Aarhus Universitet - Handelshøjskolen Århus Analyzing Consumer Choice of Beverages in Bosnia and Herzegovina (BiH) applying Means End Chain (MEC) Masters thesis Nino Jakupovic NJ64041 1/4/2009 Advisor: Liisa Lähteenmäki Dissertation Project April 2009 Nino Jakupovic NJ64041 ---------------------------------------------------------------------------------------------------------------------------------------------------------- Abstract An abstract of the dissertation of Nino Jakupovic for the Master of Marketing Title: Analyzing Consumer choice of Beverages in Bosnia and Herzegovina (BiH) applying Means End Chain (MEC) The beverage industry is one of the fiercest competed industries in the world. Competitors are always trying to find new ways to increase their sales and reach new customers. However, the question is how consumers are choosing beverages to consume, especially in the new countries in Europe like Bosnia & Herzegovina. The purpose of the present study was to determine why particular beverages for personal consumption are chosen in BiH. The study also sought to answer 1. Is the choice of different beverage categories linked to different personal consequences and values of consumer? 2. Are consumers in BIH more involved in some beverage categories than in others? 3. How can this knowledge be utilized by a hypothetical beverage producer when planning marketing strategies for their products in the future? Consumer analysis using laddering technique showed that there is to some degree link between the choice of beverage and personal values of consumers. Furthermore, analysis showed that consumers do tend to be more involved n some beverages than others. Lastly, using MECCAS model, analysis identified some alternative routes a hypothetical beverage producer can use to reach consumers in BIH _____________________________________________________________________________ Page 1 of 72 Nino Jakupovic NJ64041 ---------------------------------------------------------------------------------------------------------------------------------------------------------- Table of contents Chapter 1 ....................................................................................................................................... 5 1.0 Introduction ............................................................................................................................. 5 1.1 Problem formulation and frame .................................................................................... 5 1.2 Dissertation relevance ................................................................................................... 7 1.3 Dissertation Paradigm ................................................................................................... 8 1.4 Theoretical frame of dissertation ......................................................................................... 9 1.5 Limitations and definitions................................................................................................ 10 Chapter 2 ..................................................................................................................................... 12 2.0 Theoretical approach ............................................................................................................. 12 2.1 Cognitive consumer behavior............................................................................................ 13 2.2 Means-End Chain (MEC) approach .................................................................................. 15 2.2.1 The objectives of MEC approach ............................................................................... 16 2.2.2 MEC as a model of cognitive structure ...................................................................... 19 2.2.3. Additional problems regarding MEC approach ............................................................ 21 2.3 Application of MEC .......................................................................................................... 22 2.4 Product Involvement and MEC ......................................................................................... 24 Chapter 3 ..................................................................................................................................... 26 3.0 Research methods and data gathering of MEC ..................................................................... 26 3.1 Qualitative and quantitative data gathering techniques ..................................................... 27 3.1.1 Laddering by personal interviews .............................................................................. 27 3.1.2 Attribute elicitation and laddering interview ............................................................. 27 3.1.3 Quantitative data gathering techniques .................................................................... 30 3.1.4 “Hard” vs. “Soft” laddering ........................................................................................ 30 3.2 Analysis of laddering data ................................................................................................. 31 3.3 Why using MEC approach and laddering for this study ................................................... 33 Chapter 4 ..................................................................................................................................... 35 4.0 Research design ..................................................................................................................... 35 _____________________________________________________________________________ Page 2 of 72 Nino Jakupovic NJ64041 ---------------------------------------------------------------------------------------------------------------------------------------------------------- 4.1 Research method ............................................................................................................... 35 4.2 Selection of data gathering technique ............................................................................... 35 4.3 Selection of respondents.................................................................................................... 35 4.3 Elicitation technique.......................................................................................................... 37 4.4 Pilot research ..................................................................................................................... 38 4.5 The interviews ................................................................................................................... 38 Chapter 5 ..................................................................................................................................... 41 5.0 Research Findings and discussion ......................................................................................... 41 5.1 Participants ........................................................................................................................ 41 5.2 Content analysis ................................................................................................................ 43 5.3 Coffee ................................................................................................................................ 44 5.3.1 Involvement with coffee............................................................................................. 46 5.3.2 MECCAS for coffee ................................................................................................... 48 5.4 Fruit juice .......................................................................................................................... 49 5.4.1 Involvement with fruit juice ....................................................................................... 51 5.4.2 MECCAS for fruit juice ............................................................................................. 52 5.5 Bottled water ..................................................................................................................... 53 5.5.1 Involvement with bottled water .................................................................................. 55 5.5.2 MECCAS for bottled water ........................................................................................ 56 5.6 Carbonated soft drinks ...................................................................................................... 57 5.6.1 Involvement with carbonated soft drinks .................................................................. 59 5.6.2 MECCAS for carbonated soft drinks ......................................................................... 60 Chapter 6 Validity, reliability and generalization of the study ................................................... 62 Chapter 7 Conclusions and further research................................................................................ 64 7.0 Conclusion......................................................................................................................... 64 7.1 Implications for further research ....................................................................................... 65 _____________________________________________________________________________ Page 3 of 72 Nino Jakupovic NJ64041 ---------------------------------------------------------------------------------------------------------------------------------------------------------- Chapter 8 References .................................................................................................................. 66 _____________________________________________________________________________ Page 4 of 72 Nino Jakupovic NJ64041 ---------------------------------------------------------------------------------------------------------------------------------------------------------- Chapter 1 1.0 Introduction The author participated in marketing activities of the Coca-Cola Company Balkan and Adriatic Region in Sarajevo, Bosnia and Herzegovina (BiH), from January 2006 – March 2008. One of the main objectives of the company at the time was marketing and sales of different refreshment beverage types on different occasion. The company continuously gathered a lot of data on sales volumes and/or competitor activities in the area. However, this data only represented raw numbers on consumers’ choice, making it rather hard to understand why a certain product category was chosen in the first place, and if this choice was based on a more abstract aspects of consumer life or barely a physical need. Gutman (1984) suggested that consumers’ choice of product enables a consumer to achieve desired personal values. Since no data on consumer’s choices in BIH has been gathered to date, the author decided to investigate if the above statement is applicable to consumers in BiH as well. 1.1 Problem formulation and frame The aim of this dissertation is to analyze the following main question: Why are particular beverages for personal consumption chosen in BiH? When analyzing such marketing question/problem, it is necessary to specify the question and focus on the most relevant aspects of decision making (Reynolds & Olson, 2001). _____________________________________________________________________________ Page 5 of 72 Nino Jakupovic NJ64041 ---------------------------------------------------------------------------------------------------------------------------------------------------------- The main question covers personal beverage choice by individual consumers in BIH and the person consuming beverage(s) will be defined as the decision maker. The focus will be on the consequences/benefits a particular choice of a beverage type produces for a consumer. Furthermore, what are the personal values these benefits achieve for a consumer? The main question will cover consumer choice in context of consumption during spare (leisure) time and relaxation. Spare time is defined as the period of recreational and discretionary time before or after compulsory activities such as sleeping, going to work, running a business or attending school. Spare time will also include being with friends and/or family. There are four different beverage categories that will constitute units of analysis of consumer choice in BIH. These are carbonated soft drinks, fruit juices, bottled water and hot drinks (coffee). The consumers in BIH are analyzed for each of the four beverage categories mentioned. All of the above categories are widely represented in the area