination of a reasonable price that should be paid. The findings reveal that consumers would be willing to pay more for wines whose external characteristics are associated with quality and class. An interesting fact that also emerges from the research is that when there is no explicit information that can help consumers determine the quality of wine, then the price is used as the measure. Overall, therefore, the findings presented by Outrevile and Le Fur (2017) indicate that wine consumers are generally attracted to reasonably high prices regardless of whether or not they can judge the quality from external characteristics.
Price reductions as an approach to stimulating sales would be a poor strategy. This is because the association of price with quality would make any discount seem like a reduction in quality, thus leading to a decline in sales. This argument is strongly supported by findings by Perovic (2014), which shows that wine retailers in Russia give quantity discounts instead of price discounts as a means of avoiding losing sales. In the quantity discount approach, retailers in Russia usually give one bottle of wine free for every nine bottles specific consumer purchases. This approach, as reported by Perovic (2013) has been found to enable retailers to increase wine purchases while avoiding the decline in sales that would result if a 10% price discount was given. From this evaluation, therefore, Outrevile and Le Fur (2017) is found to present research evidence on the behavior of consumers in relation to the price of wine while Perovic (2013) confirms these findings and goes ahead to give a practical approach on how to stimulate wine purchase while keeping prices constant.
The sources evaluated in this section all agree that the perceived link between price and wine quality and class makes it impossible to employ a low-price strategy as a means of stimulating purchase. Apart from Perovic (2013) who presents research findings specific to the Russian market, the findings in the other sources are general and so it is not clear whether consumers in all markets respond to wine prices as found by Outrevile and Le Fur (2017) or there are some differences that must be put into considerations especially in markets like China and Hong Kong specifically. Muhammad et al. (2013) conducted perhaps the most detailed research study on wine importation trends in China. Their findings show that from the year 2000 to 2011, French wine was the most common foreign wine in China, accounting for an average of 44.12% of all foreign wine sold in the country. A significant highlight of the finding is that French wine is the most expensive wine in the Chinese market retailing at an average of $5.99 per litre in 2011. In the same year wine from Spain was the cheapest retailing at an average price of $3.27 per litre. For ordinary goods, it would be reasonable to assume that the cheaper product would have a very high demand, which translates to a very high level of sales. However, this is not the case for wine in China, according to the results by Muhammad et al. (2013). The presented data shows that the market share of the much more expensive French wine increased from 36.28% in the year 2000 to 55.39% in 2011. The market share of Spanish wine was only 4.86% in 2011. These findings are found to indicate the high demand for highly-priced wines, thus confirming the findings by Perovic (2013), Outrevile and Le Fur (2017) and Bisak et al. (2020). The Chinese market can, therefore, be said to behave the same way as predicted by the general results presented in research studies that do not focus on any specific market.
There are numerous research studies investigating the relationship between the price of wine and the purchase intentions of consumers. Many of these research studies find that the price of wine is closely linked to the perceived quality and the status associated with specific wine brands (Ritchie et al., 2010; Outrevile and Le Fur, 2017; Bisak et al., 2020). For that reason, it would not be advisable to use price discounts as a means of stimulating purchase (Perovic, 2013). In the specific context of China, these findings have been found to be highly relevant as a result of the high consistency they show with the statistical data presented by Muhammad et al. (2013) in relation to wine imports in the country. This section has, therefore, successfully revealed the kind of relationship that exists between wine prices and the purchase intentions of consumers. The specific focus on the Hong Kong market will, however, provide more specific insights on how this relationship is similar to or different from that observed in the general Chinese market as per the findings of Muhammad et al. (2013).
Aside from the above 5 consumers’ attributes, “Taste” was also an initial factor for research, but after taking into consideration of the fact that the Hong Kong wine selling environment is largely supermarket-based, the factor of “Taste” was dropped as most regular consumers are not given the chance to taste the wine before making a purchase. This is very different from other countries such as Europe, the US and Australia etc., which have many individual wine stores which allow consumer to taste before purchase.
CHAPTER 3: Research Gap, Design and Methodology
3.1 Conceptualisation
Consumers are encountered with some challenges when presented with different brands of wine. In normal contexts, averagely educated wine purchasers are mostly overwhelmed by variety of wine options on the shelves of various supermarkets and retail stores. In a similar version, the wine sector is increasingly changing the market because of the constantly changing consumer tastes and preferences and the need for health and safe products. In the existing studies, most of the available sources addressing the factors affecting wine purchase mostly appeared to focus on each of the listed factors in isolation. This study aims to combine the identified cues and investigate if there are any correlations among the cues and identify the most and least influential factors that affect the purchase behaviour of non-wine educated consumers. A clear rank of the external cues from the most influential to the least influential in Hong Kong will help in making practical recommendations on how wine producers can boost their sales in the market. The identification of a clear preference of some cues over others will also stimulate research in other markets so as to identify the combination of external cues that wine producers should make more investments in. In that regard, the research questions that this study seeks to answer are:
The primary objective of this research is to analyse consumer perceptions when buying wine based on brand recognition, country of origin, critic scores and wine ratings, label/packaging design, and pricing.
Specific study objectives include:
3.1.2 Research Philosophy
This research will be based on the expectation confirmation theory (ECT) largely because it is considered as the most appropriate theoretical framework for studying post-buying behaviour, service marketing, and consumer satisfaction by many scholars. The expectation confirmation theory was coined by Oliver (1980), who stated that the buying intentions of consumers are mainly connected to satisfaction, which is an attributed that was acquired via expectation and perceived performance of products. In the existing studies on consumer behaviour, expectations are conceptualized as beliefs or predictions regarding products or brands that have identified product features.
The adaption level theory support ECT, as it states that people perceive stimuli relative to or as a derivation from an adapted level of baseline stimulus level (Diddi, 2014). The physiological characteristics of the people affected by the stimulus as well as the setting of the adapted level. Acceptance expectation perform an essential role in the process of purchase decision-making processes. Mainly, the expectations rely on the data collected from external forces, such as friends or mass media. In the proposed research, these sources of information are anticipated to influence consumer perceptions towards country of origin, label design, and brand recognition. Additional factors such as winery marketing strategies and critic’s grades play a basic role in patronizing business and informing the decisions of consumers (Diddi, 2014). Buyers of wine are affected to a variety of data about brands and label designs that influence their buying alternatives. Marketing has significantly assisted wineries to form a positive product image. The rapid technology developments have made it hard for businesses to control the information that reaches customers. Thus, the buying decision processes among consumers are influenced by various expectations. Figure 1 demonstrated various factors that influence the wine purchasing decision.
Figure 1 Factors influencing wine purchasing decisions
As can be seen from figure 1, there are several factors that affect the wine purchasing decision among non-educated wine consumers. These include demographic factors such as gender, age, and education, brand recognition, critic’s grade, country of origin, label/packaging design, and pricing. This study takes a keen interest in brand recognition, pricing, and country of origin, critic’s grade, and label/packaging design.
3.1.3 Research Design
Five experiments will be conducted for Brand Recognition, Country of Origin, Critic’s Grading, Label/Packaging Design and Pricing. These five factors are the independent variables. Brand perceptions and attitudes are the dependent variables.
3.2 Research Methodology
3.2.1 Preamble
This is an experimental research that aims to test and examine the purchase consumer perceptions among non-wine educated consumers. A mixed research method will be adopted. The samples will utilize the independent and dependent study variables to test the varied elements of relationship among the variables. The respondents of the experiment will be non-wine educated buyers and consumers regularly. This research intends to collect data from 20-30 respondents for each experimental condition.
Data collection is conducted during a twelve-week period via experiments setting with systematic approach of randomly selecting 20-30 persons for each experimental condition, who did not have any wine knowledge education, across Hong Kong. The sample population will strictly comprise of wine buyers of that have attained the legal drinking age (18+ years) purchasing wine in Hong Kong. The selection of locations will to collect data from will be based on random selection of districts on the basis of the average mean of income across various districts in Hong Kong.
The collected data will be analysed with the SPSS software. Descriptive statistics will be used to provide a summary of the buying preferences of the study participants. This will be followed by multiple t-tests and regressions to compare data from participants of different demographic profiles.
3.2.2 Research Design
The study utilizes both quantitative and qualitative data. The underlying research questions are derived from the existing studies and improved with insights obtained from qualitative data collection and analysis and confirmed via developed experimental method, which is known as conjoint analysis.
The research consists five phases: phase 1 (qualitative research), phase 2 (quantitative study-pilot), phase 3 (quantitative study-experiment), phase 4 (quantitative conjoint study), and phase 5 (phase 4 and 5 result comparison). Phase 1 presents the qualitative component, which will be comprised of two focus group interviews. The next subsequent sections highlight the study methodology adopted to carry out a qualitative research, which includes sampling and data collection and analysis tools. The rest of the phases adopt quantitative methods, and detailed information is provided in the quantitative methodology section.
Phase 2 is comprised of the pilot study in form of self-administered survey, where participants are asked to provide a rating and descriptions of product profiles through manipulation of intrinsic cues, mostly in dispute with the intrinsic cues achieved through manipulation of prices and country of origin. In addition, subjective, objective, and self-confidence will be measured in the pilot study. It is noteworthy that the main objective of the pilot survey is to ensure validation of measurement tools, chosen products, product cue types, and the level of attributes. The main reason for conducting the pilot study is to identify and remedy any underlying weakness or errors before carrying out stage 3 due to its experimental nature and the related limitation of resources.
The third phase is comprised of the sensory experiments, in which participants will taste and assess manipulated product samples. Analysis of data will determine the capacity of extrinsic cues and determine the extent to which self-confidence and knowledge can moderate perceptions of price, country of origin, brand recognition, critic scores and wine ratings, and label/packaging design.
Following the completion of the pilot study, taste test experiment, and the improvement of the questionnaire, stage 4 is comprised of a survey that measure the described intrinsic and extrinsic cues. A description of intrinsic cues will be provided rather than experienced, which determines product quality on the basis of expectations instead of sensory experience.
Lastly, findings on expected product quality against sensory experiment will be made. Quantification of any considerable variations between expected quality and real sensory experience for all tested for all cues to provide insights that can be used in the conjoint analysis to offer a prediction of consumer views.
3.2.3 Qualitative Method-Phase 1
Focus groups will be adopted to find out whether customers think that extrinsic cues are critical and have influence on the purchase of selected wines. It is important to confirm the countries that would be negatively or positively associated with the selected wines given that country of origin effect been established to be country, product, and market specific (Young, 2019). Group interviews will allow interaction among respondents enabling quick gaining of valuable understanding into customer views about specific topics of interest (Tadajewski, 2016). In addition, focus groups have been chosen because they are vital in triangulating data from other sources and might also disclose new and unanticipated results for further research (Young, 2019). A significant literature exists, showing countries that have a higher likelihood to generate relevant images or associations with wine (Tadajewski, 2016); nonetheless, a presumption could not be made on this. As mentioned, effects of a country of origin are often product, market, and product specific. Thus, it is crucial to identify regions (nations) that would create different quality expectations to assess wine. Moreover, it is essential to validate vital feasible aspects of using wine in this study.
3.2.3.1 Sampling Approach
Judgement sampling will be used recruit participants to participate in four focus groups of 4-5 respondents. While such a small sample obtained through judgement sampling has some limitations, especially the opinions of a few respondents might not be applicable, the use of this sampling strategy is suitable in qualitative research (Yadav et al., 2019). Before final selection of potential participants, all members in focus groups will be screened to make sure they bought or consumed wine at least once in every two weeks. The proposed demographic information of the two focus groups is presented in table 3.1.
Table 3.1 Proposed focus groups demographics
| Demographics | Group 1 | Group 2 |
| Gender | 3 males; 2 females | 2 males; 3 females |
| Age range | 25-40 years | 30-50 years |
From table 3.1, it is clear the study intends to include engage respondents, both male and female from all legal drinking ages. This will ensure that the study findings can be generalized, as people from all ages will be represented in the study. In addition, though not indicated in table 1, the researcher intends to request the participants to provide details of their education status. This will be necessary to compare the level of participant’s education relative to their knowledge about wines.
3.2.3.2 Data Collection
Participants will be video-taped and their answers to questions will be recorded. A purchase scenario will be given to members of the focus groups to induce their thoughts through product characteristics and their significance in wine purchasing decisions. The participants will be asked to consider buying wine for a special person as a gift or have a high-class evening party. Participants will be informed that there will be no wine in the dinner party and they must buy before leaving the wine shop. In addition, group member will be asked to list product characteristics that influenced their buying decisions and score accordingly based on the overall significance when making the purchase decision. A rating scale of 10 (very important) to 0 (on important at all) will be provided. Moreover, respondents will be required to identify countries they think would manufacture low, average, and high-quality wine. It is essential to note that no suggestions will be given in regard to possible countries and product attributes.
3.2.3.3 Qualitative Data Analysis
Assessment and examination of information obtained from the interviews will support the effect of country of origin, price, brand Recognition, and critic’s grading, label/packaging design on expectation for the product recommended for testing. While the views from the focus groups represent the opinions of comparatively few people, the outcome is expected to be consistent with the previous research and support the conceptual framework. Terry et (2017) affirm that thematic analysis method has been adopted for qualitative data analysis and is preferred due to its ability to examine data to identify common themes from the collected data. The researcher will have to make sure that the identified themes are meaningful and provide accurate representations of the collected qualitative data. Following identification of themes, the researcher will need to define and name each one of theme. Defining a theme involves formulation of the exact meaning of each theme and finding out how it assists the researcher to comprehend data (Nowell et al., 2017). Naming theme includes formulation of easily understandable and succinct name of each theme. Overall, each of the identified theme will be discussed in the results and discussion chapter and will show how the analysis answered the research questions.