and constitute the biggest share of consumption beverages used for refreshment in spare time. It is anticipated that these categories will create a broad picture of factors influencing choice of different beverage categories, without having too broad or too detailed number of categories. Product involvement is another issue that can deepen the explanation of consumer behavior. Therefore, it is interesting to explore if there is a difference in product involvement when choosing different beverages. Knowledge about consumer choices can also have implications on beverage marketers. The results of this dissertation can be utilized by the beverage marketers in BIH, to structure their advertising campaign in the near future. _____________________________________________________________________________ Page 6 of 72 Nino Jakupovic NJ64041 ---------------------------------------------------------------------------------------------------------------------------------------------------------- The following, more specific questions will be used to gain insight in the main study question using a more targeted approach: Is the choice of different beverage categories linked to different personal consequences and values of consumer? Are consumers in BIH more involved in some beverage categories than in others? How can this knowledge be utilized by a hypothetical beverage producer when planning marketing strategies for their products in the future? The above questions will be answered via gathering and analyzing data from consumers in BiH. 1.2 Dissertation relevance Due to fairly recent war and political unrest in BiH, there has not either been any known research of the consumer behavior in the area, nor has the author encountered examples of earlier similar research associated with BiH. “A global look at Consumer involvement and use of products” (Zaichowsky, 1988), was a cross-cultural study of consumer involvement with various products, and the single piece of literature only briefly regarding former Yugoslavia, BiH. Although, a valuable article the author does not consider it sufficient. BiH has 4.5 million inhabitants and possible customers. The country is no longer at risk of war and people are starting to maintain regular daily activities. The research proposed in this study will therefore be of interest for a company that may wish to launch their products on BiH market. Former Yugoslavian countries are getting closer to becoming members of the European Union (EU). Even some Danish beverage companies may be interested to invest and deliver their product to this growing market. It is therefore anticipated that some parts of this dissertation could be useful for them as well. _____________________________________________________________________________ Page 7 of 72 Nino Jakupovic NJ64041 ---------------------------------------------------------------------------------------------------------------------------------------------------------- 1.3 Dissertation Paradigm Paradigm choice is a statement on the researcher’s view of the world (Ontology) and how the researcher thinks that knowledge will be created (Epistemology) (McKerchar, 2008). Furthermore, it is a question of (Methodology) which is influenced by ontology and epistemology. The choice of paradigm has an influence on the method approach one will choose later in the research process (Guba & Lincoln, 1994). Paradigm creates basis for understanding of application of theory and application of empiric in the research. The literature on the design and conduct of research traditionally suggests four (4) paradigms commonly referred as positivism, critical realism (post-positivism), critical theory and constructivism (Guba & Lincoln, 1994). Major differences/similarities between the different paradigms are summarized in Table 1. Table 1: Overview of different paradigm approaches Critical Realism Positivism Ontology Naive realism, “reality” real Epistemology Objectivist/finding truth Objectivist/ findings probably true Hypothesis testing Modified experimental. Majorly Quantitative Both quantitative qualitative Method Imperfect reality real Critical Theory Constructivism Crystallized over time Local /specific constructed reality Subjectivist / value mediated findings Subjectivist Dialogic Hermeneutical Source: Guba & Lincoln, 1994 with adjustments Critical realism is the most appropriate choice of paradigm for this dissertation. Critical realism believes that there is a world to research, but one can never be completely objective toward the world (Guba & Lincoln 1994). _____________________________________________________________________________ Page 8 of 72 Nino Jakupovic NJ64041 ---------------------------------------------------------------------------------------------------------------------------------------------------------- This approach tends to rely on dialogic methods and combines observation and interviewing (Sohlberg, 2003). This dissertation will be based on beliefs of critical realism scientists who consider the understanding of real peoples’ lived experience are very important for conducting research (Littlejohn, 2002). When research is conducted from this perspective one can use a mixture of theoretical reasoning and observable experiments. The theories in the critical realism are seen as a mean to create knowledge but every theory changes in time, so neither theory is completely exhaustive. Furthermore, both qualitative and quantitative methodologies are viewed as appropriate methodologies according to critical realism paradigm (Krauss, 2005; Guba & Lincoln, 1994) 1.4 Theoretical frame of dissertation Theoretical frame of the dissertation is based on the problem statement and cognitive consumer behavior will be the departing point of theoretical approach. The approach chosen to model the cognitive structure of consumer will be Means-End Chain (MEC). The MEC’s approach refers to various methods for interviewing consumers about the reasons for their decision choice and understanding, and analyzing of their responses in terms of linkages between outcomes (Reynolds & Olson, 2001, Wagner, 2007). The main assumption is that consumer does not choose product because of the product itself, but for the benefits consumption of the product can provide. The benefits/consequences of product consumption are important for realization of consumer’s higher values and goals (Costa Dekker & Jongen, 2004). In respect to the main purpose of the dissertation, the author will assume that the choice of specific beverage will depend on personal consequences this choice has on consumers and personal values consumers will realize from the benefits of the consumption. There are two epistemological views/approaches on MEC theory in consumer behavior (Grunert & Grunert 1995; Grunert, Beckmann & Sørensen 1995; Costa, Dekken, & Jorgen, 2004): _____________________________________________________________________________ Page 9 of 72 Nino Jakupovic NJ64041 ---------------------------------------------------------------------------------------------------------------------------------------------------------- Motivational view –MEC is concerned with gathering insights into consumer’s motives for buying in specific situations and how they are linked to shopping behavior (Grunert & Grunert, 1995; Costa, Dekken, & Jorgen, 2004). These insights are of qualitative nature. The user of this approach can evaluate if the approach is useful by assessing if he/she have acquired better understanding of consumers’ decision making (Costa, Dekken, & Jorgen, 2004). Cognitive structure view – MEC is used as a model of consumers’ consumptionrelevant cognitive structure (how consumption knowledge and experience is stored in the memory of consumer) (Grunert, Beckmann & Sørensen, 1995, Peter, Olson & Grunert, 1999). The utility of this approach is measured by its usefulness to predict consumption behavior in a given situation. The MEC literature on does not provide a clear answer on which of these two views is to be preferred (Grunert & Grunert, 1995, Costa, Dekken, & Jorgen, 2004). Grunert (1995), states that the choice of MEC approach depends on the general approach one has chosen as the basis of analysis. Costa et.al (2004) suggests that regardless of the approach chosen one has to consider the predictive ability of the conducted study. The research for this dissertation will be conducted using the cognitive structure view approach. The choice was made on the assumption that abstract reasons which guide consumers to choose a certain beverage are organized in the consumer cognitive structure, and that MEC can be used as a model of cognitive structure that concerns consumption. The relationship of product involvement and the MEC approach as well as literature overview of MEC approach application in consumer studies will also be discussed. 1.5 Limitations and definitions The main focus of this study is personal consumption of beverages, rather than the decision making which affects a family or a place of business. _____________________________________________________________________________ Page 10 of 72 Nino Jakupovic NJ64041 ---------------------------------------------------------------------------------------------------------------------------------------------------------- The research for this project was conducted in former Yugoslavia region which is comprised of four countries that speak the same language and have the same cultural background but different economic and political development and current opportunities. Due to limited resources available to author, the research will be focused on BiH as the single country in Balkan Peninsula region. It would however be interesting to include other countries and compare data in future research. Furthermore, BiH is a multiethnic and multi religious country. It is however deemed irrelevant to discuss the consumer preferences on the ethnic level as it would make the discussion highly complex and confusing. Moreover, considering the history of the area and the privacy of the informants it is not wise to impose such personal questions (i.e. religion, ethnic background). The political situation which is assumed to be highly complicated will not be a part of the dissertation. According to English language dictionary, the term ‘beverage’ is defined “a drink, especially tea, coffee, or other non-alcoholic drink” In accordance to this the focus will be on non-alcoholic beverages. Gutman (1984) also placed beer and wine in the beverages category. This research will however mainly focus on non-alcoholic beverages. Non-alcoholic beverages are also chosen because of the earlier mentioned difference in religious beliefs within the country. The beverage categories are narrowed down by excluding categories like milk and energy drinks as these are considered to be irrelevant for this study. _____________________________________________________________________________ Page 11 of 72 Nino Jakupovic NJ64041 ---------------------------------------------------------------------------------------------------------------------------------------------------------- Chapter 2 2.0 Theoretical approach The American Marketing Association defines Consumer Behavior as: “The dynamic interaction of affect and cognition, behavior, and environmental events by which human beings conduct and exchange aspects of their lives” (Peter et al, 1999; American marketing association (http://www.marketingpower.com). Consumer behavior has become more and more essential for marketers, academic researchers in recent years. Assael (1994) argued that consumers determine the profits and sales of the companies by their purchasing decisions. Consumer behavior blends different elements from many sciences such as psychology, economics, and sociology (Peter, Olson & Grunert, 1999). There are many models, theories and concepts developed with the aim to understand and use consumer behavior. However, none of the approaches have been fully accepted by all the researchers on the subject (Peter, Olson & Grunert, 1999). Marketing approach to consumer behavior can be divided in behavioral and cognitive approach. Behavioral approach explores links between the consumer environment and behavior, while cognitive approach is concerned with mental structure and thinking processes of consumers (Zanoli & Naspetti, 2002). This includes mental and emotional processes that consumers engage in when selecting, purchasing and using products (Zannoli & Naspetti, 2002). This particular dissertation begins with a cognitive approach to consumer behavior, acknowledging that both approaches are acceptable to analyze consumer behavior, emphasizing a high degree of complementarities between the two (Zanoli & Naspetti, 2002; Peter et al, 1999). _____________________________________________________________________________ Page 12 of 72 Nino Jakupovic NJ64041 ---------------------------------------------------------------------------------------------------------------------------------------------------------- 2.1 Cognitive consumer behavior Theoretical analysis of the problem statement is conducted via analysis of cognitive consumer behavior, description of the MEC approach, and by relating MEC approach to cognitive consumer behavior to establish unified understanding framework between concepts. According to Grunert & Grunert (1995), cognitive view of consumer behavior views behavior as the interaction between cognitive structures and cognitive processes. A cognitive structure is the organization of knowledge and information in human memory. It is usually described as network models1 consisting of two types of elements: cognitive categories (nodes) and their associations (links) (Grunert, Beckmann, Sørensen, 1995). Cognitive processes are processes consumers consciously or subconsciously use to change cognitive structure with new information from the environment (Peter, Olson & Grunert, 1999). Cognitive processes are also used to collect information from the cognitive structure and direct future behavior (Grunert & Grunert, 1995). There are two categories of cognitive processes (Grunert & Grunert, 1995): Automatic processes –are subconscious, do not limit capacity but it takes time to learn them, and once learned they are very effective (i.e. an experienced driver can drive from point A to B, without being consciously aware of anything that occurred during the drive).Strategic process are conscious, they limit the capacity and are used as problem solving activities, i.e. planning the above mentioned example performed for the first time in that drivers life. The driver will consciously try to locate point B. 1 There are also other descriptions such as scripts and schemas (Peter et al. 1999) _____________________________________________________________________________ Page 13 of 72 Nino Jakupovic NJ64041 ---------------------------------------------------------------------------------------------------------------------------------------------------------- Cognitive Information Processing Models (CIPMs) are developed to explain how the interaction between cognitive processes and cognitive structures occur in the cognitive system (Peter, Olson & Grunert, 1999). There are different cognitive models in literature of consumer behavior such as one constructed by Howard and Sheth (1969), which relates information processing to decision making. Bettman’s Consumer Information Processing Model (CIPMC) of choice, describes consumers as problem solvers with a limited capacity to process information so called “bounded reality” (Bettman, 1979; Bettman, Luce & Payne, 2002). Figure 1 shows the decision making Cognitive Processing Model by Peter, Olson & Grunert (1999). Figure 1: Cognitive processing model of Consumer Decision Making Source: Peter, Grunert & Olson, (1999) The model illustrates2 how information stimuli from the outside environment are interpreted by consumer in order to create new knowledge, meanings and beliefs. 2 Blue squares indicate cognitive process. Black squares indicate outside stimuli and behavior _____________________________________________________________________________ Page 14 of 72 Nino Jakupovic NJ64041 ---------------------------------------------------------------------------------------------------------------------------------------------------------- Interpretation processes include comprehension and attention. They are cognitive processes which determine how to create personal knowledge from information, which information to keep (store in the memory), which to discard, and which information to integrate with already stored knowledge in memory. Integration processes are used to integrate new and old knowledge, to form attitudes and decisions which result in given behavior. The knowledge kept in the memory of consumers is organized in structures of knowledge – cognitive structure. The literatures often use MEC approach as means of eliciting consumers’ cognitive structure (Grunert, Lahteenmaki, Nielsen, Poulsen, Ueland, Astrom, 2001; Peter, Olson & Grunert, 1999). 2.2 Means-End Chain (MEC) approach Means End Chain (MEC) was conceptualized by Jonathan Gutman in 1982 in the article entitled “A Means End Chain model based on consumer categorization process”. Gutman believes that people think at different levels of abstraction. Physical product attributes are not always the main focus of consumer thinking, but consumers think of more abstract levels such as benefits of product use, or even the personal values that the products can help them achieve (Howard & Warren, 2001). In order to conceptualize MEC Gutman discussed the relationship between earlier research and theories which tried to relate consumer’s values to the individual behavior (Gutman, 1982). These theories included Rokeach’s theory of human values which people use to formulate attitudes and opinions. Furthermore, Howards Choice Theory which says that product knowledge in consumers is hierarchically organized by different levels of abstraction. In 1975, Young and Feigns described The Grey Benefit Chain (GBC) approach by debating the possibility to gather data of consumer’s psychological perceptions of the products (Gutman, 1982; Rokeach, 1968 & 1973; Howard 1977; Barrena, 2009; Young & Feigin, 1975). Gutman used background research and elements from these theories to introduce the MEC into the field of marketing and _____________________________________________________________________________ Page 15 of 72 Nino Jakupovic NJ64041 ---------------------------------------------------------------------------------------------------------------------------------------------------------- consumer research (Gutman, 1982; Barrena, 2009). He proposed MEC as a cognitive structure of meaning that relates product attributes and consequences of product use with personal values of consumers. In other words, MEC explains the relationship between consumer knowledge and consumer behavior (Barrena, 2009; Howard & Warren, 2001). 2.2.1 The objectives of MEC approach The main focus of Gutman’s work was presented in Thomas J. Reynolds and Gutman’s well known article ‘Laddering: Theory, Method, Analysis and Interpretation’ (1988): “Means End theory specifically focuses on attributes that exist in products (the means), the consequences for the consumer provided by the attributes and the personal values (the ends) the consequences reinforce.” The above definition defines consumer relationship to product by hierarchical cognitive structure with linkages between attributes of the product, consequences of product use and personal values/goals of consumers (Hofstede, Audenaert, Wedel & Steenkamp, 1998). Concepts of attributes, consequences and values create content of consumer knowledge, and the hierarchical linkages between them form the structure of consumer knowledge-cognitive structure (Hofstede et.al, 1998). Values (V) ↑ Consequences (C) ↑ Attributes (A) Figure 2: Attributes, Consequences and Values and linkages between them as three level chains. _____________________________________________________________________________ Page 16 of 72 Nino Jakupovic NJ64041 ---------------------------------------------------------------------------------------------------------------------------------------------------------- Source: Devised by author The hierarchical linkages between the three concepts represent higher levels of abstraction. The higher the level of abstraction, the closer one gets to the personal values of consumer. Following are the descriptions of the three concepts of consumer knowledge: Attributes represent knowledge about characteristics of the product (i.e. the characteristics of the products which are important to the preferences of the consumers (Gutman, 1982; Reynolds & Gutman, 1988). Attribute knowledge consists of knowledge of concrete tangible attributes and abstract intangible attributes (Snelders & Schoorman, 2004). For example, sodium can be an attribute of carbonated soft drink. Consequences are positive or negative implications of the consumers purchase or use of products (Bech-Larsen, Nielsen et al., 1997). Most researchers also agree that positive feelings are the results of benefits from consumption; therefore consequences can be used as benefits of product consumption (Snelders & Schoorman, 2004). Furthermore, consequences can be divided into: psychological and functional consequences of product use (Peter, Olsen & Grunert, 1999). Skyte and Bove (2004) further describe consequences as “the outcomes produced by the attributes”. For example, the tingling feeling of soda in the carbonated soft drink is a consequence produced by the soda. Values are preferred end states of being that drive the behavior, and the top of the hierarchy. One can make distinction between instrumental values (i.e. to be ambitious) and more broad psychological terminal values of end states (i.e. happiness) (Peter et al., 1999). Continuing with previous examples the tingling feeling of soda makes a person feel happy and happiness can be a value of a person. _____________________________________________________________________________ Page 17 of 72 Nino Jakupovic NJ64041 ---------------------------------------------------------------------------------------------------------------------------------------------------------- Values play dominant role in guiding consumer choice patterns, and they represent the consequence consumer strives for when choosing a specific product among others (Gutman, 1982). MEC approach is based on this fundamental assumption. There are other assumptions of MEC approach - Consumers cope with magnitudes of products by grouping them into sets of categories based on their functions in order to reduce the complexity of choice; - Consumers learn to associate particular consequences with particular actions they take, and finally - All actions have consequences Some researchers tend to divide the above three levels of abstraction in to further two sublevels in order to create finer analysis of concepts (Claeys, Swinnen, Abeele, 1995). The three lower levels of MEC will then present the consumer’s knowledge of the product while the three upper levels will comprise the consumers self knowledge. Figure 2 illustrates this. Figure 3: 6 levels of MEC Source: (Peter et al., 1999) There could be more than one MEC concerning a single product. On the other hand the MEC does not have to be complete (Lind, 2006). Incomplete MEC can point to consumer inexperience with a product or simplicity of the product (Reynolds & Olson, 2001). Additionally, more symbolic products can reach higher abstraction levels than product which are bought for its functionality (Gutman, 1997). Studies that involved _____________________________________________________________________________ Page 18 of 72 Nino Jakupovic NJ64041 ---------------------------------------------------------------------------------------------------------------------------------------------------------- MEC of food products showed more attributes and consequences than values due to its habitual purchase, and lack of expertise in food (Costa et al., 2001). Olson & Reynolds (1983) and Reynolds & Olson (2001) suggested that the number of levels one will adopt in MEC depends on complexity of the product in question. Six levels of MEC are rather too complex according to the literature (Reynolds & Olson, 2001). If a product is simple fewer chains will be developed. Therefore, it is assumed that beverages belong to a group of simpler products and the assumption of this dissertation will be that cognitive structure of consumers’ choice of beverages will be four standard and most common levels of MEC (Reynolds & Olson, 2001) attributes, functional consequences psychological consequences and values. MEC approach models consumers cognitive structure primarily by using data gathered utilizing qualitative interviewing technique called laddering (Grunert et al. 2001, Reynolds & Gutman, 1988). The data derived from these interviews give foundation for creating an aggregate map of consumer cognitive structure called hierarchical value map (HVM) (Reynolds &Olson, 1983; Grunert, Beckmann & Sørensen, 1995). More detailed explanation of these techniques will be conducted in methodical part of the dissertation. 