This research will ensure trustworthiness by establishing credibility and reliability. The main aim of validity is ensuring the credibility and trustworthiness of the study (Terry et al., 2017). Thematic analysis adopted for data analysis will address validity through member checking, which is a technique to assist in improving the applicability, internal validity, as well as transferability of a study (Nowell et al., 2017). Member checking is selected for this study due to its ability to verify the interpretation of research findings, which ensures the validity and reliability of the research method and research design used. Establishing reliability involves finding dependable outcomes that can be applied to different settings (Terry et al., 2017). To ensure reliability, the sample population will be comprised of respondents that have the most understanding and knowledge about wines. Reliability of data is critical as ensures consistency and applicability of the research findings to various settings (Nowell et al., 2017). In addition, member checking improves data reliability.
3.2.4 Quantitative Methodology
The knowledge gained from the qualitative methodology will be useful in developing the conjoint analysis design and data collection instruments. This section provides a description of the measure that will be used to quantify the knowledge and self-confidence of consumers and a brief description of the data analysis techniques.
3.2.4.1 Design
The research utilises a correlational survey research design due to its ability to select sample population and administer of questionnaires to gather the required information (Goertzen, 2017). In addition, this design was selected for this research to expose the study participants to various questions to enhance a fair comparison of the collected data. Regression and correlational designs were also considered for this research. Regression analysis facilitates determination of whether a single study variable predicts other study variables (Mellinger and Hasnon, 2016). On the other hand, correlation analysis helps researchers to determine the strength of the relationship among the study variables (Goertzen, 2017). Since one of the aims of this research is to determine the link between extrinsic cues and the purchasing behaviour non-wine educated participants, a correlation analysis is the most suitable.
3.2.4.2 Sampling
Non-probability, convenience sampling method has been adopted for this research. Samples will be comprised of wine buyers in Hong Kong. The selection of locations across Hong Kong will be random. It is known that samples from the general public can be contribute to limitations as they might not represent the entire population due to restricted nature of profiles based on demographic data. Due to the negative demographics typical of the general population, the general public is more likely to represent general population. To participate in this research, participants must be at least 18 years.
3.2.4.3 Data Collection Instrument
The surveys in this study will be designed on the basis of research questions and the previous studies reviewed in this research. The demographic data of the respondents such as gender and age are required for purposes of comparing the sample population with the general population in Hong Kong. The outcome of data analysis will be used to make necessary enhancements in the next survey versions.
The survey will utilize conjoint analysis-full profile, which is one of the common conjoint analysis and is beneficial because each profile is individually examined, which allows participants to concentrate on their profiles only. Nonetheless, the related risk increases participant fatigue when the number of the profiles to be examined is high. By use of this approach, respondents will be required to review a range of ten wine profiles that will be assessed on a metric scale. This puts the focus of participants on the levels of acceptability feature as opposed to comparison of characteristics among optional offers, with the typical result being rating decision dominance by a few attributes. Nevertheless, this will not be viewed as a limitation to this research since the design will be restricted to a total of three attributes. Moreover, the study objectives do not include identification of product profile, which is viewed the more desirable by participants. Instead, the objective is to examine the influence of the extrinsic cues on product quality expectations and determine the impact on wine quality via taste tests. Thus, a survey design, which makes use of a complete profile conjoint approach is considered the most suitable approach for this study.
While about fifteen product profiles are feasible when participants are examining the product profiles through description, the use of multiple wine products in a single tasting experiment might be monotonous for respondents for the tasting experiment of the study. It is challenging for the respondents to remain vulnerable to the experiment variations in the sampled wines because of the likely desensitized taste buds and fatigue because of the extended time required for the tasting tasks (Jaipakdee et al., 2016). To address this issue, an orthogonal fractional factorial design will be adopted to reduce the number of overall profiles while making sure that each level of attribute is sufficiently represented and controlled to approximate a parameter of major impact of each level and product characteristic (Miah, 2016). The use of an SPSS fractional factorial design will make it possible to taste the worth contribution of all attribute levels (Berger et al., 2018).
It is essential noting that profiles are not examined simultaneously. Therefore, respondents might not receive immediate experience of likely product attributes and their differences to the extent they taste multiple product profiles. To address this limitation, two practice product profiles are proposed for participants to have a sense of the number of attributes and likely levels. This will be essential as it would stabilize ratings and improve accuracy (Berger et al., 2018). However, this idea is countered by Durakovic (2017) who argue that categories of products are common and frequent purchases are made, the experience and market knowledge by consumers are likely to offer sufficient attribute referencing. Nevertheless, these researchers caution that when profiles of products are only described without realistic pictorial representation, training of respondents becomes necessary where familiarization of the product profiles is suggested. Thus, two practice product profiles will be included to train participants to improve the validity of the results. While the two practice product profiles will not be used to determine the attribute of part worth, they will be examined to test validity of the experiment. In addition, the part worth results for characteristics in the practice product profiles will be compared to the part worth results from the product profiles in the fractional factorial design for the purposes of checking consistency of respective attribute and influence level.
3.3 Ethical Considerations
This study will be carried out following approval from all relevant bodies. The researcher will inform potential participants about the research and what is expected of them, and their questions will be answered to satisfaction. After this, the respondents will be needed to make an informed choice on whether to participate or not in this study before completing the survey. In addition, the researcher will enlighten the respondents about their participation and withdrawal rights. The researcher will ensure the confidentiality of the responses and the respondents throughout the study by keeping their identity anonymous.
3.4 Data Analysis
The collected data must meet the validity and reliability criteria for effective analysis. Validity is the level to which a research tool determines the expected research outcome (Mellinger and Hasnon, 2016), and the researcher intends to conduct a pilot survey in which ten questionnaires will be administered to validity and improve the data collection tool by making necessary adjustments. Reliability is the degree to which a research instrument produces same results when research is conducted under identical parameters (Goertzen, 2017), and this will be ensured via enhanced data collection research tool, the researcher will keep cleansing the collected data regularly, and lastly the researcher will ensure a consistent data organization throughout the entire research period.
Following data collection, data analysis will be done through descriptive statistics. The collected data in this study was analysed by SPSS software to generate data array to be utilised in statistical data analysis (Mellinger and Hanson, 2016). Data will be coded to group responses into various categories. It is essential noting that the survey data will be analysed in the form of a 5-point Likert-scale.
3.4.1 Measures used for Consumer Behaviour
3.4.1.1 Subjective Knowledge
The perceived expertise level or knowledge of product level by consumers is the subjective knowledge (Lewis et al., 2019). In majority of the existing research, subjective knowledge is measured by a single report item, ad hoc multi item scales, and semantic differential scales specifically developed for the pertinent research (Niimi et al., 2017). The result of using subjective knowledge covers a broad range of strategies with few methods verified via experiment or use in different research (Li et al., 2019; Priilaid et al., 2019; Canziani et al., 2016). This research will adopt an eight scale item, and this has been used and validated by Flynn and Goldsmith (1999) and Goldsmith et al. (1998) via tasting experiments across eight distinct product categories. The six items that have been recommended for this research are provided in table 3.2.
Table 3.2 Scale Items for Subjective Knowledge
Format: 1-Strongly Disagree; 5=strongly Agree
| 1. | I understand how to judge the wine quality |
| 2. | I have adequate information about different wines |
| 3. | I feel that I am not knowledgeable about wine (r). |
| 4. | I am less knowledgeable of wine compared to my peers (r) |
| 5. | I have heard of majority of the available wines. |
| 6. | I am able to determine whether or not a wine is worth its price. |
The scale of items for subjective knowledge presented in table 3.2 will play a very crucial role to play because high-quality research rests on understanding the study participants. This will enable the researcher to make informed observations during data analysis and presentation.
3.4.1.2 Objective knowledge
Definitions of types of objective knowledge and measurement tools, there has been inconsistencies, which makes comparisons between studies risky and challenging. Objective knowledge is defined as the information of a product kept in long-term memory on the basis of cognitive knowledge and experience, with customers keeping high objective knowledge levels by finding current information about the product. Genuine expertise is an integration of high objective knowledge and product familiarity levels by consumers. It is not feasible to anticipate any substantial consumer market segment to possess expertise levels consistent with professionals working in fields aligned or classified with a specific product. For instance, one cannot expect consumers of wine to have the same knowledge as vineyard owners or wine makers. Thus, it is essential to come up with objective knowledge questions that non-wine educated consumers are expected to be knowledgeable about. Relevant literature will be examined, and product experts consulted to design a set of questions for wine to measure participant objective product knowledge. The four objective items that have been recommended for this research are provided in table 3.3.
Table 3.3 Scale Items for Objective Knowledge
Format: 1-Strongly Disagree; 5=strongly Agree
| 1. | Country of origin of wine is an essential when considering the kind of wines to buy. |
| 2. | Wine brands and branded influence my purchasing decision making process |
| 3. | Price is the most important factor to consider when buying wine. |
| 4. | I must check the label and the packaging design before I buy wine. |
The scale of items for objective knowledge presented in table 3.3 is important to the study because it improves and facilitates the discovery of the most appropriate explanation to a scientific research problem identified in the first chapter of the proposed study. Objectivity will ensure that that evidence and not personal bias or desires is used to answer the identified scientific research questions.
3.4.1.3 Self-confidence
High confidence levels are considered to empower buyers to acts on their individual beliefs, in spite of their accuracy or basis (Apuke, 2017). Conventional self-confidence measurements among consumers are based on tools assessing personal confidence; research implies a connection between confidence and confidence in their perceptions (Mellinger and Hanson, 2016). Perceived locus of control, previous experiences, and dominance are other vital factors contributing to self-confidence (Apuke, 2017). Table 3.4 presents the questions that will be asked to determine the personal self-confidence levels of non-wine educated consumers.
Table 3.4 Personal self-confidence questions
Answer format: 5=strongly Agree; 1-Strongly Disagree
| 1. | Normally, I think my opinions are inferior. (r) |
| 2. | I do not concentrate so much on what people think about me. |
| 3. | I seldom fear actions that would make others have a low opinion of me |
| 4. | When introduced to a stranger, I am never at loss for words. |
| 5. | My first reaction is always inferiority and shyness when confronted by strangers. (r) |
| 6. | I do not have a good first impression on people. (r) |
The personal self-confidence questions presented in table 3.4 will allow the researcher to understand the self-confidence levels of participants, which is crucial for the researcher to experience freedom from negative making negative observation about their conduct during the entire research period. For example, greater self-confidence levels among participants would enable the researcher to collect adequate and rich data concerning the research phenomenon under investigation.
The aim of this research to obtain subjective and objective knowledge, and personal self-esteem of non-wine educated consumers and measure all of them independently. The researcher will investigate the nature and the existence correlations between the knowledge of consumers and self-esteem and the levels of utility measured for all product attribute that will be determined from conjoint analysis. Besides, during assessment of the relationships, any variation in self-confidence and knowledge will be examined among groups to identify common response patterns.
3.4.1.4 Survey Data Analysis
The study adopts factor analysis and correlation to analyse the collected data. Correlation is used to determine the connection between at least two study variables, with coefficients ranging between -1.00 and -1.00, where -1.00 represents a perfect negative relationship and +1.00 a perfect positive relationship (Mellinger and Hanson, 2016). When the analysis of two study variables find a significant relationship, with a coefficient close to +1.00 or -1.00, there is a strong linear correlation between the variables (Goertzen, 2017). In the current study, correlations will be used to test the subjective and objective knowledge and self-confidence and brand recognition, country of Origin, critic’s grading, label/packaging design and pricing. The principle factor analysis will be used for testing and proving latent variables are determined from multiple items grouped together (Ghauri et al., 2020), and this analysis will be applied for the scales used in different stages for purposes of examining and measuring self-esteem and subjective knowledge. Due to change of validity checks and reverse coded items, the mean for each participant that reflects their self confidence levels and subjective knowledge will be calculated. Standardized score for objective knowledge will be computed to compare objective and subjective knowledge levels.
The examination of the preferences of wine and the expected product levels from the conjoint analysis will be compared with SPSS. Data analysis expects to reveal features of a product that will have the most impact on the assigned score. Exclusive analysis will be done at all phases of the study, starting with the overall sample and finishing with segments in relation to levels of knowledge and levels of self-esteem. This analysis will enable the testing and comparison of tasting scores to answer the research questions.
3.5 Attribute Importance Determination
Conjoint analysis seeks to create a set of additive utilities that make use of scores assigned to product profiles for purposes of deriving attributive utility scores (Goertzer, 2017). Fundamentally, these are index numbers that correspond to coefficients of regression that measure how desirable or valuable a specific cue is to the participant (Pattern and Newhart, 2017). In addition, the Ordinary Least Regression (OLS) approach to ratings on the basis of conjoint analysis, as it provides a simple yet strong approach that can be used to derive different utility values used to in the computation of attribute part worth for all components.
Part worth utility values will be both negative and positive, and expressed on a common scale that sums to zero for all cues; while part worth statistics within a cue might be compared, they might not be comparable across all cues (Pattern and Newhart, 2017). Thus, the most appropriate approach is to interpret the part worths to examine the gaps among levels of utility levels within cues (Ghauri et al., 2010). A large gap between utility levels within cues will suggest that that respondents think that change within that specific cues have considerable effect on their overall analysis. Therefore, cues with higher ranges are the one used frequently by participants to distinguish between product profiles and have greater levels of comparative significance in their rating (Pattern and Newhart, 2017). Importance scores will be calculated to compare the relative importance of cues, and this will be done through taking into consideration of the least and the highest utility value for a cue and diving it by the total all utilities (Ghauri et al., 2020). Overall, values of average importance disclose the comparative importance of each cue to the decision of the participant and the part worths show the cues that are most and least preferred. Moreover, a score or perceived worth can be calculated to determine the cues that consist the most desirable levels. The comparison means will need non-parametric tests; therefore, this research adopt a non-categorical measure to facilitate comparisons of means both within different cues and investigation between the study variables.