2.2.2 MEC as a model of cognitive structure It is important to relate the concept of how memory is organized in cognitive structure and MEC approach to evaluate advantages and disadvantages of MEC as model of cognitive structure. The applicability of MEC as a model of cognitive structure was discussed by Grunert, Beckmann & Sørensen (2001). They fitted MEC with the following 6 dimensions basic network models of cognitive structures apply: Episodic vs. Semantic - Is the information in the cognitive structures built on general or episodic information? If a person should provide knowledge about beverages it will include general knowledge about beverages and not only _____________________________________________________________________________ Page 19 of 72 Nino Jakupovic NJ64041 ---------------------------------------------------------------------------------------------------------------------------------------------------------- knowledge on single individual usage event. Therefore, it is assumed that MEC for choice of beverages can be regarded as semantic. Declarative vs. Procedural - Can knowledge be verbalized or non-verbalized? MEC’s can be regarded as declarative as the knowledge gathered from the laddering interviews can primarily be verbalized by the respondent. Verbal vs. imagery - Is the knowledge organized in verbal form or is it structured in the image form? MEC can be regarded as verbal due to language used to measure results. However, it is also likely that some consumers store mental imagery much better. However, in this case the focus is not on brand or label therefore limited imagery knowledge is required. Hierarchical vs. Not hierarchic - Is knowledge organized in hierarchical or non hierarchical manner? It is assumed that MEC is hierarchic due to A – C –V levels of abstraction. Positional vs. Distributional - Is knowledge stored in specific position or distributed across cognitive structure? MEC can be regarded as positional due to one knowledge category refers to a specific concept (Attribute, Consequence or Value). Types of Associations - MEC has associations restricted to a certain casual type; Attribute – Consequences - Values Fitting MEC to these dimensions shows that MEC can be used as a model of cognitive structure. However, there is no evidence that MEC should be regarded as the most appropriate model of cognitive structures as it is possible to find examples of consumers’ knowledge which follows different dimensions from MEC (Grunert et al. 1995). For example, mental imagery is sometimes easier to memorize than verbal. Furthermore, consumers can sometimes tend to remember more episodic information from previous experiences i.e. drinking coffee and being sick. The data gathered from laddering and the MEC itself are also influenced by previously described constant interaction between cognitive structures and processes. _____________________________________________________________________________ Page 20 of 72 Nino Jakupovic NJ64041 ---------------------------------------------------------------------------------------------------------------------------------------------------------- Being influenced by cognitive process MEC can only be an estimate and not a complete model of cognitive structure which can explain and predict behavior (Grunert et al. 1995). Grunert, Beckmann & Sørensen (1995) discuss that a lot more research is required to position MEC into larger cognitive theory of consumer behavior, 1995). The possibility of building on theory of reasoned action by Fishben & Ajzen (1985) is suggested. Due to lack of overall theory that establishes MEC as a complete cognitive model of consumer behavior, Grunert & Grunert (1995), suggest using MEC as an excerpt of the cognitive structure considered relevant to explain buying behavior. It is anticipated that the information gathered from this study will not show the complete cognitive structure of consumers in BIH but only the estimate of the structure concerning beverage consumption. The overall paradigm of the dissertation also states that reality is imperfect and that findings are probably true (Chapter 1.3). The activation of knowledge from the cognitive structure is explained by spreading activation theory of memory (Anderson, 1983, Grunert & Grunert 1995, Rusell et al., 2004). This theory suggests that cognitive category will be recalled by extern stimuli (in this case MEC data elicitation technique). Once a cognitive category is activated, the activation spreads to other cognitive categories making any related association more accessible (i.e. values and consequences) (Russel et al., 2004) 2.2.3. Additional problems regarding MEC approach The literature has also identified additional problematic issues with the application of MEC approach. Theory and method for MEC are developed informally which often represents a major problem when certain recommendations should be provided from data gathered using MEC techniques (Van Rekom & Wierenga, 2007; Reynolds & Olson, 2001). _____________________________________________________________________________ Page 21 of 72 Nino Jakupovic NJ64041 ---------------------------------------------------------------------------------------------------------------------------------------------------------- Furthermore, Van Rekom & Wierenga (2007) question the existence of hierarchy with consumers. MEC theory only assumes that a hierarchy structure exists. Researchers argue that the existence of hierarchy would establish MEC as asymmetrical construct which is not certain. An overview of earlier research that used MEC as well as statistical tests of data gathered, have shown that structure of linkages between A → C →V sometimes headed in the direction of A → V → C or even in a looping way (Van Rekom & Wierenga, 2007). This puts hierarchy assumption into question. This problem is also partly confirmed by Scholdrer and Grunert (2005) who have made analysis of different assumptions of MEC and where they state that hierarchy question was “particularly problematic”. 2.3 Application of MEC Reynolds & Gutman (1988) suggested using MEC as an instrument of segmenting consumers according to their values, orientations for a product or a brand. Gutman (1984) applied MEC in his study of beverage consumption. He used MEC analysis to map consumers’ value orientation toward different beverages. The study concentrated on single beverage category as well as overall relations across all beverages. The study showed that MEC model and its data gathering techniques are applicable for studies on beverages categories, which is well in line with this dissertation. Furthermore, results derived from the MEC approach are used as basis for evaluating advertising or developing basis for advertising. This is done in conjunction with model called MECCAS (Means-End Chain Conceptualization for Advertising Strategy), (Gutman, 1982, Reynolds & Craddock, 1988). MECCAS integrate the primary consumer research data from MEC analysis in order to create objective basis upon which advertisement strategic process can be grounded (Reynolds et. al 1988). The MECCAS model identifies five levels of an advertising strategy: driving force, leverage point, consumer benefit, message elements and executional framework. _____________________________________________________________________________ Page 22 of 72 Nino Jakupovic NJ64041 ---------------------------------------------------------------------------------------------------------------------------------------------------------- Gutman (1984) used MECCAS to suggest advertising advice producers can use for different beverage categories. Bech-Larsen (2001) also used MECCAS and concluded that use of MEC analysis in conjunction with MECCAS can create stronger product value associations with consumers as well as better communication between advertising agencies and their clients. Reynolds and Whitlark (1995) are focusing on using means end framework to develop six thinking tools for developing a communication strategy and advertising copies. Additionally, MEC is well known for its application in the food products studies.MEC provides an understanding of drivers behind food choice and marketing “hooks” for the promotion of the products (Rusell, Leppard & Cox, 2003). There are numerous articles regarding food products and use of MEC model. Lind (2006) has used the motivational view of MEC in order to identify motivational structure of consumers in relation to purchase of different kinds of pork meat brands in Sweden as well as the consumer involvement in purchase of pork meat. Barrenas et al. (2009) used MEC to conduct beef consumption frequency study; Costa et al. (2007) to study a different choice of meal ; and Krystallis et al. (2008) to define the most important functional food attributes that affect consumers’ purchasing decisions . Other studies that used MEC compared cognitive structures between groups of consumers in different countries such as study of perception of genetically modified food products (Grunert et al., 2001), or the cognitive structures of consumers for a specific clothing line across three countries by Botschen and Hemetsberger (1998). MEC analysis also includes studies of more intangible issues such as schools (Veludo de Oliviera & Ikeda, 2004), courses (Goldenberg et al, 2000) or recycling (Bagozzi & Dabholkar, 1994). One can see that MEC analyses are applicable in various marketing and consumer research fields. Furthermore, many studies using food products can be related to this dissertation due to similar assumptions in relation to consumer cognitive structure _____________________________________________________________________________ Page 23 of 72 Nino Jakupovic NJ64041 ---------------------------------------------------------------------------------------------------------------------------------------------------------- regarding foods and beverages. Previously conducted research definitely justifies the application of MEC approach to this study. Furthermore, an interesting point in respect to previous studies overview, concerns the use of MECCAS models. In accordance with problem statement (Chapter 1.2), one of the objectives deals with possible use of knowledge about consumers by a beverage producer. Using MECCAS for this goal can be valuable. Possibly to use MECCAS as a suggestion for different beverage producers advertising campaigns in BiH. Moreover Linds (2006) study of involvement regarding different kinds of pork meat can possibly be used as a starting point when analyzing involvement with different types of beverages. 2.4 Product Involvement and MEC Product involvement is conceptualized as the motivation to process information by consumers (Celsi & Olson 1988; Zaichkowski, 1985). The concept of involvement is used by researchers and marketers to understand subjective importance of purchase to the consumers (Grunert, 2000). The degree of involvement has an impact on choice process and evaluation of alternatives. With high involvement the consumers tend to use many product characteristics/attributes when making a choice. With a low degree of involvement the choice will be based on choice heuristics (Grunert, 2000). Heuristics are choice tactics consumers use in order to reduce the effort in making judgments (Hoyer & Macinnis, 2007). There are various studies concerning involvement and its relation to MEC (Claeys Swinnen & Vanden Abeele, 1995; Celsi & Olson, 1988; Reynolds & Gutman, 1988, Lind, 2006). In order to relate the concept of involvement with overall subject of this dissertation the departure point can be taken in the definition of product involvement made by Peter, _____________________________________________________________________________ Page 24 of 72 Nino Jakupovic NJ64041 ---------------------------------------------------------------------------------------------------------------------------------------------------------- Olson & Grunert (1999): “Involvement refers to consumers' perceptions of importance or personal relevance for an object, event, or activity.” Furthermore, Celsi & Olson (1988) are offering a similar view on involvement: “a consumer level of involvement with an object, situation or action is determined by the degree of which he/she perceives that concept to be personally relevant” Previously depicted picture of MEC (Figure 3) can describe product involvement by looking at how product knowledge (Attributes, Consequences) is related to the selfknowledge (Values). In other words involvement with a product is higher if there are stronger relations between product knowledge and self knowledge. Walker and Olson (1991) are researching the degree of involvement when shopping for wedding cards vs. thinking of you cards. The wedding cards were more self-relevant to the consumers; hence the level of involvement was higher. Therefore, if a consumption of a specific beverage is considered to be highly self-relevant the involvement in particular beverage will be higher. There are different opinions on how involvement with products can be measured. According to Kapferer & Laurent (1985) involvement can be compared to presence or absence of its major determinants/conditions such as perceived risk, emotional value, sign value and functional importance of a product. In their work they used scales and measured involvement profiles on each determinant. However, when looking at the means-end basis of involvement one can say that the level of product involvement is determined by the type of means end knowledge activated in the situation (Peter, Olson & Grunert, 1999). If a consumer postulates that product attributes are strongly linked with the personal values he/she will feel higher level of involvement. On the other hand if product attributes only are weakly linked to the important values it will present lower level of involvement (Peter,Olson & Grunert, 1999). _____________________________________________________________________________ Page 25 of 72 Nino Jakupovic NJ64041 ---------------------------------------------------------------------------------------------------------------------------------------------------------- Claeys, Swinnen and Vanden Abeele (1995) as well as Lind (2006) are making distinction between “think” and “feel” products. “Think” products are chosen due to its functional value. In MEC approach, the functional consequences are considered to be central for these products (Lind, 2006). “Feel” products are more connected to the enhancing of self and expressing values which corresponds to the psychological consequences and end values. Chains for “think” product will stop at the lower level of abstraction and have higher ratings for levels related to attributes and functional consequences (Lind, 2006). Feel products, on the other hand will likely reach psychological consequences and value level (Claeys Swinnen & Vanden Abeele, 1995). Costa et al. (2004) and Bech-Larsen et al. (1997) consider food products to be low involvement products. As mentioned earlier this dissertation is considering similarity between food and beverages as products and therefore assumes that beverages exhibit characteristics of low involvement as well as ‘think’ products and they will generate chains at lower levels of abstraction. However, one can presume that some beverage categories can have more of ‘feel’ characteristics. This study therefore investigates if consumers MEC for different types of beverages will be different in terms of “think” and “feel” characteristics. Chapter 3 3.0 Research methods and data gathering of MEC There are different techniques of gathering data when conducting studies using MEC approach. The techniques can be divided into more “soft” techniques such as laddering by personal interviews, and more “hard” quantitative techniques such as laddering by paper and pencil and Association Pattern Technique (APT) (Reynolds & Phillips, 2009). The following chapter will present both qualitative and quantitative data gathering techniques to enable the author to choose the technique best suited for research design. _____________________________________________________________________________ Page 26 of 72 Nino Jakupovic NJ64041 ---------------------------------------------------------------------------------------------------------------------------------------------------------- 3.1 Qualitative and quantitative data gathering techniques 3.1.1 Laddering by personal interviews The laddering technique by personal interview has become the most commonly used technique in relation with MEC measurements and it is sometimes interchangeably mentioned with MEC approach (Reynolds & Gutman, 1988, Rusell et al. 2004). The method is ‘one to one’ semi - structured dialog that elicits the MEC’s of Attributes, Consequences and Values associated with a brand or product (Reynolds, Dethloff, Westberg, 2001). The fact that method is semi structural differentiates laddering from some other qualitative techniques. Laddering is structured in a way to reveal more abstract but personally important reasons for product choice and personal benefits of usage. 3.1.2 Attribute elicitation and laddering interview The two major parts of the laddering technique by personal interview are: Attribute elicitation and One on one Interview The attribute elicitation is a starting point of every laddering interview (Reynolds and Gutman 1988). It provides interviewers with an opportunity to identify key choice criteria that consumers use when choosing among different brands or products. Thereafter it gives interviewer an opportunity to learn why those criteria are important, salient or relevant to the consumer (Reynolds & Olson, 2001). There are many attribute elicitation techniques proposed in the literature. Bech- Larsen, Nielsen Grunert & Sørensen (1997) identified the five most commonly used attribute elicitation techniques for low involvement products: 1. Triadic sorting – A technique used with the purpose of mapping cognitive structures (Bech –Larsen, Nielsen, Grunert & Sørensen, 1997, Kelly 1955). In _____________________________________________________________________________ Page 27 of 72 Nino Jakupovic NJ64041 ---------------------------------------------------------------------------------------------------------------------------------------------------------- this procedure three products are presented to respondents. Respondents are asked to mention important attributes on which two products are similar and at the same time differ from the third. Bech-Larsen et al. (1997) describe the method as complex and time consuming and question if the results obtained can justify the time and complexity of the method. 2. Direct elicitation – A technique where respondents are asked to freely name the attributes which are important to them when choosing between products presented to them (Bech –Larsen et al. 1997). This technique does not limit the respondent and attributes mentioned can be quite broad. It is often suggested when conducting explorative research. 3. Free sorting - The respondents are freely placing the products presented to them in groups of either similar products in some important aspect and different from products in the other groups. Afterwards, respondents name the perceived differences between the groups. This technique is suggested by Bech-Larsen for investigations of cognitive structures. 4. Ranking - Informants are asked to prioritize products presented to them and state what the reasons behind their ranking were. 5. Picking from the attribute list - In this technique the informants are asked to decide on attributes from already made list. This list could be gathered from focus groups or other previous research (Bech –Larsen et al. 1997). There are no theoretical criteria on which to base elicitation technique choice, and there is no conclusive research if the specific elicitation technique affects the resulting set of attributes (Beckmann, Sørensen, 2001). However, Bech-Larsen et al. (2001) reviewed this issue and have made some recommendations. Consequently, the main recommendation is to base the elicitation technique choice on the study purpose. When one of the elicitation techniques is chosen and conducted the laddering interview starts by approaching the interview from the attributes elicited by the consumer and _____________________________________________________________________________ Page 28 of 72 Nino Jakupovic NJ64041 ---------------------------------------------------------------------------------------------------------------------------------------------------------- asking a series of questions similar to: “Why is that (Attribute) important to you?” One has to have in mind is situational specificity related to the purpose of the study. Is it cognitive structure in specific situation or more broad usage situation? If one is interested in specific usage situation the respondent has to be aware of this during entire interview. If it is more broad consumption situation the interviewer has to cover different situations (Grunert et al. 2001). The shape and number of the questions asked by the interviewer depends also on how verbalized the respondents answers are (Reynolds, Dethloff, Westberg, 2001). Less in depth answers from the respondents requires more probes by the interviewer. The interviewer constructs each subsequent question on the basis of the answers gained to previous question and moves “up” the ladder of abstractness. The laddering process continues in this manner until respondent cannot provide more meaningful information. Mental block by respondents is the problem often encountered by interviewers preventing them to move from one level to another and exposing the interviewer to dangers of becoming biased and giving the answers to the respondent (Reynolds, Dethloff, Westberg, 2001). Interviewer can either remain silent, to allow respondent more time to think about the answer, or skip a question and come back to it later. Furthermore, one question may generate forked answers that result in more cognitive categories retrieved for one question. The answers should all be recorded and laddering technique continued from each one of them (Grunert & Grunert, 1995). Reynolds and Gutman (1988) provided further advice on how to successfully conduct laddering interviews. The first advice is to create the appropriate interview environment so informant does not feel in any way “threatened” by the interviewer. This will encourage respondents to look deeper inside them in order to find underlying motives for purchase of product class in question. Furthermore, it should be clear that respondent is regarded as the expert in the interview and interviewer as neutral party. _____________________________________________________________________________ Page 29 of 72 Nino Jakupovic NJ64041 ---------------------------------------------------------------------------------------------------------------------------------------------------------- 3.1.3 Quantitative data gathering techniques Recently, more quantitative “hard “methods of gathering MEC data have emerged as an alternative to ‘soft’ laddering. The main characteristic of hard laddering methods is using either paper and pencil or computerized questionnaires (Walker & Olson, 1991; Hofstede et al. 1998, Botschen & Hemetsberger 1998). Walker and Olson (1991) used paper and pencil questionnaires in their study on how two different decision situations affect consumer generated MECs. The respondents were required to identify attributes they attach to selecting greeting cards and then to complete a questionnaire which linked those attributes to consequences and goals. The respondents wrote their answers in boxes preceded by the statement of: this is important to me because… However respondents were not permitted to fork their answers. The study by Mulvey, Celsi and Walker (1994), used the same method to explore differences in means-end knowledge structures for consumers with different levels of involvement. Hofstede, Audenaert, Steenkamp & Wedel (1998) suggested Association Pattern Techniques (APT) as newest alternative to laddering. This technique in contrast to ‘soft’ laddering technique resembles the quantitative surveys where informants check boxes on the readymade lists of Values, Consequences and Attributes (Phillips & Reynolds, 2009; Van Rekom & Wieranga, 2007). 