3.6 Validity of Research Instruments
For self-confidence and subjective knowledge scale, Cronbach Alpha coefficient ratings has been adopted to evaluate the reliability of the scale. Cronbach Alpha scores range from 0-1, and the higher the score the higher the reliability of the scale. A score of at least 0.7 is considered acceptable for a research collection instrument. Therefore, if the current research attains a Cronbach alpha coefficient of at least 0.7, it will be considered reliable. For objective knowledge, the questions that would be used to measure objective knowledge of non-wine educated consumers will be compiled. The researcher will seek the help of the experts in the field to ensure that the questions are as objective as possible.
Regarding conjoint analysis, it needs checks for internal and external validity (Ghauri et al., 2020). Internal validation will be done by carrying out a ‘goodness of fit’ of the approximated approach with r2 values and Kendall’s Tau statistics to measure the relationships of ranked data. The value ranging from 0 to 1 demonstrate the relationship between the approximated and the observed preferences and should always tend towards 1 because they are better as the move towards and closer to 1; models with values tending to 0 are considered to be poor fit and are suspect. The utility comparison will offer a further internal check, which is based on the extent to which the sample population represent the population under study and whether or not the chosen cues reflect the credibility of items. Comparison of the sample and the general population will be done and analysed at phase independently. Cues and levels were chosen based on extensive literature review, review of the products available, and quantitative data analysis.
Overall, adequate information about quantitative methodology in terms of sampling techniques and designing of data collection tools has been provided. This includes the rationale behind each selected method, validation procedures, and data collection and analysis. The following section expands the methodology based on the data analysis of the pre-test specific to the phase 2 of the research, which is the pilot research.
3.7 Pilot Study
The pilot study has been carried out to check the suitability of the conjoint analysis methodology, attributes, products, as well as measures of self-confidence and knowledge. Pilot data collection and analysis discloses a few issues that relate to perceptions of respondents regarding the type of wines and cues employed. This has created the need to re-examine some aspects of the survey and instructions to participants. Overall, the pilot study provided an invaluable opportunity in terms of addressing the revealed issues, while still validating the adopted methodology.
A sample of 15 respondents agreed to take part in the survey and 12 respondents agreed to participate in the focus group. The 12 participants in the focus group were divided in four groups of three members each. These respondents agreed to participate on a voluntary basis and completed the survey and the sensory experiment.
The pilot self-administered survey was designed to collected demographic information of the respondents and divided into three main parts: scale of items for subjective knowledge, objective knowledge, and self-confidence. Consideration of the few issues revealed in the pilot survey, the following changes have been proposed, as regards the format of the final survey:
Survey Format
Front Cover: Demographic information
Part One: (a) Different wine products (b) Subjective knowledge questions (c) Wine consumption and buying habits.
Part Two: Objective knowledge questions
Part Three: Self-confidence questions
Regarding the extrinsic cues, different countries were selected to represent the different quality levels of wine. Table 3.5 demonstrates the chosen extrinsic cues.
Table 3.5 Specification of Cues
| Attribute | Type of cue | Level |
| Country of origin | Extrinsic | France
United States Australia |
| Price | Extrinsic | USD 40
USD 15 USD 10 |
Various studies have adopted extensive descriptions to represent highest and lowest rating scores include ‘very desirable’ vs ‘very undesirable’, ‘worst possible’ vs ‘best possible’, ‘like extremely’ vs ‘dislike extremely’. Following a review of these alternatives, the pilot study settled on utilization of ‘Terrible’ vs ‘Excellent’ in a scale of 1-10, because it was observed that the descriptor motivated participants to provide an opinion of the wine to that is clearly positive or negative, without linking this experiment directly to buying intention. After rating each product, the respondents were asked to indicate whether they could purchase the wine based on tasting, with a ‘yes’ or ‘no’ response.
3.7.1 Validation of the research Instruments
A survey pre-test was carried out using 15 respondents to determine the time needed to complete the survey. In addition, the pre-test was as well conducted to provide insight into the likely validity of the scales and objective knowledge tests used to measure independent study variables. Moreover, participant reactions to the profile evaluation activities and an examination of the preliminary data from the conjoint experimental design were possible.
3.7.1.1 Subjective knowledge and self confidence
Whereas the items used to measure the subjective knowledge had been validated in previous researches, the items used to measure self-confidence had not this procedure before. In previous research, the items were simply used as statements in which agreements and disagreements of the items facilitated the determination of whether a participant was considered confident or not. The first stage in the validation process is to determine internal reliability with the help of the Cronbach Alpha coefficient for each measure. Cronbach analysis revealed that the survey is well designed, and this is provided in table 3.6. The coefficients indicate sound reliability of the chosen study variables.
Table 3.6 Reliability test of Scales (pilot)
| Subjective knowledge | No. of items=6 | Alpha=0.866 |
| Self-confidence | No. of items=6 | Alpha=0.728 |
| Sample population (n) = 15 | ||
3.7.1.2 Objective Knowledge Items (pilot)
The feedback from the pilot participants was positive; the comments the researcher received showed that the respondents found the tests simple to complete and interesting. Some respondents were eager to know how well they scored. In accordance with the existing literature, generally, respondents found that their objective knowledge was lower than anticipated. No statement was flagged as unrealistic or unreasonable. The feedback from data collection does not give any signal that participants had not answered truthfully, which shows that no statement was misunderstood or ignored. Therefore, it is apparent that both internal and external validity were not compromised, meaning that the research instrument is suitable for research.
3.7.1.3 Sensory Experiment-Focus Groups
Before the start of the experiment, a strict protocol of briefing and discussion about the experiment took place. The researcher provided background information of the study and general introductions. Participants, in their various groups, were informed about the format of the experiment and the type and quantity of products they would taste. The researcher provided information on when participants were to stop and wait for further guidelines. During the experiment the researcher observed that the atmosphere of the experiment was kept friendly and relaxed due to the small number of participants in the four groups. Therefore, from this observation, a small number of 3-5 individuals will be used in the main experiment.
In general, the feedback obtained during the pilot data collection through the focus group interviews show that the set of questions and experiment test designs will remain unchanged, as the participants expressed no concern about how the experiments were conducted. However, the researcher realized that due to the need for participants to take a significant time to participate in the study, a cash incentive would be appropriate. Therefore, the researcher intends to offer a cash incentive of 20 dollars per participant. The payment will be made once the respondent completes the experiment.
In this study, participants have not been trained or coached to enhance their knowledge of wine. In addition, they have not been tested to determine their natural capabilities to distinguish between product samples. Nonetheless, these potential participant limitations are irrelevant in this research because the researcher is not attempting to quantify the sensitivity levels of consumers. This is the main reason why the study intends to recruit non-wine educated consumers. The main aim of this study is to determine and test the relationships identified between the study variables. The experiment seeks to examine the perspectives of non-wine educated consumers to quantify the respective influence of extrinsic cues on wine purchase, as determined via their sensory evaluation. This explains why product training, sensitivity, and reliability tests have not been included in the methodology of experiments.
3.7.1.4 Sample Demographics
Table 3.7 and 3.8 present the sample demographic profiles of the participants that participated in the pilot study.
Table 3.7 Sample demographic profile-Survey
| Gender | Percentage (%) | Age (Years) | Percentage (%) |
| Male
Female |
40
60 |
18-25
26-35 36-50 51 and above |
6.67
20.00 66.67 6.67 |
Table 3.8 Sample demographic profile-Focus Interview
| Gender | Percentage (%) | Age (Years) | Percentage (%) |
| Male
Female |
50
50 |
18-25
26-35 36-50 51 and above |
8.33
25.00 58.33 8.33 |
From the comparison of the two tables, it is clear that majority of the respondents that participated in the pilot study were between 36 to 50 years. There were no adequate participants between the age brackets 18-25 and 51 and above. This might be attributed to the convenience of the employed methodology used to sample pilot participants, such biases are anticipated and are less severe compared to other types of convenience samples. In addition, considering the products used and the age brackets, demographic skew is likely to give a highly representative sample of wine consumers in the chosen markets. Moreover, it is essential noting that while student sample was used, only students from the evening classes were considered, and because of the diverse demographics typical of evening students, they have a higher likelihood to constitute a group representative of the general population compared to traditional samples.
3.8 Summary
All items used in the survey satisfied tests for external and internal validity, which confirms their usability in subsequent research stages. Items designed to measure knowledge and self-confidence remains unchanged. However, the a few changes were in terms of changing the survey layout and the order of various components to improve the presentation of the survey. From the observation in the focus groups, the researcher realized that the research participants needed reassurance that the assessment of foreign products does not mean that the researcher or the participant prefer them to local products. The researcher will need to assure the respondents that wine from foreign countries is preferable in this study because the respondents are used to the local brands of wine, and this might not negatively affect the outcome of the sensory experiment because of this bias, which might make it difficult to obtain objective responses. This will be indicated in the methodology chapter in the section of sensory experiments (focus groups). The researcher also discovered that a detailed review will be given to participants in the focus groups before their participation in the sensory experiment in order to address any potential hostility and resistance towards foreign products. These changes and improvements will be discussed comprehensively in the final dissertation.
The next subsequent chapter, chapter four: Results and Analysis, will be present the findings of the collected data through the methodology presented in this chapter. The collected data will seek to examine the perceptions of wine purchase among the non-wine educated consumers. To achieve this goal, the research instruments are designed to collect data that can be used to investigate the attitude of consumers towards brand recognition, country of origin, critic scores and wine ratings, label/packaging design, and pricing. With findings of the study, the study aims to establish the effects of highlighted five factors on the buying behaviours of the consumers in Hong Kong and rank the five factors in terms of their importance. Overall, chapter four will analyse, present, and discuss the findings, as obtained from the survey and the sensory experiment. In addition, the chapter will test hypotheses that to prove and determine the extent to which the five factors affect the consumer buying behaviour of wine consumers.
CHAPTER 4 – RESULTS AND ANALYSES
4.1 Introduction and Overview
The primary objective of this research has been to analyse consumer perceptions when buying wine based on brand recognition, country of origin, critic scores and wine ratings, label/packaging design, and pricing. Moreover, the objectives of the study sought to identify consumer perceptions towards wine buying and to determine the linkage between the identified factors and wine buying behaviour of consumers in Hong Kong. Therefore, detailed analysis for each of the data feedback has been presented throughout the study with the focus to examine the perceptions of wine purchase among the non-wine educated consumers. Thus, the current chapter presents results and a review of the methods employed. In the first section, the response rate of the feedback from the participants has been explicated. In the second section, the review of missing values, outlier and normality tests, correlations, and descriptive statistics for data feedback of focus groups has been captured. Then the third section features the results for Part 1-3 feedback and fourth section provides the outlook for the surveys and in both cases data screening has also been implemented. The fifth section presents the test of significance where main relationships have been explored relying on various tests such as General Linear Models, t-tests, Hierarchical Regression Analysis, Mediated Regression Analysis, and Moderated Regression Analysis. The outline sought in the chapter is as shown below under Figure 4.1 below.
Figure 4.1: Structure of Chapter Four
4.2 Response Rates and Distribution of Surveys
4.2.1 Distribution of Surveys
The surveys were coordinated across consumers in Hong Kong to decipher the perceptions of wine purchase among the non-wine educated consumers. The unit of analysis was non-wine consumers comprising of a total of 30 participants. The participants were invited to provide their honest feedback amidst presentation of various wines types to determine the rating on taste among others.
4.2.2 Response Trends
The study’s sample consisted of non-wine enthusiasts targeted through judgment and convenience sampling. Thus, focus groups were administered to 30 participants and contained eight items while the surveys were guided by 7 item constructs coordinated to the sample of participants. In the focus groups, some individual items failed to achieve 100% response rate since some of the answers were missing. However, the surveys and Parts 1-3 had 100% response rate.
4.3 Screening and analysis of data from Focus Group
In this section, the findings are based on the results generated from the focus group. The first attempt is to evaluate the data for reliability using Cronbach’s Alpha especially for the scores generated from the participants in different levels since string variables could not be subjected to the same reliability test as SPSS program could not recognise the function as evidenced in the screenshot below in Figure 4.2.
Figure 4.2: Declined reiability test of string variables
Therefore, the option was to run only scale variables which in this case represented all the scored values details of which have been discussed in due course of the study. Table 4.1 below depicts the Cronbach’s Alpha results.
Table 4.1: Cronbach’s test results I
| Case Processing Summary | |||
| N | % | ||
| Cases | Valid | 30 | 100.0 |
| Excludeda | 0 | .0 | |
| Total | 30 | 100.0 | |
| Reliability Statistics | ||
| Cronbach’s Alphaa | Cronbach’s Alpha Based on Standardised Itemsa | N of Items |
| -.590 | -.483 | 6 |
From the results, it is evident that a negative Cronbach’s Alpha is manifested at -.590 which means there is a presence of a negative average covariance within items generated by the scoring model. As a result, the outcomes show to violate the reliability model assumption. One of the solutions was to re-examine the items representing all scores coding and re-run the reliability test to achieve a better model. Nonetheless, the case summary proves that there are no missing values in the dataset meaning all the cases are valid. To address the problem above, the researcher sought to transpose the score values where a second test of reliability generated the following results, as shown under Table 4.2.
Table 4.2: Cronbach’s test results II
| Reliability Statistics | |
| Cronbach’s Alpha | N of Items |
| .519 | 30 |
The results depict that the negative average covariance is eliminated and the reliability rising to .519 which is same as 51.9%. However, in the future it would be desirable to increase the sample to above 200 to achieve a higher reliability value of above 70%. The researcher evaluated the same variables using Principal Component Matrix as depicted under Table 4.3 below to evaluate the degree of validity.
Table 4.3: Factors loading analysis on Critic Scores
| Communalities | |||||||||
| Initial | Extraction | ||||||||
| Score_Blind | 1.000 | .755 | |||||||
| Score_Not Blind | 1.000 | .741 | |||||||
| Score_Blind | 1.000 | .815 | |||||||
| Score_Not Blind | 1.000 | .751 | |||||||
| Score_Blind | 1.000 | .773 | |||||||
| Score_Not Blind | 1.000 | .464 | |||||||
| Extraction Method: Principal Component Analysis.
|
|||||||||
| Total Variance Explained | |||||||||
| Component | Initial Eigenvalues | Extraction Sums of Squared Loadings | |||||||
| Total | % of Variance | Cumulative % | Total | % of Variance | Cumulative % | ||||
| 1 | 1.791 | 29.857 | 29.857 | 1.791 | 29.857 | 29.857 | |||
| 2 | 1.439 | 23.989 | 53.846 | 1.439 | 23.989 | 53.846 | |||
| 3 | 1.068 | 17.799 | 71.645 | 1.068 | 17.799 | 71.645 | |||
| 4 | .812 | 13.540 | 85.185 | ||||||
| 5 | .510 | 8.495 | 93.680 | ||||||
| 6 | .379 | 6.320 | 100.000 | ||||||
| Extraction Method: Principal Component Analysis.
|
|||||||||
From the establishments above in Table 4.3, it is evident that all the communalities representing the results for each of the scores are above the threshold for 0.4 or 40% meaning all the cases are valid. Therefore, the study’s results confirm that the feedback from the focus group on the scoring facets in relation to the taste of wine. In light of the explained total variance it is evident that there are statistical differences although not possible to determine if it is significant or not-significant.
Further analysis aimed to establish whether the data for the scores mentioned above consisted of a normal distribution, as shown under Table 4.4 below.
Table 4.4: Normality test results for the scores feedback
| Tests of Normality | ||||||
| Kolmogorov-Smirnova | Shapiro-Wilk | |||||
| Statistic | df | Sig. | Statistic | df | Sig. | |
| Score_Blind | .168 | 30 | .031 | .941 | 30 | .098 |
| Score_Not Blind | .120 | 30 | .200* | .947 | 30 | .136 |
| Score_Blind | .233 | 30 | .000 | .892 | 30 | .005 |
| Score_Not Blind | .169 | 30 | .028 | .907 | 30 | .012 |
| Score_Blind | .175 | 30 | .020 | .932 | 30 | .054 |
| Score_Not Blind | .236 | 30 | .000 | .908 | 30 | .013 |
From the outcomes, especially Kolmogorov-Smirnov, it is evident that only once incidence of the scores (Sig. = .200) shows to have a normality distribution while the rest do not give that the p-values are above 5% margin of error. Although the researcher checked the same trend using the transposed values and the p-values were above 5% margin of error for all the cases. In that regard, the problem for a lack of normality was considered weak and that the reported scores could still be modelled for predictive analytics on the influence of extrinsic cues on wine purchase behaviour in Hong Kong.
Using box plot an evaluation of the presence of outliers in the reported scores was captured as shown next. For instance, Figure 4.3 below illustrates the outlier trend for the first score.
Figure 4.3: Outlier test for Score 1_A
The results under Figure 4.3 above indicate that there exist no outliers in the dataset meaning majority of the scored cases followed a uniform pattern.
Figure 4.4: Outlier test for Score 1_B
The results captured under Figure 4.4 above indicate that the 20th score is an outlier verses the rest which means it does not align to the pattern absorbed in other slots. However, despite the presence of the outliers the researcher holds the opinion that the problem is not broad since just one incidence is non-uniform.
Figure 4.5: Score 1_C
As depicted under Figure 4.5 above, it can be seen that the third scoring model has the 9th feedback as an outlier meaning it fails to follow the uniform pattern manifest in the rest of the trends.
Figure 4.6: Score 1_D
The results under Figure 4.6 above are indicative of the fourth scoring as having no outliers in the dataset; hence, confirming that all the patterns are not deviating from a uniform pattern.
Figure 4.7: Score 1_E
Figure 4.7 above presents the fifth level of the scoring from the participants which proves to have no cases for outliers. In this regard, the data patterns can be said to have a uniform process that is not deviating from the mean.
Figure 4.8: Score 1_F
Lastly, Figure 4.8 shows that the scoring feedback under Score 1_E label has numerous cases of outliers compared to the rest of the feedback. In that case, there are outliers that do not align to the uniform performance reported in other patterns; the outliers points are at participant 10, 14, 16, 20, 21, 28, and 30.
Overall, it is evident that the dataset representing the various scores to a better extent meets criteria for reliable and valid data despite incidences that lack a normal distribution.
4.3.1 Descriptive statistics on critic scores derived from Focus Group
In this section, descriptive statistics for the reported scores has been captured representing the mean trend between the scales of 1-10. The results are as shown under Table 4.5 below.
Table 4.5: Summary statistics for Scores Estimates
| Descriptives | ||||
| Statistic | Std. Error | |||
| Score 1_A | Mean | 6.1000 | .47790 | |
| 95% Confidence Interval for Mean | Lower Bound | 5.1226 | ||
| Upper Bound | 7.0774 | |||
| 5% Trimmed Mean | 6.1481 | |||
| Median | 7.0000 | |||
| Variance | 6.852 | |||
| Std. Deviation | 2.61758 | |||
| Minimum | 1.00 | |||
| Maximum | 10.00 | |||
| Range | 9.00 | |||
| Interquartile Range | 4.25 | |||
| Skewness | -.290 | .427 | ||
| Kurtosis | -1.014 | .833 | ||
| Score 1_B | Mean | 6.4333 | .47145 | |
| 95% Confidence Interval for Mean | Lower Bound | 5.4691 | ||
| Upper Bound | 7.3975 | |||
| 5% Trimmed Mean | 6.5741 | |||
| Median | 7.0000 | |||
| Variance | 6.668 | |||
| Std. Deviation | 2.58221 | |||
| Minimum | .00 | |||
| Maximum | 10.00 | |||
| Range | 10.00 | |||
| Interquartile Range | 3.25 | |||
| Skewness | -.594 | .427 | ||
| Kurtosis | .147 | .833 | ||
| Score 1_C | Mean | 7.4333 | .41157 | |
| 95% Confidence Interval for Mean | Lower Bound | 6.5916 | ||
| Upper Bound | 8.2751 | |||
| 5% Trimmed Mean | 7.6111 | |||
| Median | 8.0000 | |||
| Variance | 5.082 | |||
| Std. Deviation | 2.25424 | |||
| Minimum | 1.00 | |||
| Maximum | 10.00 | |||
| Range | 9.00 | |||
| Interquartile Range | 3.00 | |||
| Skewness | -1.087 | .427 | ||
| Kurtosis | .950 | .833 | ||
| Score 1_D | Mean | 7.1000 | .47307 | |
| 95% Confidence Interval for Mean | Lower Bound | 6.1325 | ||
| Upper Bound | 8.0675 | |||
| 5% Trimmed Mean | 7.2407 | |||
| Median | 8.0000 | |||
| Variance | 6.714 | |||
| Std. Deviation | 2.59110 | |||
| Minimum | 1.00 | |||
| Maximum | 10.00 | |||
| Range | 9.00 | |||
| Interquartile Range | 4.25 | |||
| Skewness | -.640 | .427 | ||
| Kurtosis | -.556 | .833 | ||
| Score 1_E | Mean | 6.3667 | .45608 | |
| 95% Confidence Interval for Mean | Lower Bound | 5.4339 | ||
| Upper Bound | 7.2995 | |||
| 5% Trimmed Mean | 6.4630 | |||
| Median | 6.5000 | |||
| Variance | 6.240 | |||
| Std. Deviation | 2.49805 | |||
| Minimum | 1.00 | |||
| Maximum | 10.00 | |||
| Range | 9.00 | |||
| Interquartile Range | 3.25 | |||
| Skewness | -.279 | .427 | ||
| Kurtosis | -.415 | .833 | ||
| Score 1_F | Mean | 5.6333 | .34735 | |
| 95% Confidence Interval for Mean | Lower Bound | 4.9229 | ||
| Upper Bound | 6.3437 | |||
| 5% Trimmed Mean | 5.6667 | |||
| Median | 6.0000 | |||
| Variance | 3.620 | |||
| Std. Deviation | 1.90251 | |||
| Minimum | 1.00 | |||
| Maximum | 10.00 | |||
| Range | 9.00 | |||
| Interquartile Range | 1.25 | |||
| Skewness | -.365 | .427 | ||
| Kurtosis | 1.662 | .833 | ||
The results depicted under table 4.5 above affirm that the mean or average scores reported in each of the cases supersede the standard deviation; for that reason it means that there is minimal deviation from the average scores which seem to be above 6. In the survey, a score of 6 and above which points to an optimal estimation that is close to excellent outcome upon tasting of the wine by the participants.
4.3.2 Correlation test results for Critic Scores
In this part of the analysis the focus is to further establish the existence of a linearity trend for the outcomes of each scores using Pearson Product Moment Correlation as shown under Table 4.6 below. Noteworthy, the critic scores are based on the test outcomes with three samples of wines where the first round was blind tasted and participants requested to give a score from 1-10. In the second round the process relied on the same but letting the participants see the label of wine. There were two sets of scores recorded on three individual wines leading to a total of six outcomes as manifested in the correlation matrix under Table 4.6 below.
Table 4.6: Correlation test results
| Correlations | |||||||
| Score_Blind | Score_Not Blind | Score_Blind | Score_Not Blind | Score_Blind | Score_Not Blind | ||
| Score_Blind | Pearson Correlation | 1 | .248 | -.276 | -.174 | .057 | -.124 |
| Sig. (2-tailed) | .186 | .139 | .357 | .763 | .514 | ||
| N | 30 | 30 | 30 | 30 | 30 | 30 | |
| Score_Not Blind | Pearson Correlation | .248 | 1 | .186 | -.316 | -.042 | -.100 |
| Sig. (2-tailed) | .186 | .326 | .089 | .828 | .599 | ||
| N | 30 | 30 | 30 | 30 | 30 | 30 | |
| Score_Blind | Pearson Correlation | -.276 | .186 | 1 | -.391* | .130 | .215 |
| Sig. (2-tailed) | .139 | .326 | .032 | .493 | .253 | ||
| N | 30 | 30 | 30 | 30 | 30 | 30 | |
| Score_Not Blind | Pearson Correlation | -.174 | -.316 | -.391* | 1 | -.400* | -.048 |
| Sig. (2-tailed) | .357 | .089 | .032 | .028 | .800 | ||
| N | 30 | 30 | 30 | 30 | 30 | 30 | |
| Score_Blind | Pearson Correlation | .057 | -.042 | .130 | -.400* | 1 | .174 |
| Sig. (2-tailed) | .763 | .828 | .493 | .028 | .357 | ||
| N | 30 | 30 | 30 | 30 | 30 | 30 | |
| Score_Not Blind | Pearson Correlation | -.124 | -.100 | .215 | -.048 | .174 | 1 |
| Sig. (2-tailed) | .514 | .599 | .253 | .800 | .357 | ||
| N | 30 | 30 | 30 | 30 | 30 | 30 | |
| *. Correlation is significant at the 0.05 level (2-tailed). | |||||||
As captured in Table 6 above, there are only two cases where there is a significant negative linearity given the Pearson Correlation at -.391** and -.400** where both are above 95% confidence interval. For instance, in the former case, it manifests that there is significant inverse relationship in the scores awarded to the second type of wine while blinded tasted and not blind tasted. The latter case evidences that the critic score for the third type of wine while blind tasted has significant inverse relationship to the scores attributed to the second type while not blind tasted. However, the rest of the scores have p-values that are way above 5% margin of error meaning the correlations are not significant. From these evaluations, the researcher purports that there is not high correlation in the reported critic scores around the taste of wine either blind tasted or not blind tasted as rated by the focus groups in the scale of 1-10.
4.3.3 Demographic characteristics of the Focus Group
In this part, the age and gender of the participants has been reported which is an important evaluation in this study since they might have an influence over the consumer buying behaviour towards wine in Hong Kong. For instance, Figure 4.9 below manifests the trend for gender of the participants.
Figure 4.9: Gender of the participants
The trend shown under Figure 4.9 above shown that majority of the participants at 60% are males while 40% are females. Later, the researcher has shown whether gender has any influence or rather association with consumer buying behaviour of wine in Hong Kong. The same case applies to age of the participants whose trend has been captured under Figure 4.10 below.
Figure 4.10: Age of the participants
As shown in Figure 4.10 above, majority of the focus group participants fell in the age bracket 36-50 years which is 36.7% of the total cohort; while 26-35 years being the second ranking age at 26.7% while 18-25 years covering 20% of the respondents. Lastly, 51+ years is the least age representing 16.7% of the participants.
4.3.4 Type of Wine
The focus group was asked to illustrate the types of wine they like to drink; this was an important feedback since it gauged the preferences of the non-wine educated consumers. Moreover, the feedback around preferences equated the perceptions the cohort has on various types of wines. In more specific terms, the study sought to establish the specific kind of wines the participants usually buy based on four brands namely: white, red, sweet, sparkling, and other where the results are reported in Figure 4.11 below using a bar chart.
Figure 4.11: Preferences on wine among Focus Group
The indication is that red wine seems to be the most preferred by 63.3% of the respondents followed by white label at 16.7%; then sparkling wine at 13.3% and lastly other at 6.7%.
4.3.5 Different wines for different occasions
Moreover, the study sought to examine whether the respondents were inclined to buy different wines for different occasions as reported under Figure 4.12 below.
Figure 4.12: Inclination to buy different wines for different occasions
The feedback in the graphical output above in Figure 4.12 illustrates that a higher frequency i.e. 46.7% of the participants said yes and that they bought wine for different occasions. Along with these, it was sought whether an association exists between inclination to buy wine on different occasions and the preferred type of wine; here, the researcher applied a Chi-Square test whose results are as reported below under Table 4.7.