3.1.4 “Hard” vs. “Soft” laddering Both hard and soft MEC data gathering techniques have distinct advantages and disadvantages. The major disadvantages of traditional “soft” laddering technique are time consumption, relatively high costs of interviews and finding the right interviewers to complete the task (Phillips & Reynolds, 2009, Botschen & Hemetsberger, 1998). On the other hand the ‘soft’ laddering interviews have major advantages in possibility to elicit more concrete attributes, possibility to bypass forked answers, and possibility to “help” respondents by using previously mentioned advices from Reynolds and Guttman (1988). _____________________________________________________________________________ Page 30 of 72 Nino Jakupovic NJ64041 ---------------------------------------------------------------------------------------------------------------------------------------------------------- The advantages of the more quantitative approach are lower the costs of conducting research, saving time for the interviewer, no need for transcribing or coding data (Grunert et al. 2001). Furthermore, interviewer is not attending the interviews, larger sample sizes are easier to cover and interviewer bias is eliminated (Botschen & Hemetsberger, 1998; Grunert & Grunert, 1995). However, as stated by Phillips and Reynolds (2009) it is debatable if consumers would reach the higher levels of abstraction on their own, by using the ’paper and pencil’ approach and without probing questions by the interviewer. Moreover, it can be assumed that by just checking the boxes from readymade lists like in APT the informants will limit their cognitive process and just use tick boxes with answers they recognize and would not use something they may recall. Gengler and Reynolds (1995) believe that hard laddering interferes in one of the core assumptions of the method, which is to enable the interface with the answers created spontaneously by the respondent. There is still no research that clearly suggests which technique is preferable to use or study which definitely tests validity of results from both techniques (Valudo de Oliviera et al., 2006). Russell , Busson et al. (2004) as well as Reynolds and Phillips (2009) have researched this issue, however no firm recommendation on which technique to use was made yet. Grunert and Grunert (1995) assert that the soft approach is potentially better when types of problems, on the respondent’s side, are caused by very weak or very elaborate cognitive structures. In fact, they believe that when a respondent’s level of knowledge about a product is too low or too high, the interviewer should prefer the soft laddering method because there are more chances to conduct an appropriate interview. 3.2 Analysis of laddering data The data one has gathered is raw and rather complex, and responses can be distinctive from one respondent to another (Gutman, 1984). In order to interpret and present the complex data obtained from the laddering, a content analysis has to be made first. Content analysis groups large numbers of personal meanings into limited number of common categories under the levels of Attributes, Consequences and Values coding _____________________________________________________________________________ Page 31 of 72 Nino Jakupovic NJ64041 ---------------------------------------------------------------------------------------------------------------------------------------------------------- them individually. There can be different formulations, with the same meaning mentioned during the interviews and they should be grouped in the same category (i.e sparkling taste or tingling feeling can both fit into attribute level - carbonization). Furthermore, one has to bear in mind that categories have to be broad enough to include more than one respondent making a statement that fits the category but also narrow enough prevent a risk of lost of meaning (Reynolds, Gutman, 1988). This provides a rather subjective aspect to the laddering, thus it is subject to the interviewer to make the grouping. The content analysis serves as a background for the second step of the analysis which is constructing an implication matrix that shows the number of associations between different concepts in the interview. This could be in direct connection A – C, or in indirect connection A – V. The sum of different connections indicates the strength of a given association. Implication matrix creates basis for constructing HVM. Marketers can use HVMs to generate product, for advertising and to apply segmentation strategies (Lin, 2009). According to Reynolds & Gengler (1995) the finished HVM must include all associations from raw data over a chosen threshold level (cut-off level). The cut off level is defined by Skytte & Bove (2004) as …”the minimum number of times two elements must have an indirect link to be mentioned in the hierarchical maps”. However, the direct links should be added to this definition. The determination of the cut off level is also subjective decision by the analyst who is constructing HVM, as there is no theoretical guideline concerning the determination of the cut off level (Grunert & Grunert, 1995). However Reynolds and Gengler (1995) assert that minimum threshold of associations shown in the HVM should never be lower than 70% and typically in the range of 75% to 85%, otherwise valuable information can be lost. On the other hand Skytte and Bove (2004) recommend a cut-off level that will represent between 50-70% of the relations. In most studies the cut off level is from 1-5 depending on the number of respondents and number of categories identified. _____________________________________________________________________________ Page 32 of 72 Nino Jakupovic NJ64041 ---------------------------------------------------------------------------------------------------------------------------------------------------------- According to Grunert, Grunert & Sørensen (1995) the obstacle of HVM is the question of redundancy between linkages. If there is a linkage between categories at abstraction levels A – V - C than linkage between A - V would be redundant, and could lead to misleading characteristics of HVM. However, if there is homogeneity among respondents Grunert et al. (2001) argue that redundant HVM’s should be allowed. 3.3 Why using MEC approach and laddering for this study The theoretical frame of this dissertation identified different advantages and limitations of MEC approach to deal with study objectives. However, it also anticipated to show the possibility of applying MEC approach to complete these objectives. The following reasons justify the use of MEC to conduct a consumer study in BIH: It is believed that MEC approach is a very useful method to model cognitive structure of consumers in BIH when choosing beverages. Previous studies have shown that MEC is applicable framework for similar studies. Furthermore, theoretical approach has also shown that every variable identified in the problem formulation either has been and/or can be analyzed using MEC approach. In addition, another valuable component of MEC is its commodity of allowing the respondent(s) to use own frame of reference to provide reasons for choosing specific beverage. In accordance with the theoretical frame of the dissertation provided as the background for conducting a consumer study in BIH, the author will analyze the objectives proposed in Chapter 1.1 in the following way: Is the choice of a different beverage category linked to different personal consequences and values of consumer? This question/objective will be analyzed departing from Gutman (1984), and by using data gathered from consumers to construct MEC’s for each beverage category as a unit _____________________________________________________________________________ Page 33 of 72 Nino Jakupovic NJ64041 ---------------------------------------------------------------------------------------------------------------------------------------------------------- of analysis. These MEC’s are hereafter analyzed and HVM’s are created for each unit of analysis. This will provide the estimate of cognitive structure model of consumers in BIH concerning beverages. What degree of product involvement describes the choice of a different beverage category? Data gathered from consumer study and Lind’s (2006) theoretical background will be used to analyze this particular objective and investigate how consumers in BiH are involved in consumption of different beverages. The level of involvement will be presented by means of number of ladders elicited by the respondents and the complexity of the HVMs. Furthermore, the quantity of each of those concepts will be used to investigate whether it is a “think” or a “feel” product. How can this knowledge be utilized by a beverage producer when advertising their products in the future? This objective will be analyzed by using data from the consumer study, applying it into the MECCAS model. The results of the laddering interviews represented by HVM’s, will provide the opportunity to select a MEC to create suggestions for future marketing strategies by beverage producers. The suggestions will concentrate on creating a message in the mind of the target group, for a hypothetical beverage, for each beverage category. HVM’s also provide the opportunity to explore whether the present marketing activities correspond with the results gathered by the consumer study. _____________________________________________________________________________ Page 34 of 72 Nino Jakupovic NJ64041 ---------------------------------------------------------------------------------------------------------------------------------------------------------- Chapter 4 4.0 Research design The research design was developed using parts of the theory and the description of MEC methods from previous chapters. 4.1 Research method In line with paradigm, established in Chapter 1.3, qualitative research design will be applied to gather data to establish consumer reasons to choose a particular beverage. Qualitative research will enable both researcher and the respondents to put more “words” into the subject which would be made rather difficult if a quantitative method (i.e. questionnaires) was used. 4.2 Selection of data gathering technique There is no conclusive evidence on preferable technique (Chapter 3.1.4) for data collation. The laddering interviewing techniques was used as previously described and the choice was based on following reason. Author has an assumption that using paper and pencil or ATP technique in BiH combined with consumer(s) lack of experience with these types of studies would not be efficient. It would encourage respondents to get through the survey ‘easily’ and without using their cognitive processes, and therefore will provide less real picture of their cognitive structure. 4.3 Selection of respondents When conducting a study applying MEC and laddering it is preferable to establish some criteria for selection of the respondents. It is important to obtain a certain degree of _____________________________________________________________________________ Page 35 of 72 Nino Jakupovic NJ64041 ---------------------------------------------------------------------------------------------------------------------------------------------------------- homogeneity of respondents, in order for HVM to reveal cognitive differences that are not related to socio-demographic profiles of respondents (Grunert & Grunert, 1995), and respondents should provide most information on the subject (Reynolds & Olson, 2001). These considerations let to inclusion of the following respondents: Both male and female between 18 – 30 years of age (students and young people in BiH), Young adults are considered to be trendsetters in the area of beverage consumption. Most beverage products are marketed towards young people (e.g. Coca-Cola, Evian etc). Young people are also a group which is most likely to be consuming all four beverage categories on daily basis. Furthermore, this group is considered to be more open minded and, current in their cognitive structure. Younger generation would be harder to interview due to lack of focus during the interview, and many of them are still guided by their parents when choosing beverages. This generation might not be user of all beverage categories (e.g. coffee). An older generation is more influenced by joint decisions of a family in terms of purchase and this generation might also not consume all beverage categories (e.g. carbonated soft drinks). The respondents were chosen from either Bosnian capital, Sarajevo or the second largest town in BiH, Mostar. The two different cities are chose to elicit more diversity between participants (big-smaller city). All potential respondents are daily consumers of chosen beverage. They were approached and asked if they are interested to participate in the consumer study. The primary data for this research was gathered outside Denmark. The main problem the author encountered was time and resource constraints to enable him to travel to BiH _____________________________________________________________________________ Page 36 of 72 Nino Jakupovic NJ64041 ---------------------------------------------------------------------------------------------------------------------------------------------------------- to ensure that enough respondents are recruited to create some type of tendencies in the research. Reynolds and Olson (2001) use the rule of thumb and consider 20 people to be the minimum size for any subgroup investigated. One subgroup of respondents between the ages of 18 – 30 was included in this research. However, using 20 people seemed as a rather small sample size. The decision was made that 30 interviews will be conducted as a minimum sample size. Finally, one has to consider language variation. Bosnian language is researcher’s native language and no problems were encountered while conducting interviews. Moreover, raw data was gathered in Bosnian and most of data processing is in Bosnian language. However, the coding and structuring of HVM’s is conducted in English. 