Table 4.7: Chi-Square test 1
| 3a. Type of Wine * 3b. Different Occasions Crosstabulation | |||||||||
| Count | |||||||||
| 3b. Different Occasions | Total | ||||||||
| De | No | Ye | |||||||
| 3a. Type of Wine | Other | 1 | 0 | 1 | 2 | ||||
| Red | 7 | 5 | 7 | 19 | |||||
| Sparkling | 1 | 1 | 2 | 4 | |||||
| White | 0 | 1 | 4 | 5 | |||||
| Total | 9 | 7 | 14 | 30 | |||||
|
Chi-Square Tests |
|||||||||
| Value | df | Asymptotic Significance (2-sided) | |||||||
| Pearson Chi-Square | 4.262a | 6 | .641 | ||||||
| Likelihood Ratio | 5.982 | 6 | .425 | ||||||
| N of Valid Cases | 30 | ||||||||
As per the results above in Table 4.7, despite the cross-tabulation showing interactions between the two phenomena the Pearson Chi-Square returned a non-significant p-value meaning there is no association. In other words, there exists no relationship in the two categorical variables meaning the preference for the type of wine especially red wine is not influenced in any way by the emerging occasion.
4.3.6 Factors about the kind of wine to drink or buy
The study evaluated the kind of factors the participants consider when thinking about the kind of wine to drink or buy. Figure 4.13 below provides a summary overview of the feedback results.
Figure 4.13: Feedback on factors influencing decisions to buy or drink
The results shown in Figure 4.13 above prove that price is considered a major factor when choosing wine to drink or buy and rated at 43.3%. Taste is ranked at the second place at 26.7% while bottle coming at a distant third at 13.3% and brand has an impression at 10.0%. However, both design and origin are the least factors at only 3.3% respectively. In other words, the latter two can be considered to be the minimal factors influencing decisions to drink or buy wine.
4.3.7 Source of best wines
In the interactions with the focus group the study sought to establish from the participants where they would expect the best wines to come from and the countries that come to their mind. In fact, the phenomenon of the country of origin is linked to the consumer perceptions when buying wine as noted earlier in the study. The first impression is as depicted under Figure 4.14 below.
Figure 4.14: Feedback on country of origin for the best wine
As evident under Figure 4.14 above, France seems to be the major country that comes to mind as the origin of the best wine at 63.3% while Italy ranks second at 23.3%. However, U.S and Spain both at 6.7% and 3.3% respectively come a distant third meaning they are not recognised as origins of best wine and lastly China is the least considered at 3.3%. As a matter of fact, both Spain and China seem to be the lowest in terms of the considerations by the focus group regarding the origin of best wine. Another related assessment on the country that comes to mind in association to the original of best wine is as shown below in Figure 4.15.
Figure 4.15: Related feedback on country of origin for the best wine
In the results under Figure 4.15 above it can be seen that Italy and France are ranked the highest in terms of the country of origin that comes to mind regarding the best wine; from this outcome, it can be inferred that there is cogent fact that the two countries are associated with the development of the best wine. Due to this the researcher sought to establish whether there exists mean differences among country of origin, factors about the kind of wine to drink or buy, and inclination to buy different wines for different occasions. The mean differences have been estimated using Chi-Square for the three categorical variables as reported under Tables 4.8 next.
Table 4.8: Test of association between country of origin and choice of wine to drink or buy
| 5. Countries come to mind * 3b. Different Occasions Crosstabulation | |||||
| Count | |||||
| 3b. Different Occasions | Total | ||||
| De | No | Ye | |||
| 5. Countries come to mind | China | 0 | 0 | 1 | 1 |
| France | 7 | 4 | 8 | 19 | |
| Italy | 2 | 0 | 5 | 7 | |
| Spain | 0 | 1 | 0 | 1 | |
| US | 0 | 2 | 0 | 2 | |
| Total | 9 | 7 | 14 | 30 | |
| Chi-Square Tests | |||
| Value | df | Asymptotic Significance (2-sided) | |
| Pearson Chi-Square | 13.981a | 8 | .082 |
| Likelihood Ratio | 14.725 | 8 | .065 |
| N of Valid Cases | 30 | ||
The results depicted under table 8 above show that there is cross-tabulation between countries that come to mind as origin for the best wine and intention to purchase or drink wine for different occasions. However, considering a 95% confidence interval an association between the two situations cannot be confirmed meaning origin and consumption of wine for different occasions may not have had influence on each other.
4.4 Data screening and analysis on Part I-III Findings
In this section data screening and analysis on the feedback for Parts I-III has been addressed. The first analysis is aimed to examine the results in terms of missing values and valid cases as reported in Table 4.9 below.
Table 4.9: Missing Values Assessment for Parts I-III
| Case Processing Summary | |||
| N | % | ||
| Cases | Valid | 30 | 100.0 |
| Excludeda | 0 | .0 | |
| Total | 30 | 100.0 | |
The indication is that the dataset consists of no missing values rendering all the cases to be non-excluded. The other assessment features the Principle Component Matrix meant to determine the degree to which the same data consists of valid cases as depicted in Table 4.10 below.
Table 4.10: Factor analysis on feedback for Parts I-III
| Communalities | ||
| Initial | Extraction | |
| Think about your next red wine purchase to have for dinner with some friends and family, if the wines shown in figure 1 are the only ones available, kindly select the one you would buy (select only one). “ | 1.000 | .836 |
| Brand | 1.000 | .876 |
| Design | 1.000 | .783 |
| Price | 1.000 | .772 |
| Scores | 1.000 | .897 |
| Origin | 1.000 | .844 |
| In case you do not settle on buying any of the five wines, would you shop elsewhere? | 1.000 | .868 |
| Below 100 | 1.000 | .883 |
| 101-300 | 1.000 | .943 |
| 301-600 | 1.000 | .909 |
| 601-999 | 1.000 | .874 |
| Above 1000 | 1.000 | .908 |
| 2b – France | 1.000 | .806 |
| US | 1.000 | .804 |
| Australia | 1.000 | .769 |
| Italy | 1.000 | .824 |
| Spain | 1.000 | .764 |
| Wine with an attractive bottle design: | 1.000 | .766 |
| Based on your first impressions of this bottle, how likely would you be to buy the wine?” | 1.000 | .946 |
| 3b (1). The design of the bottle is very attractive | 1.000 | .946 |
| 3b (2). The design of the bottle is very desirable | 1.000 | .861 |
| 3b (3). The bottle looks exactly the way a wine bottle should | 1.000 | .636 |
| 3b (4). This is a high quality wine | 1.000 | .770 |
| The wine is approximately: “ | 1.000 | .912 |
| Based on your first impressions of these bottles, how likely would you be to buy any of the wine?” | 1.000 | .886 |
| 4b (1). The design of the bottles are very attractive | 1.000 | .912 |
| 4b (2). The design of the bottle is very desirable | 1.000 | .674 |
| 4b (3). The bottles look exactly the way a wine bottle should | 1.000 | .886 |
| 4b (4). These are high quality wines | 1.000 | .750 |
| Each of the wine presented in this figure is approximately: “ | 1.000 | .908 |
| Based on your first impressions of this bottle, how likely would you be to buy the wine? | 1.000 | .979 |
| 5b (1). The design of the bottle is very attractive | 1.000 | .815 |
| 5b (2). The design of the bottle is very desirable | 1.000 | .645 |
| 5b (3). The bottle looks exactly the way a wine bottle should | 1.000 | .908 |
| 5b (4). This is a high quality wine | 1.000 | .979 |
| The wine is approximately | 1.000 | .840 |
| Extraction Method: Principal Component Analysis. | ||
The results are indicative that all the cases for Parts I-III have a high validity score above 40% or 0.4 which means the feedback has capacity to measure whatever was aimed to be measured as expressed in each of the constructs. In due course, the exact issues captured in the surveys i.e. Parts I-III is going to be evaluated and implications to the study identified.
The next analysis evaluates the descriptive statistics on feedback for part 1 of the survey. The results are presented under Table 4.11 below.
Table 4.11: Descriptive statistics on Part I
| Descriptive Statistics | |||||
| N | Minimum | Maximum | Mean | Std. Deviation | |
| Think about your next red wine purchase to have for dinner with some friends and family, if the wines shown in figure 1 are the only ones available, kindly select the one you would buy (select only one). | 30 | 1.00 | 5.00 | 3.0667 | 1.25762 |
| Brand | 30 | 1.00 | 5.00 | 3.3333 | 1.47001 |
| Design | 30 | 1.00 | 5.00 | 2.5000 | 1.19626 |
| Price | 30 | 1.00 | 5.00 | 2.9000 | 1.42272 |
| Scores | 30 | 1.00 | 5.00 | 2.9667 | 1.42595 |
| Origin | 30 | 1 | 5 | 3.3000 | 1.489 |
| In case you do not settle on buying any of the five wines, would you shop elsewhere? | 30 | 1.00 | 2.00 | 1.6000 | .49827 |
In part I of the study, the participants were invited to reflect about their next red wine purchase for dinner with some friends and family upon which they were to select the one they would prefer to buy (Mean = 3.0667, S.D. = 1.2576). The results indicate that on average the participants are inclined to buy bottle 3 of 5 which is 36.7% of the total respondents. The standard deviation proves to be lower than the mean value which in this case means there is consistency and stability in the outlook of this response as depicted in Figure 4.16 below.
Figure 4.16: Graphical Analysis on feedback for selection of wine to buy
The illustration shown above shows illustrates the frequencies around other choices of wine in terms of what the participants would prefer to buy; for instance, 23.3% of the participants indicate they would buy bottle 2 of 4 while 20% would purchase bottle 1 of 5. On the other hand, bottle 4 of 5 would appeal to 10% and the same case for bottle 1 of 5. The results manifested above, therefore, are an important trajectory for the ongoing analysis in the sense that they reflect the wine buying behaviour of consumers in Hong Kong. Later, the feedback trend has been used as one of the dependent variables against other predictor variables such as brand recognition, country of origin, critic scores and wine ratings, label/packaging design, and pricing in creating prediction outcome models.
Further review under Table 4.11 above indicates that Brand (Mean = 3.3333, S.D = 1.47001) is consistent and stable; the actual interpretation means that for the average participants brand recognition in a neutral manner has influenced their selection of wine. However, packaging design (Mean = 2.5000, S.D = 1.1963) shows that on average the participants disagreed to have been influenced by this factor in decisions to buy wine. A neutral position is evident in the case of Price (Mean = 2.9000, S.D = 1.4227), Scores (Mean = 2.9667, S.D = 1.4260), and Origin (Mean = 3.3000, S.D = 1.4890) as factors influencing decisions to buy or select a preferred wine brand among the participants.
In line with the above results, the study found that 60% of the participants said no regarding the matter of themselves not settling on buying any of the five wines hence opting to buy or shop elsewhere while 40% said yes to the matter. See Figure 4.17 below for the actual distribution of the feedback results.
Figure 4.17: Graphical analysis on decision to shop wine elsewhere
The results captured under Figure 4.17 above are also an important trajectory for this study in the sense that they reveal consumer behaviour towards wine among the participants. In due course, the trend is going to be invoked when building prediction outcomes around consumer buying behaviour of wine in Hong Kong.
The next review is on Part II results where the participants were invited to showcase their preference based on cost of wine at various values approximated using Hong Kong Dollars. The summary results have been captured under Table 4.12 using descriptive statistics.
Table 4.12: Summary statistics on preference to buy based on cost factors
| Descriptive Statistics | |||||
| N | Minimum | Maximum | Mean | S.D | |
| Below 100 | 30 | 1.00 | 5.00 | 4.0667 | 1.08066 |
| 101-300 | 30 | 1.00 | 4.00 | 2.5667 | 1.10433 |
| 301-600 | 30 | 1.00 | 3.00 | 1.8000 | .84690 |
| 601-999 | 30 | 1.00 | 4.00 | 2.5000 | .97379 |
| Above 1000 | 30 | 1.00 | 5.00 | 4.0667 | 1.41259 |
| Valid N (listwise) | 30 | ||||
The results captured in Table 4.12 above indicate that the feedback in each of the cases is consistent and stable since all the standard deviations are below the mean values. Moreover, as for wine costing below 100 Hong Kong dollars (Mean = 4.0667, S.D = 1.0807) the participants agreed they would buy it and similar affirmation is evident for Above 1,000 Hong Kong dollars. From the two feedbacks it can be subsumed that the participants agree to prefer wine that is at high cost and low cost. Nonetheless, approximations around 101-300 (Mean = 2.5667, S.D = 1.1043), 301-600 (Mean = 1.8000, S.D = .8469), and 601-999 (Mean = 2.5000, S.D = .9740) are cases where the respondents disagreed to prefer wines that fall within the mentioned costs. The researcher considers that the mentioned preferences are indicative of consumer buying behaviour of wine in Hong Kong; due to the importance of this feedback to the study, a correlation analysis was sought as shown in Table 4.13 below.