4.3 Elicitation technique The ranking technique was chosen as the elicitation technique. This choice was based on the limited number of categories respondents have to choose from and on the assumption that this technique will be best suited as the starting point for the interview. Free sorting technique, frequently used for uncovering cognitive structures, or list picking (the quickest technique) (Bech-Larsen, 2001), were also strongly considered and could have been efficiently used. However, there was no previous attribute list made for this study from the focus group so the list could not have been used. The ranking technique was also chosen to enable the author to establish which beverages will be ranked at the top and provide a picture of peoples´ choice. Furthermore, there are not many categories for respondents to choose from and free sorting technique would not be optimally utilized. Additionally, according to Costa et.al. (2003), ranking technique makes a good balance between concrete and abstract attributes gathered as well as relevance of the attributes gathered is considered to be high. Moreover, this technique is also considered to have a good predictive ability which is well in line with previous discussion about cognitive view of MEC from Chapter 2. _____________________________________________________________________________ Page 37 of 72 Nino Jakupovic NJ64041 ---------------------------------------------------------------------------------------------------------------------------------------------------------- 4.4 Pilot research A small pilot study was conducted by the author prior to doing the main research for the study. The pilot study enabled the interviewer to change and improve various aspects of the interview design or technique to make it more appropriate. Three young individuals, independent from either similar studies or beverage industry were initially interviewed. The informants objected to using a voice recorder during the interview. They deemed this as ‘material’ likely to be misused. Therefore, in the main interview it will be optional and left to the informants to approve the use of recorder. If particular respondent does not approve usage of voice recorder, the interviewer will note the answers in hand. The elicitation technique was also improved by making the author realize the importance of focusing on the situational context. The answers about choice of different beverages can considerably vary if the consumption situation context is not clearly explained to the respondent, very much in line with Grunert et al. (2001). 4.5 The interviews Data was collected from individual interviews lasting approximately thirty minutes per person. All interviews were conducted in the surroundings chosen by respondent. That is being their flat, dorm room, and café bar. The respondents were first explained the purpose of the research and especially focusing on the fact that there were no right or wrong answers to any of the questions as the interviewer is only interested in the opinions of the participants. Respondents were also assured that the interviewer has no affiliation to any of the producers or brands which might be mentioned during the course of the interview. This approach was taken to relax the participants. Each interview started with ranking task. The four groups of different beverage categories were presented to participants. The participants were asked to rank each _____________________________________________________________________________ Page 38 of 72 Nino Jakupovic NJ64041 ---------------------------------------------------------------------------------------------------------------------------------------------------------- beverage category in order of preference in their spare time from 4 to 1: with 4 being the most likely choice and 1 the least likely choice. The preferences were later related to the attributes of chosen product category. This method is inspired by suggestion made by Grunert, Grunert & Sørensen (1995). Once the participant gave their reason(s) (attributes), they were asked to elaborate further on and to state the importance of the reasons given. Each answer that followed was also further elaborated, continuing the laddering process until the participant had no more information to give. One of the respondents chose coffee as the most preferred beverage and the following is a sample interview with male participant aged 26: Interviewer: You said that coffee is the most preferred beverage you choose in your spare time. Why is that? Respondent: Well, I usually have some spare time after a long day of school and coffee is then one thing I really need. Interviewer: Why is it important for you to have the coffee after school? Respondent: I get really tired than I need something to give me a kick for the rest of the day. Interviewer: What do you mean by getting a kick? Respondent: I need something as a stimulant to do things at home. Well, coffee has caffeine and I believe that when I have coffee I get rested and clear in my mind. It energizes me to do more house work, to study or whatever… Interviewer: Why is this energy boost important to you as a person? Respondent: I am a rather active person I like to get things done. Then I feel I have accomplished something and my day was not wasted. Otherwise I would sit on the couch and do nothing. I like to feel that when a day is gone I have done something I am happy about. The day was not wasted. The answers gathered from the interview were later categorized into Attributes Consequences and Values as illustrated below: Accomplishment (V) ↑ Can do more (Pc) ↑ Energizing (Fc) ↑ Caffeine/Stimulant (A) Figure 5: Example of categorizing the results of the laddering interviews Source: Devised by the author _____________________________________________________________________________ Page 39 of 72 Nino Jakupovic NJ64041 ---------------------------------------------------------------------------------------------------------------------------------------------------------- The second choice of beverages was also examined and followed by the rest of choices. The length of interviews depended on the time needed for respondent to answer the questions, non consumption, forked answers etc. In addition to recording answer on the pre-prepared answering sheets the interviewer also recorded interviews using his personal computer, and only if agreed by the respondent. Only 2/30 participants didn’t consent to their interviews being recorded. An example of pre-prepared answering sheet can be seen in Appendix nr. 1. Furthermore some background data of the participants have been gathered, regarding age and gender. The complete interview guide can be seen in Appendix nr.2. _____________________________________________________________________________ Page 40 of 72 Nino Jakupovic NJ64041 ---------------------------------------------------------------------------------------------------------------------------------------------------------- Chapter 5 5.0 Research Findings and discussion The approaches chosen to discuss the results were guided with questions established in Chapter 1. The analysis of the results from interviews for each beverage category is followed by the discussion of the involvement and whether the product should be regarded as “think” or “feel” product. This discussion provides the grounds for conducting recommendations for future marketing strategies for these categories by using MECCAS. 5.1 Participants A total of 30 participants (n=30) was interviewed. 17 Female (56%) and 13 Male (44%) respondents participated in research. The mean age of the respondents was 24,2 (Range: 18-30). The results of the ranking tasks can be seen in the following table. In the appendix 3 a more detailed table is shown regarding ranking of each beverage by each participant. Furthermore, in one can see each beverage category divided by the preference. The table shows how many times each beverage category was selected as first, second, third or fourth choice, by the respondents. _____________________________________________________________________________ Page 41 of 72 Nino Jakupovic NJ64041 ---------------------------------------------------------------------------------------------------------------------------------------------------------- Table 2: Results of the ranking task conducted as the attribute elicitation technique. Total Beverages 1st choice 2nd choice 3rd choice 4th choice freq.points Coffee 14 5 7 4 89 Fruit juice 4 10 12 4 74 Bottled water 10 8 8 4 84 Carbonated soft drinks 2 7 3 18 53 Source: Laddering interviews in BiH. Devised by author The above ranking results show that the highest number of ranking points 89/300 (29,6 %) was allocated for coffee. Coffee is the most likely beverage category to be chosen by consumers interviewed in BiH, it is closely followed by water with the score of 84, and fruit juice 74 and carbonated soft drinks 53. The main reason for choosing each category was further elaborated through laddering interviews. Previously mentioned context of consumption played a large role in the ranking part of all interviews. The respondents clearly needed to put the rankings in the context as they would often mention that ranking of the beverages will be different in another context (i.e. choosing a beverage while at work/school could differ from choosing one at home). This can be illustrated through extract from the interview with a participant: Interviewer: How would you rank frequency of consuming these 4 beverage categories during your spare time? Respondent: Well, if by free time you mean when I am at home then coffee would be my first choice. However, when I am at school I could go for bottled water first as its easier to carry it with me in the classroom or maybe ½ liter of cola. Interviewer: No, I am firstly interested in your spare time. Your time off school, while relaxing at home. Respondent: In that case coffee would be my first choice. Yes, coffee. From this point the respondent would concentrate on the given context and laddering would continue. _____________________________________________________________________________ Page 42 of 72 Nino Jakupovic NJ64041 ---------------------------------------------------------------------------------------------------------------------------------------------------------- 5.2 Content analysis Content analysis is used to classify raw data into a limited number of categories. To be able to carry out this classification, 28/30 interviews were recorded on the computer. Two participants chose not to be recorded and their answers were noted. A definition of the means, the ends and their relations, was made and content analysis of all concepts was conducted from the recording. Categories were developed to capture the main thoughts expressed by the informants on an Attribute, a Consequence and a Value level in a coding table. Coding tables can be seen in Appendix nr.4-7. Coding tables were designed for each of the following beverages: 1. Coffee 2. Fruit juice 3. Bottled water 4. Carbonated soft drinks All these areas serve as units of analysis and will be treated separately in upcoming subchapters. The relationships of A-V-C for each category were analyzed using software called MEC Analyst developed by SkyMax. The answers from each respondent were typed into the program and program secured the development of implication matrix. Implication matrix for each