Table 4.13: Pearson Correlation on Preferences and Costs
| Correlations | ||||||
| Below 100 | 101-300 | 301-600 | 601-999 | Above 1000 | ||
| Below 100 | Pearson Correlation | 1 | .690** | -.324 | -.754** | -.590** |
| Sig. (2-tailed) | .000 | .081 | .000 | .001 | ||
| N | 30 | 30 | 30 | 30 | 30 | |
| 101-300 | Pearson Correlation | .690** | 1 | .015 | -.818** | -.755** |
| Sig. (2-tailed) | .000 | .938 | .000 | .000 | ||
| N | 30 | 30 | 30 | 30 | 30 | |
| 301-600 | Pearson Correlation | -.324 | .015 | 1 | .125 | -.450* |
| Sig. (2-tailed) | .081 | .938 | .509 | .013 | ||
| N | 30 | 30 | 30 | 30 | 30 | |
| 601-999 | Pearson Correlation | -.754** | -.818** | .125 | 1 | .451* |
| Sig. (2-tailed) | .000 | .000 | .509 | .012 | ||
| N | 30 | 30 | 30 | 30 | 30 | |
| Above 1000 | Pearson Correlation | -.590** | -.755** | -.450* | .451* | 1 |
| Sig. (2-tailed) | .001 | .000 | .013 | .012 | ||
| N | 30 | 30 | 30 | 30 | 30 | |
The correlation matrix reveals that majority of the cases return a significant p-value at 95% confidence interval meaning the preferences around each of the mentioned costs of wine are cross-related. For instance, cost below 100 (-.590**, Sig. = .001) means that the two preferences are moderately and inversely correlated. The results indicate more moderate and strong inverse correlations across the variables while there are instances of moderate and strong direct proportionality. The implications of the findings are that the preferences on wine influenced by cost factors have supported relationship meaning there is a convergence in the sense in which they impact on consumer buying behaviour towards wine in Hong Kong.
Other issues addressed in the survey were preferences of wine considering the origin i.e. countries of origin. The study, therefore, aimed to assess the level of agreement or disagreement with the matter. The summary results are presented in Table 4.14 below.
Table 4.14: Preference of wine on basis of the country of origin
| Descriptive Statistics | |||||
| N | Minimum | Maximum | Mean | Std. Deviation | |
| France | 30 | 1.00 | 4.00 | 1.6667 | .92227 |
| US | 30 | 1.00 | 5.00 | 3.3333 | 1.09334 |
| Australia | 30 | 1.00 | 5.00 | 3.8333 | 1.26173 |
| Italy | 30 | 1.00 | 5.00 | 2.1667 | 1.05318 |
| Spain | 30 | 2.00 | 5.00 | 4.0000 | 1.08278 |
| Valid N (listwise) | 30 | ||||
Foremost, the results are indicative that the standard deviations are below the mean values and that confirms stability and consistency in the feedback gauged using Likert scale for strongly agree to strongly disagree. As such, the interpretation is that France (Mean = 1.6667, S.D = .9223) and Italy (Mean = 2.1667, S.D = 1.0532) are cases that the respondents seemed to disagreed meaning they would not prefer wine from these locations. However, a neutral position is evident for the cases of US (Mean = 3.3333, S.D = 1.0933) and Australia (Mean = 3.8333, S.D = 1.2617). Spain (Mean = 4.000, S.D = 1.0829) has the highest approval rate since majority of the participants would prefer to wine from this origin. In the study, country of origin is operationised as a major predictor variable hence establishing the correlation of the feedback results would be meaningful, as shown in table 4.15 below.
Table 4.15: Pearson Correlation on Preference of wine on basis of the country of origin
| Correlations | ||||||
| 2b – France | US | Australia | Italy | Spain | ||
| France | Pearson Correlation | 1 | -.467** | -.138 | .095 | -.311 |
| Sig. (2-tailed) | .009 | .466 | .619 | .095 | ||
| N | 30 | 30 | 30 | 30 | 30 | |
| US | Pearson Correlation | -.467** | 1 | .117 | -.469** | -.291 |
| Sig. (2-tailed) | .009 | .539 | .009 | .118 | ||
| N | 30 | 30 | 30 | 30 | 30 | |
| Australia | Pearson Correlation | -.138 | .117 | 1 | -.653** | -.530** |
| Sig. (2-tailed) | .466 | .539 | .000 | .003 | ||
| N | 30 | 30 | 30 | 30 | 30 | |
| Italy | Pearson Correlation | .095 | -.469** | -.653** | 1 | .181 |
| Sig. (2-tailed) | .619 | .009 | .000 | .337 | ||
| N | 30 | 30 | 30 | 30 | 30 | |
| Spain | Pearson Correlation | -.311 | -.291 | -.530** | .181 | 1 |
| Sig. (2-tailed) | .095 | .118 | .003 | .337 | ||
| N | 30 | 30 | 30 | 30 | 30 | |
The areas marked in different colours are those that reveal significant correlation although all the incidences are an inverse proportionality. The conclusion is that there exist uncogent grounds to purport that the cases attributable to country of origin are linear when it comes to the preferences of wine among consumers in Hong Kong. In other words, there are weak outcomes that prove that these trends have a relationship whatsoever.
Further to the results above, the study sought to examine the reactions of the participants regarding wine with an attractive bottle design on whether it would increase their preferences. The results are captured below under Figure 4.18 using a graphical model.
Figure 4.18: Graphical trend on preferences on attractive bottle design
The trend above i.e. Figure 4.18 manifests that 50% being the largest sample of the participants disagreed with the fact of wine with an attractive bottle design having an influence of on their preferences. In fact, the general assertion is that this facet does not influence behaviour towards buying of wine among sampled consumers in Hong Kong. Further interpretation revealed that 20% of the respondents were neutral about the matter while 13.3% strongly disagreed and another 13.3% agreeing to the matter while the least of all at 3.3% indicate to strongly agree. In that case, attractiveness of the bottle design seems to not bother the cohort hence may not be adopted as a marketing mix frontier of wine in Hong Kong.
Part III of the surveys presented pertinent issues that included: (a) likelihood to buy the wine on grounds of the first impression of the bottle design (b) cost of wine. A summary of the cases has been explicated next in detail. Foremost, the participants were invited to state the likelihood of buying the wine based on their first impressions of the bottle. The average impression is as reported below under Figure 4.19.
Figure 4.19: Trend on likelihood of buying the wine
The indication is that majority of the participants at 50% depicted that they would purchase the wine as compared to 40% that might or might not purchase; whereas 10% established that they would probably not buy the wine on first impression of the bottle. The researcher considers the outlook above as showing some moderately strong sense of likelihood to buy among the participants since on a very rare occasion they indicated they would probably not buy the wine. Below in Table 4.16 is a Pearson correlation matrix indicating whether the responses around the same feedback are linear.
Table 4.16: Correlation on likelihood of buying the wine on first impression of the bottle
| Correlations | ||||
| Based on your first impressions of this bottle, how likely would you be to buy the wine?” | Based on your first impressions of these bottles, how likely would you be to buy any of the wine?” | Based on your first impressions of this bottle, how likely would you be to buy the wine? | ||
| Based on your first impressions of this bottle, how likely would you be to buy the wine?” | Pearson Correlation | 1 | .251 | -.070 |
| Sig. (2-tailed) | .180 | .714 | ||
| N | 30 | 30 | 30 | |
| Based on your first impressions of these bottles, how likely would you be to buy any of the wine?” | Pearson Correlation | .251 | 1 | -.060 |
| Sig. (2-tailed) | .180 | .751 | ||
| N | 30 | 30 | 30 | |
| Based on your first impressions of this bottle, how likely would you be to buy the wine? | Pearson Correlation | -.070 | -.060 | 1 |
| Sig. (2-tailed) | .714 | .751 | ||
| N | 30 | 30 | 30 | |
The results indicate that none of the p-values returns a confidence interval above 95% meaning the outlook on each of the likelihood to buy wine based on the first impression of the bottle is not correlated. Therefore, each of the matter does not relate to the other and cannot be used interchangeably to debate on consumer behaviour of wine shoppers in Hong Kong.
Further to the issues above, the study aimed to establish the perceptions of the participants around issues of attractiveness, desirability, fitness, and quality of wine in lieu of the of the design of the bottle. The summary of the results are going to be captured using descriptive statistics as shown under Table 4.17 below.
Table 4.17: Summary statistics on likelihood of buying the wine on first impression of the bottle
| Descriptive Statistics | |||||
| N | Min | Max | Mean | Std. Deviation | |
| 3b (1). The design of the bottle is very attractive | 30 | 1.00 | 4.00 | 2.0333 | .85029 |
| 3b (2). The design of the bottle is very desirable | 30 | 1.00 | 4.00 | 2.1000 | .92289 |
| 3b (3). The bottle looks exactly the way a wine bottle should | 30 | 1 | 5 | 3.2300 | 1.10400 |
| 3b (4). This is a high quality wine | 30 | 1 | 5 | 3.2000 | 1.12600 |
| 4b (1). The design of the bottles are very attractive | 30 | 1 | 4 | 2.0700 | .78500 |
| 4b (2). The design of the bottle is very desirable | 30 | 1 | 3 | 2.1700 | .69900 |
| 4b (3). The bottles look exactly the way a wine bottle should | 30 | 1.00 | 5.00 | 2.7333 | 1.17248 |
| 4b (4). These are high quality wines | 30 | 1.00 | 4.00 | 2.5333 | 1.00801 |
| 5b (1). The design of the bottle is very attractive | 30 | 1.00 | 5.00 | 3.5000 | 1.13715 |
| 5b (2). The design of the bottle is very desirable | 30 | 2.00 | 5.00 | 3.9000 | 1.02889 |
| 5b (3). The bottle looks exactly the way a wine bottle should | 30 | 1.00 | 5.00 | 2.9000 | 1.29588 |
| 5b (4). This is a high quality wine | 30 | 1.00 | 5.00 | 3.0667 | 1.20153 |
Overall, the data results above indicate that there is consistency and stability in the reported perceptions meaning the impressions given by the participants were not deviating much from the average feedback. The same can be used as a ground to rule out high cases or risks of outliers in the data feedback. Other than that there are indicative cases that the participants disagreed or remained neutral about the matter in question which is somewhat a pessimistic concern about the attractiveness, desirability, fitness, and quality of wine in context of the of the design of the bottle.
4.5 Data screening and analysis on Survey Findings
The current section is going to report the key findings of the surveys that capture subjective knowledge, objective knowledge, and self-confidence. To start the general information of the participants in terms of the demographic were captured i.e. gender, age, and frequency of buying wine. Figure 4.20 below illustrates the trend for gender.
Figure 4.20: Gender of the participants surveyed
As evidenced above, a high rate of the participants at 60% are males whiles 40% are females. The age of the participants is further captured under Figure 4.21 below.
Figure 4.21: Age of the participants surveyed
The indication is that a high percentage at 36.7% of the surveyed participants are in the age bracket 35-50 years followed by 26.7% at 26-35 years while 20% being the cohort with 18-25 years. The lowest rate of the surveyed participants at 16.7% is in the age bracket 51 and above.
Further, it was established the degree to which the surveyed participants would frequently buy wine as shown in Figure 4.22 below.
Figure 4.22: Frequency to buy wine
The results depict that 60% of the participants being the highest rate report to frequently buy wine less than six times in a year while 23.3% stated that they do so once every month; then 10% indicated to buy wine once per fortnight while 6.7% once per week. The feedback is essential to this ongoing study in the sense that it captures the trend of consumer buying behaviour towards wine in Hong Kong.
Therefore, using Likert scale the study further gauged the participants to comprehend various perceptions and behavioural orientation in an environment of wine products. The results are reported below under Table 4.18 using Principle Component Matrix meant to determine the validity of the main constructs.
Table 4.18: Factor analysis on perceptions and behavioural orientation
| Communalities | ||
| Initial | Extraction | |
| I understand how to judge the wine quality | 1.000 | .662 |
| I have adequate information about different wines | 1.000 | .794 |
| I do not feel knowledgeable about wine | 1.000 | .816 |
| I am less knowledgeable of wine compared to my peers | 1.000 | .666 |
| I have heard of majority of the wines around | 1.000 | .765 |
| I can tell if a wine is worth its price or not | 1.000 | .879 |
| Country of origin of wine is an essential when considering the kind of wines to buy. | 1.000 | .768 |
| Wine brands and branded influence my purchasing decision making process | 1.000 | .710 |
| Price is the most important factor to consider when buying wine. | 1.000 | .760 |
| I must check the label and the packaging design before I buy wine. | 1.000 | .849 |
| Normally, I feel my opinions are inferior. | 1.000 | .693 |
| I do not concentrate so much on what people think about me. | 1.000 | .858 |
| I seldom fear actions that would make others have a low opinion of me | 1.000 | .733 |
| When introduced to a stranger, I am never at loss for words. | 1.000 | .788 |
| My first reaction is always inferiority and shyness when confronted by strangers. | 1.000 | .824 |
| I do not have a good first impression on people. | 1.000 | .780 |
| Extraction Method: Principal Component Analysis. | ||
The results above indicate that all the cases have extraction values above 40% meaning there is validity; therefore, the survey feedback data has capacity to measure what it was purposed to measure in each of the items. For instance, one of the derivations from the results under Table 4.18 is that the ability to tell whether a wine is worth its price or not (.879) has the highest factor loadings meaning it is one of the major concerns that the participants have in Hong Kong. Thus, in a descending manner each of the extraction values can be used to comprehend the most felt issues among the participants around wine as a consumable product. For instance, it can be inferred that the capacity to judge the wine quality (.662) is the least felt concern among the participants since it has the lowest factor loadings.
4.6 Test of Significance
In the study the main cues have been brand recognition (BR), country of origin (COO), critic scores and wine ratings (CSWR), label/packaging design (L), and pricing (P) serving as the predictor variables while wine purchase behaviour being the criterion variable. All the demographic information such as age and gender has been considered as moderator variables. The first model relationship is as shown under Table 4.19 below and modelled as follows:
WPB = α + β1BR + β2COO + β3CSWR + β4L + β5P + ε ………………………………. (1)
Table 4.19: Regression test result 1
| Model Summary | |||||||||||
| Model | R | R Square | Adjusted R Square | Std. Error of the Estimate | |||||||
| 1 | .814a | .662 | .484 | .667 | |||||||
|
ANOVAa |
|||||||||||
| Model | Sum of Squares | df | Mean Square | F | Sig. | ||||||
| 1 | Regression | 16.530 | 10 | 1.653 | 3.722 | .007b | |||||
| Residual | 8.437 | 19 | .444 | ||||||||
| Total | 24.967 | 29 | |||||||||
| Coefficientsa | |||||||||
| Model | Unstandardised Coefficients | Standardised Coefficients | t | Sig. | Collinearity Statistics | ||||
| B | Std. Error | Beta | Tolerance | VIF | |||||
| 1 | (Constant) | -.563 | 2.029 | -.278 | .784 | ||||
| I have adequate information about different wines | .634 | .262 | .552 | 2.422 | .026 | .342 | 2.922 | ||
| a. Dependent Variable: Frequency | |||||||||
Noteworthy, is that the results captured in Table 4.19 above only reflect one of the criterion variables which represent brand recognition (β = .634, Sig. = .026); the assertion is that it is the only aspect that shows to have significant predictive effects towards buying behaviour of wine among the participants. The rest of the variables do not in the sense in which their p-values are above the 5% margin of error. The full results are captured in Appendix A. Now, further analysis of the model shows that the R2 (.662) is a goodness of fit meaning brand recognition (BR), country of origin (COO), critic scores and wine ratings (CSWR), label/packaging design (L), and pricing (P) explain 66.2% of the cases for buying behaviour towards wine. Moreover, the Analysis of Variance (F = 3.722, Sig. = .007) is proof that all the extrinsic cues have significant relationship to wine purchase behaviour that cannot be taken for granted in real life. Suffice to mention that the VIF metrics are at 2.922 for the significant predictor rendering the entire model to be valid and consistent as it lacks no traces of multicollinearity problem. In fact, due to this affirmation and the significance of the entire model the following hypothetical judgment is approved:
Ha: Extrinsic cues i.e. brand recognition (BR), country of origin (COO), critic scores and wine ratings (CSWR), label/packaging design (L), and pricing have statistically significant relationship on frequency of wine purchase trend among non-wine educated consumers in Hong Kong
The next model examined the moderating effects of gender and age of the participants on the influence of the identified extrinsic cues on wine purchase behaviour, as captured under Table 4.20 below.
Table 4.20: Moderated regression test results
| Model Summary | |||||||||
| Model | R | R Square | Adjusted R Square | Std. Error of the Estimate | Change Statistics | ||||
| R Square Change | F Change | df1 | df2 | Sig. F Change | |||||
| 1 | .814a | .662 | .484 | .666 | .662 | 3.722 | 10 | 19 | .007 |
| 2 | .831b | .691 | .473 | .673 | .029 | .806 | 2 | 17 | .463 |
| ANOVAa | ||||||
| Model | Sum of Squares | df | Mean Square | F | Sig. | |
| 1 | Regression | 16.530 | 10 | 1.653 | 3.722 | .007b |
| Residual | 8.437 | 19 | .444 | |||
| Total | 24.967 | 29 | ||||
| 2 | Regression | 17.260 | 12 | 1.438 | 3.173 | .015c |
| Residual | 7.706 | 17 | .453 | |||
| Total | 24.967 | 29 | ||||
As per the results depicted under Table 4.20 above, the moderating effects of gender and age are insignificant since the predictive effects for each of the predictor variables have not been proven beyond 95% confidence interval. Moreover, the change to the R value upon the moderating effects of the demographic factors has not returned a significant p-value although the entire model is still considered to be significant at 0.015. The full details of the regression model are featured in Appendix B. Therefore, the hypothetical argument is accepted as follows:
Ha: Extrinsic cues i.e. brand recognition (BR), country of origin (COO), critic scores and wine ratings (CSWR), label/packaging design (L), and pricing have statistically significant relationship to the frequency of wine purchase trends among non-wine educated consumers in Hong Kong when moderated by demographic factors such as age and gender
4.7 Review of Key Findings
The main findings have been on the extent to which extrinsic cues in this case brand recognition, country of origin, critic scores and wine ratings, label/packaging design, and pricing have real-life effects on frequency of buying behaviour of wine in Hong Kong among educated non-wine consumers. The findings have affirmed that the relationship is real and cannot be taken for granted at both theoretical and marketing context of wine in the region. The moderating or intervening effects of the demographics has also been exonerated meaning they must be issues to be considered in staging a feasible marketing plan for wine products in Hong Kong. Nonetheless, intricate features of wine design such as the bottle, colour or shape among others have not revealed to influence the perceptions and attitudes educated non-consumers of wine have towards various brands.
4.8 Summary of Chapter
The study has provided a thorough analysis on the wine purchase behaviour of educated non-wine consumers in Hong Kong. The information has powerful marketing insights that can be used by vendors of wine in Hong Kong to build appropriate market propositions and or value propositions. The results can be used to determine the best business model and product design for the wine products that are going to have an appeal to the target market. The main hypotheses of the study were as follows: extrinsic cues have statistically significant relationship to the trends in frequency of buying of wine among the educated non-wine consumers in Hong Kong which has been duly confirmed. The other hypothesis evaluated whether with the moderating effects of demographic factors such as age and gender of the educated non-wine consumers in Hong Kong, extrinsic cues maintained a statistically significant relationship to the trends in frequency of buying wine; the hypothesis was also duly confirmed in the study.
Appendix A: Hierarchical Regression model
| Model Summary | ||||
| Model | R | R Square | Adjusted R Square | Std. Error of the Estimate |
| 1 | .814a | .662 | .484 | .666 |
| ANOVAa | ||||||
| Model | Sum of Squares | df | Mean Square | F | Sig. | |
| 1 | Regression | 16.530 | 10 | 1.653 | 3.722 | .007b |
| Residual | 8.437 | 19 | .444 | |||
| Total | 24.967 | 29 | ||||
| Coefficientsa | ||||||||
| Model | Unstandardised Coefficients | Standardised Coefficients | t | Sig. | Collinearity Statistics | |||
| B | Std. Error | Beta | Tolerance | VIF | ||||
| 1 | (Constant) | -.563 | 2.029 | -.278 | .784 | |||
| 4a. I understand how to judge the wine quality | .057 | .249 | .048 | .229 | .821 | .411 | 2.434 | |
| b. I have adequate information about different wines | .634 | .262 | .552 | 2.422 | .026 | .342 | 2.922 | |
| c. I do not feel knowledgeable about wine | .003 | .263 | .003 | .010 | .992 | .234 | 4.271 | |
| d. I am less knowledgeable of wine compared to my peers | .135 | .190 | .119 | .710 | .486 | .633 | 1.580 | |
| e. I have heard of majority of the wines around | .452 | .290 | .475 | 1.559 | .136 | .192 | 5.211 | |
| f. I can tell if a wine is worth its price or not | -.190 | .270 | -.162 | -.705 | .490 | .336 | 2.976 | |
| g. Country of origin of wine is an essential when considering the kind of wines to buy. | -.272 | .161 | -.305 | -1.687 | .108 | .546 | 1.832 | |
| h. Wine brands and branded influence my purchasing decision making process | -.041 | .139 | -.042 | -.292 | .773 | .840 | 1.190 | |
| i. Price is the most important factor to consider when buying wine. | .050 | .180 | .053 | .278 | .784 | .499 | 2.005 | |
| j. I must check the label and the packaging design before I buy wine. | .059 | .158 | .058 | .372 | .714 | .729 | 1.372 | |
| a. Dependent Variable: Frequency | ||||||||
Appendix B: Moderated Regression model
| Model Summary | |||||||||
| Model | R | R Square | Adjusted R Square | Std. Error of the Estimate | Change Statistics | ||||
| R Square Change | F Change | df1 | df2 | Sig. F Change | |||||
| 1 | .814a | .662 | .484 | .666 | .662 | 3.722 | 10 | 19 | .007 |
| 2 | .831b | .691 | .473 | .673 | .029 | .806 | 2 | 17 | .463 |
| ANOVAa | ||||||
| Model | Sum of Squares | df | Mean Square | F | Sig. | |
| 1 | Regression | 16.530 | 10 | 1.653 | 3.722 | .007b |
| Residual | 8.437 | 19 | .444 | |||
| Total | 24.967 | 29 | ||||
| 2 | Regression | 17.260 | 12 | 1.438 | 3.173 | .015c |
| Residual | 7.706 | 17 | .453 | |||
| Total | 24.967 | 29 | ||||
| Coefficientsa | ||||||||
| Model | Unstandardised Coefficients | Standardised Coefficients | t | Sig. | Collinearity Statistics | |||
| B | Std. Error | Beta | Tolerance | VIF | ||||
| 1 | (Constant) | -.563 | 2.029 | -.278 | .784 | |||
| 4a. I understand how to judge the wine quality | .057 | .249 | .048 | .229 | .821 | .411 | 2.434 | |
| b. I have adequate information about different wines | .634 | .262 | .552 | 2.422 | .026 | .342 | 2.922 | |
| c. I do not feel knowledgeable about wine | .003 | .263 | .003 | .010 | .992 | .234 | 4.271 | |
| d. I am less knowledgeable of wine compared to my peers | .135 | .190 | .119 | .710 | .486 | .633 | 1.580 | |
| e. I have heard of majority of the wines around | .452 | .290 | .475 | 1.559 | .136 | .192 | 5.211 | |
| f. I can tell if a wine is worth its price or not | -.190 | .270 | -.162 | -.705 | .490 | .336 | 2.976 | |
| g. Country of origin of wine is an essential when considering the kind of wines to buy. | -.272 | .161 | -.305 | -1.687 | .108 | .546 | 1.832 | |
| h. Wine brands and branded influence my purchasing decision making process | -.041 | .139 | -.042 | -.292 | .773 | .840 | 1.190 | |
| i. Price is the most important factor to consider when buying wine. | .050 | .180 | .053 | .278 | .784 | .499 | 2.005 | |
| j. I must check the label and the packaging design before I buy wine. | .059 | .158 | .058 | .372 | .714 | .729 | 1.372 | |
| 2 | (Constant) | .447 | 2.325 | .192 | .850 | |||
| 4a. I understand how to judge the wine quality | .119 | .258 | .100 | .463 | .649 | .391 | 2.558 | |
| b. I have adequate information about different wines | .556 | .273 | .485 | 2.038 | .057 | .321 | 3.113 | |
| c. I do not feel knowledgeable about wine | -.074 | .276 | -.078 | -.269 | .791 | .218 | 4.596 | |
| d. I am less knowledgeable of wine compared to my peers | .072 | .206 | .063 | .349 | .731 | .553 | 1.809 | |
| e. I have heard of majority of the wines around | .345 | .314 | .362 | 1.101 | .286 | .168 | 5.961 | |
| f. I can tell if a wine is worth its price or not | -.254 | .277 | -.217 | -.917 | .372 | .325 | 3.079 | |
| g. Country of origin of wine is an essential when considering the kind of wines to buy. | -.265 | .163 | -.297 | -1.628 | .122 | .544 | 1.837 | |
| h. Wine brands and branded influence my purchasing decision making process | -.105 | .157 | -.111 | -.670 | .512 | .666 | 1.501 | |
| i. Price is the most important factor to consider when buying wine. | -.024 | .191 | -.025 | -.125 | .902 | .450 | 2.220 | |
| j. I must check the label and the packaging design before I buy wine. | .152 | .177 | .150 | .859 | .402 | .594 | 1.683 | |
| Gender | .296 | .281 | .159 | 1.052 | .307 | .796 | 1.257 | |
| Age | -.149 | .169 | -.162 | -.886 | .388 | .540 | 1.853 | |
| a. Dependent Variable: Frequency | ||||||||
CHAPTER 5 – DISCUSSION AND CONCLUSION
References
Apuke, O.D., 2017. Quantitative research methods: A synopsis approach. Kuwait Chapter of Arabian Journal of Business and Management Review, 33(5471), pp.1-8.
Arnold, R. C. G., 2012. Consumer perceptions of wine packaging design styles. The Moderating Role of Involvement, University of Edinburgh.
Berger, P.D., Maurer, R.E. and Celli, G.B., 2018. Two-Level Fractional-Factorial Designs. In Experimental Design (pp. 371-421). Springer, Cham.
Bizjak, P., Hristov, H. and Kuhar, A., 2020. Dimensions of the perceived value for wine from the perspective of Slovenian wine consumers. Acta Agriculture Slovenia, 115(1), pp. 89-96
Bobik, M., 2014. Thirst for wine – inside China’s wine industry: the success factors of marketing wine in China. Anchor Academic Publishing, Hamburg, Germany
Boncinelli, F., Dominici, A., Gerini, F. and Marone, E., 2018. Consumers Wine Preferences According to Purchase Occasion: Personal Consumption and Gift-Giving. Food Quality and Preference, 71(2019), pp. 270-278
Brochado, A., Vinhas da Silva, R. and LaPlaca, P. 2015. Assessing brand personality associations of top-of-mind wine brands. International Journal of Wine Business Research, 27(2), pp.125-142.
Bruwer, J. and McCutcheon, E. 2017. Marketing implications from a behaviorism perspective of consumption dynamics and socio-demographics of wine consumers. Asia Pacific Journal of Marketing and Logistics, 29(3), pp.519-537.
Canziani, B., Hwang, J. and Byrd, E.T., 2016. Further exploration of subjective knowledge in the wine sector. International Journal of Wine Business Research.
Carsana, L. and Jolibert, A., 2017. The effects of expertise and brand schematicity on the perceived importance of choice criteria: a Boudreaux wine investigation. Journal of Product and Brand Management, 26(1), pp. 80-90
Casas, R. and Makauskiene, B., 2013. Attitudes of Lithuanian consumers towards the country of origin of wine. Ekonomika, 92(1), pp. 133-155
Celhay, F., Cheng, P., Masson, J. and Li, W., 2020. Package graphic design and communication across cultures: an investigation of Chinese consumers’ interpretation of imported wine labels. International Journal of Research in Marketing, 37(1), pp. 108-128
Chocarro, R. and Cortiñas, M. 2013. The impact of expert opinion in consumer perception of wines. International Journal of Wine Business Research, 25(3), pp.227-248.
Defrancesco, E., Orrego, J.E. and Gennari, A., 2012. Would ‘new world’ wines benefit from protected geographical indications in international markets? The case of Argentinean Malbec. Wine Economics and Policy, 1(2012), pp. 63-72
Diddi, S. 2014. Understanding ethical consumption decisions: the role of values, attitudes and expectations in the apparel purchasing context. [Online] Available at:<http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.676.4729&rep=rep1&type=pdf> [Accessed 10/04/2020].
Durakovic, B., 2017. Design of experiments application, concepts, examples: State of the art. Periodicals of Engineering and Natural Sciences, 5(3).
Flynn, L.R. and Goldsmith, R.E., 1999. A short, reliable measure of subjective knowledge. Journal of business research, 46(1), pp.57-66.
Elliot, S. and Barth, J. E., 2012. Wine label design and personality preferences for millennials. Journal of Product and Brand Management, 21(3), pp. 183-191
Ellis, D., 2015. Consumer knowledge and its implications for aspects of consumer purchasing behavior in the case of information-intensive products. Royal Institute of Technology, Stockholm, Sweden.
Forbes, S. L. and Dean, D., 2013. Consumer perceptions of wine brand names, 7th AWBR International Conference June 12-15, 2013.
Garcia-Gallego, J. M., Chamorro-Mera, A. and Garcia-Galan, M., 2015. The region-of-origin effect in the purchase of wine: the moderating role of familiarity. Spanish Journal of Agricultural Research, 13(3), pp. 1-11
Gartner, W. C., Lee, W. F., Song, H., Marlowe, B., Choi, J. W. and Jamiyansuren, B., 2017. Effect of extrinsic cues on willingness to pay of wine – evidence from Hong Kong blind tasting experiment. British Food Journal, 120(11), pp. 2582-2598
Ghauri, P., Gronhaug, K. and Strange, R., 2020. Research methods in business studies. Cambridge University Press.
Gibbs, G. 2008. Analysing qualitative data (qualitative research kit),: Sage.
Goertzen, M.J., 2017. Introduction to quantitative research and data. Library Technology Reports, 53(4), pp.12-18.
Goldsmith, R.E., d’Hauteville, F. and Flynn, L.R., 1998. Theory and measurement of consumer innovativeness. European Journal of Marketing.
GWS, 2018. How do critic scores affect wine prices? a study of Napa valley wines, global wine score. [Online] Available at: https://medium.com/the-global-wine-score/how-do-critic-scores-affect-wine-prices-a-study-of-napa-valley-wines-226d8d08adb7 Accessed on May 12, 2020.
Heine, K., Phan, M. and Atwal, G., 2016. Authenticity and prestige: what luxury brands could learn from the wine industry?’ Luxury Research, 1(2), pp. 177-190
Heslop, L. A., Cray, D. and Armenakyan, A., 2010. Cue incongruity in wine personality formation and purchasing. International Journal of Wine Business Research, 22(3), pp.288-307.
Henley, C. D., Fowler, D. C., Yuan, J. Stout, B. L. and Goh, B. K., 2011. Label design: impact on millennials’ perceptions of wine. International Journal of Wine Business Research, 23(1), pp. 7-20
Hirche, M. and Bruwer, J., 2014. Buying a product for an anticipated consumption situation: observation of high-end low-involved wine buyers in a retail store. International Journal of Wine Business Research, 26(4), pp.295-318.
Hunag, A., Dawes, J., Lockshin, L. and Greenacre, L., 2017. Consumer response to price changes in high-priced brands. Journal of Retailing and Consumer Services, 6(9), pp. 1-24
Jaipakdee, N., Limpongsa, E. and Pongjanyakul, T., 2016. Optimization of minoxidil microemulsions using fractional factorial design approach. Pharmaceutical development and technology, 21(1), pp.86-97.
Kelley, K., Hyde, J. and Bruwer, J., 2015. US wine consumer preferences for bottle characteristics, back label extrinsic cues and wine composition. Asia Pacific Journal of Marketing and Logistics, 27(4), pp. 516-534
Lau, Y., Ng, A. K. and Guerrero, D., 2014. Becoming a major hub in the distribution of wine: Hong Kong as a gate to Asian markets, IFSPA. International Forum on Shipping.
Lee, F. S., 2012. Wine and the consumer price-perceived quality heuristics. International Journal of Marketing Studies, 4(3), pp. 31-35
Lewis, G., Charters, S., Lecat, B., Zalan, T. and Wolf, M.M., 2019. The impact of setting on wine tasting experiments. International Journal of Wine Business Research.
Li, J., Streletskaya, N.A. and Gómez, M.I., 2019. Does taste sensitivity matter? The effect of coffee sensory tasting information and taste sensitivity on consumer preferences. Food Quality and Preference, 71, pp.447-451.
Li, L., Hu, X. and Jun, Z., 2011. Do Chinese consumers prefer to buy imported wine? – The effect of country-of-origin. University of Gloucestershire.
Mellinger, C.D. and Hanson, T.A., 2016. Quantitative research methods in translation and interpreting studies. Taylor & Francis.
Miah, A.Q., 2016. Experimental design. In Applied Statistics for Social and Management Sciences (pp. 325-351). Springer, Singapore.
Morse, J. M., Barrett, M., Mayan, M., Olson, K. and Spiers, J. 2002. Verification strategies for establishing reliability and validity in qualitative research. International Journal of Qualitative Methods, 1(2), pp.13-22.
Mueller, S., Lockshin, L., Louviere, J., Francis, L. and Osidacz, P., 2011. How does shelf information influence consumers’ wine choice?, The Australian Wine Research Institute, Adelaide, AU
Mueller, S. and Szolnoki, G., 2010. Wine packaging and labeling – do they impact market price? A Hedonic Price Analysis of US Scanner Data, 5th International Academy of Wine Business Research Conference, Auckland, NZ.
Muhammad, A., Leister, A. M., McPhail, L. and Chen, W., 2013. The evolution of foreign wine demand in China. Australian Journal of Agricultural and Research Economics, 58, pp. 392-408
Niimi, J., Danner, L., Li, L., Bossan, H. and Bastian, S.E., 2017. Wine consumers’ subjective responses to wine mouthfeel and understanding of wine body. Food Research International, 99, pp.115-122.
Nowell, L.S., Norris, J.M., White, D.E. and Moules, N.J., 2017. Thematic analysis: Striving to meet the trustworthiness criteria. International journal of qualitative methods, 16(1), p.1609406917733847.
Oliver, R. L. 1980. A cognitive model of the antecedents and consequences of satisfaction decisions. Journal of Marketing Research, pp.460-469.
Outreville, J. F. and Le Fur, E., 2017. Determinants of the price of wine as a consumer good or investment plan. Brugundy School of Business, Dijon, France.
Perovic, N., 2014. Psychological Pricing in the Modern Retail: Case of the Wine Sector in Hard-Discounter Chains of Russia. University of Donja Gorica-Podgorica, Montenegro
Priilaid, D., Human, G., Pitcher, K., Smith, T. and Varkel, C., 2017. Are consumers’ quality perceptions influenced by brand familiarity, brand exposure and brand knowledge? Results from a wine tasting experiment. South African Journal of Business Management, 48(2), pp.45-54.
Ritchie, C., Elliot, G. and Flynn, M., 2010. Buying wine on promotion is trading-up in UK supermarkets; a case study in Wales and Northern Ireland. International Journal of Wine Business Research, 22(2), pp. 102-121
Robertson, J., Ferreira, C. and Botha, E., 2018. The influence of product knowledge on the relative importance of extrinsic product attributes of wine. Journal of Wine Research, 29(3), pp. 159-176
Silverman, D. 2009. Doing qualitative research. 3rd ed., London: Sage.
Sjostrom, T., Corsi, A. M. and Lockshin, L., 2014. Consumer perceptions of premium and luxury wine brands. Wine and Viticulture Journal, 29(3), pp. 68-70
Snow, A. P. and Weckman, G. R., 2018. Wine critic scores and consumer behavior in a major USA Metropolitan market. The Fourth International Conference on Big Data, Small Data, Linked Data and Open Data, Athens, Ohio USA.
Stolle, M., 2018. Labeling and packaging: the first taste is almost always with the eye. Wine and Viticulture Journal, 34(2), pp. 59
Tadajewski, M., 2016. Focus groups: history, epistemology and non-individualistic consumer research. Consumption Markets & Culture, 19(4), pp.319-345.
Tamas, A., 2016. Is the country of origin an important cue when buying wine?. Journal of Eastern Europe Research in Business and Economics, 2016(2016), pp. 1-14
Tang, V. C. and Cohen, E., 2014. Perception of wine labels by Hong Kong Chinese consumers. Academy of Wine Business Research, 8th International Conference,
Tang, V. and Tchetchik, A., 2015. Perception of wine labels by Hong Kong Chinese consumers. Wine Economics and Policy, 4(1), pp. 12-21
Terry, G., Hayfield, N., Clarke, V. and Braun, V., 2017. Thematic analysis. The Sage handbook of qualitative research in psychology, pp.17-37.
Uzgoren, E. and Guney, T., 2012. The snop effect in the consumption of luxury goods. Procedia – Social and Behavioral Sciences, 62(2012), pp. 628-637
Vesela, J. and Zich, R., 2015. The country-of-origin effect and its influence on consumer’s purchasing decision. ACTA Universitatis Agriculturae Et Silviculturae Mendelianae Bruenensis, 63(2), pp. 667-673
Viot, C. and Passebois-Ducros, J., 2010. Wine brands or branded wines? The specificity of the french market in terms of the brand. International Journal of Wine Business Research, 22(4), pp. 406-422
Vlachvei, A., Ourania, N. and Efterpi, T., 2012. Branding strategies in Greek wine firms. Procedia Economics and Finance, 1, pp. 421-430
Xiong, R. and Li, W., 2017. The Wine consumption and purchase behavior of college students: an investigation in the South of China. Universal Journal of Management, 5(6), pp. 271-277
Yadav, S.K., Singh, S. and Gupta, R., 2019. Sampling Methods. In Biomedical Statistics (pp. 71-83). Springer, Singapore.
Young, V., 2019. Focus on focus groups. College & research libraries news, 54(7), pp.391-394.
Appendices
Appendix A: Pilot Survey
Hello, my name is Steve Hung, a DBA student at the University of Wales Trinity Saint David. As part of the requirement for the award of the degree, I am obligated to research on: The influence of Extrinsic Cues on Wine Purchase in Hong Kong. Therefore, I am requesting for your help in filling out the questionnaire. Please answer all questions and note that there are no correct or wrong answers; I am only interested in your opinions based on your experiences. This research will only be used for academic purposes only and will be treated with a high level of confidentiality.
Please tick where applicable.
Gender: Male ( ) Female ( )
Age Bracket: 18-25 years ( ) 26-35 years ( ) 35-50 years ( ) 51 and above ( )
| # | Statement | Strongly agree | Agree | Neutral | Disagree | Strongly disagree |
| Scale of items for subjective knowledge | ||||||
| a. | I understand how to judge the wine quality | |||||
| b. | I have adequate information about different wines | |||||
| c. | I do not feel knowledgeable about wine | |||||
| d. | I am less knowledgeable of wine compared to my peers | |||||
| e. | I have heard of majority of the wines around | |||||
| f. | I can tell if a wine is worth its price or not | |||||
| Scale of items for objective knowledge | ||||||
| a. | Country of origin of wine is an essential when considering the kind of wines to buy. | |||||
| b. | Wine brands and branded influence my purchasing decision making process | |||||
| c. | Price is the most important factor to consider when buying wine. | |||||
| e. | I must check the label and the packaging design before I buy wine. | |||||
| Scale of items for self-confidence | ||||||
| a. | Normally, I feel my opinions are inferior. | |||||
| b. | I do not concentrate so much on what people think about me. | |||||
| c. | I seldom fear actions that would make others have a low opinion of me | |||||
| d. | When introduced to a stranger, I am never at loss for words. | |||||
| e. | My first reaction is always inferiority and shyness when confronted by strangers. | |||||
| f. | I do not have a good first impression on people. | |||||
Thank you for your Cooperation
Appendix B: Pilot Focus Group
Experiment Guideline
Hello, my name is Steve Hung, a DBA student at the University of Wales Trinity Saint David. As part of the requirement for the award of the degree, I am obligated to research on: The influence of Extrinsic Cues on Wine Purchase in Hong Kong. Therefore, I am requesting for your help in filling out the questionnaire. Please answer all questions and note that there are no correct or wrong answers; I am only interested in your opinions based on your experiences. This research will only be used for academic purposes only and will be treated with a high level of confidentiality.
Please tick where applicable.
(a). Think about your next red wine purchase to have for dinner with some friends and family, if the wines shown in the 1 below are the only ones available, kindly select the one you would buy (select only one).
(b.) On a five-point Likert Scale (Strong Agree to Strong Disagree), please indicate the extent to which the following factors influenced your selection:
Brand recognition
Label/Packaging design
Price
Critic Scores and Wine Ratings
(c). In case you do not settle on buying any of the five wines, would you shop elsewhere?
(YES) (NO).
Thank you for your Cooperation
Delivering a high-quality product at a reasonable price is not enough anymore.
That’s why we have developed 5 beneficial guarantees that will make your experience with our service enjoyable, easy, and safe.
You have to be 100% sure of the quality of your product to give a money-back guarantee. This describes us perfectly. Make sure that this guarantee is totally transparent.
Read moreEach paper is composed from scratch, according to your instructions. It is then checked by our plagiarism-detection software. There is no gap where plagiarism could squeeze in.
Read moreThanks to our free revisions, there is no way for you to be unsatisfied. We will work on your paper until you are completely happy with the result.
Read moreYour email is safe, as we store it according to international data protection rules. Your bank details are secure, as we use only reliable payment systems.
Read moreBy sending us your money, you buy the service we provide. Check out our terms and conditions if you prefer business talks to be laid out in official language.
Read more