unit of analysis can be viewed in appendix 8-11 respectively. Consequently HVM’s were developed for each unit of analysis as seen in sections below. The lines between concepts in the presented HVM’s represent the complexity of the HVM and the thickness of the lines represents the number of respondents mentioning the link, directly or indirectly, or the strength of the link. _____________________________________________________________________________ Page 43 of 72 Nino Jakupovic NJ64041 ---------------------------------------------------------------------------------------------------------------------------------------------------------- 5.3 Coffee Coffee has a highest ranking/preference of all units analyzed in the research. The HVM illustrated below in Figure 6 represents the most common and dominant relations and categories between the MECs when it comes to the coffee. The coffee answers elicited the cut off level of four as the most appropriate level to provide most generalizations. The cut off level of four represents that a connection between two concepts was mentioned directly or indirectly by at least four respondents during the interviews. This cut off level has excluded many weaker links between the categories that occur when cut off level of three is chosen. Furthermore, this cut off level is giving the most clear and readable HVM picture. The first ladder on the left of the HVM shows that respondents are choosing coffee due to the concrete attribute of caffeine which energizes people and have them ready to cope with the rest of the day. Similar attribute of coffee is a more abstract attribute of stimulant which enables respondents to do more in their spare time (e.g. homework, housework). This in turn leads to the value of accomplishment, which includes doing more at home, having better results at school and so on. The results of this chain are rather similar to Gutman´s study (1984) with exception of value of self-esteem which was not mentioned during this research. However, it can be discussed if accomplishment in turn would lead to better/higher self esteem. The other part of Figure 6 represents rather social reasons for the choice of coffee as the spare time beverage. For a considerable number of respondents having coffee means the end of working day and the beginning of the leisure time. Here you can have your friends over, be social with them, talk about your day and fulfill a value of belonging to a group and be affiliated with people or a group. _____________________________________________________________________________ Page 44 of 72 Nino Jakupovic NJ64041 ---------------------------------------------------------------------------------------------------------------------------------------------------------- The traditionally prepared Bosnian coffee3, is an abstract attribute respondents are rather likely to link with consequence of performing a ritual that they have inherited from previous generations (parents, grandparents). By performing this ritual the respondents are fulfilling or preserving the traditions of the country. As many respondents answered it is very important for them to keep this part of the culture sacred and unique for the country. This part of Bosnian culture is important to the respondents and they want to have it separated from the European integration and world globalization. The last part of the HVM represents attribute category of taste. This category includes different attributes such as sweetness, sourness, or strong taste. The taste of coffee leads to the consequence of fulfilling the habit of having a coffee on daily basis. On the other hand there is a strong link mentioned regarding complementarities between coffee and cigarettes which in the end helps respondents fulfill the value of being happy and content. This can be explained by considering a rather large number of smokers in BiH in part due to rather low price of cigarettes4 and relatively low influence of anti smoking lobbies compared to Western Europe, Nevertheless, coffee and cigarettes are still regarded as complementary products in BiH. This research also shows the strong link between the two products. 3 Bosnian coffee is rather similar to the Turkish coffee; it is mostly homemade but can be bought at the cafés. 4 Price of a package of cigarettes is approx. 1EUR Source: http://www.fds.ba/loc/ _____________________________________________________________________________ Page 45 of 72 Nino Jakupovic NJ64041 ---------------------------------------------------------------------------------------------------------------------------------------------------------- CutOff=4 Accomplishment Happy/Content Tradition Affiliation/belonging Can do more Habit Be social Relaxing Ritual Energizing Good with cigarettes Caffeine Colour Stimulant Traditional (locally prepared) Taste Abstraction level Attributes Functional consequences Psychological consequences Values Figure 6 Hierarchical value map for coffee Source: Mecanalyst result from laddering interviews 5.3.1 Involvement with coffee The analysis of involvement will begin from Lind (2006), looking at the number and complexity of the chains elicited from the respondents and showed in the HVM’s. _____________________________________________________________________________ Page 46 of 72 Nino Jakupovic NJ64041 ---------------------------------------------------------------------------------------------------------------------------------------------------------- The total number of chains gathered from coffee interview is 70 and the mean, number of ladders is 2,33 per respondent. The highest number of ladders for a single respondent is 4, and the lowest number is only a single ladder. Both 4 ladders and 1 ladder were elicited from single respondents. HVM for coffee shows that ladders for coffee are reaching higher abstraction levels of psychological consequences and values. However, according to Grunert, Sørensen, and Nielsen (1996) the length of chains is not significantly correlated with involvement. On the other hand Gengler, Klenosky and Mulvey (1995) claim that complexity of HVM regarding interconnections between levels of abstraction depicts the level of involvement. The HVM for coffee can be described as rather complex, with different interconnections between the components of knowledge. The thickness of the lines in the HVM shows that there are many strong connections between product knowledge (Attributes, Functional Consequences) and self knowledge (Psychological consequences, Values) indicating higher level of involvement for coffee than for other products in the study, Therefore, it can be concluded that HVM for coffee represents higher level of involvement. Regarding “think” and “feel” aspect of coffee as a product the discussion is also based on the Lind (2006) as well as Claeys, Swinnen, Vanden Abeele (1995). The HVM indicates that coffee has higher level of abstraction and consequently is not only consumed for its functional value but also for its symbolic value. Furthermore, 59,7 % of ladders were complete, having direct connection between attribute and value which also suggests more symbolic reasons for consuming coffee. Finally, the relatively high number of psychological consequences and values also suggest that coffee can be treated as a product of choice with more symbolic value, or a “feel” product. _____________________________________________________________________________ Page 47 of 72 Nino Jakupovic NJ64041 ---------------------------------------------------------------------------------------------------------------------------------------------------------- 5.3.2 MECCAS for coffee Coffee is a product usually marketed as something to start the day with or connected with the concrete attributes such as smell or taste of coffee. The market leader for coffee in BiH Grand kafa (www.grandkafa.com/en/kampanje), advertise coffee using print and media by focusing on the sensual experiences of coffee drinking. Even the slogan of the company is targeted in that direction called “both taste and smell”. Figure 6 illustrates that both marketing strategies are indeed included in the cognitive structure of consumers in BiH. The ladder of Taste → Habit → Happy, corresponds to mostly used coffee marketing campaigns while ladder Stimulant → Can do more → Accomplishment is also present in the area and rather usual way of marketing coffee. Grand kafa is following this ladder by portraying young business people tired at their office until tasting Grand coffee (www.grandkafa.com). If MECCAS for a hypothetical coffee brand should be developed, a new kind of message should be provided in order to achieve differentiation compared to the existing competitors. One strong ladder in the HVM is in favor of coffee/cigarette compatibility. However, this cannot be used in the modern world due to the hazards of smoking. Therefore, alternative strong ladder in the HVM could be used Traditional → Ritual → Tradition. The suggestion for application of this ladder in the MECCAS framework can be as follows: Message elements (The specific attributes. consequences or features about the product that are communicated verbally or visually (Gutman, 1984)5: Coffee that is meant to be traditionally prepared. Bosnian way of making coffee. 5 All definitions for MECCAS elements are from Gutman (1984) _____________________________________________________________________________ Page 48 of 72 Nino Jakupovic NJ64041 ---------------------------------------------------------------------------------------------------------------------------------------------------------- Consumer benefit: (The major positive consequences for the consumer that are to be explicitly communicated, verbally or visually, in the advertising): The message for coffee has to contain visuals of people being together performing a ritual of drinking coffee, relaxing and talking. Making deeper friendships and family ties (i.e. using motives of traditional Bosnian family). Leverage point: (The way in which value is linked to the specific features in advertising) The focus here can possibly be on staying true to the value of tradition in the ever changing environment. Driving force (The end level to be focused on in the advertising): In this part the driving force of advertising should be to keep the traditions alive through consummation of coffee. 5.4 Fruit juice Fruit juice is mentioned as the third choice by the respondents in the interviews. Fruit juice has gathered 74 points on the frequency of choice scale (Table 2), and it is mostly chosen as a second choice with ten respondents choosing it as their second beverage choice (Table 3). Figure 7 shows the fruit juice HVM that illustrates the connection between fruit juice being the second choice, and the cognitive structure of the respondents. The first two attributes shown on the far left of the HVM describe fruit juice as a substitute for other beverages in this research. When respondents substitute water for fruit juice it is because they want a change in the form of a more flavor and specific taste compared to the neutral taste of water. _____________________________________________________________________________ Page 49 of 72 Nino Jakupovic NJ64041 ---------------------------------------------------------------------------------------------------------------------------------------------------------- Furthermore, there is a strong connection between using fruit juice as a coffee alternative, strongly leading to the similar consequence that coffee produce and that is being social. This is possibly related to the limited daily amount of coffee one can intake and fruit juice is percept as a substitute that can be consumed with the same consequence. The interesting detail is that no other beverage except fruit juice had an alternative/substitute attribute strong enough to be represented in the HVM. Moreover, the important attribute of fruit juice is to be cold. Cold fruit juice makes people feel less warm, and they feel that they can do more during the day. The attribute of coldness is connected to BiH being warm southern country, and possibly the warm weather when interviews were taken. It would be interesting to see whether attribute cold would be mentioned if a similar research was to be conducted during winter period. The next chain produced from the interviews regards low or no sugar content of the fruit juice. The low level of sugar is important to the respondents as it keeps the weight down, and in turn leads to better looks. By looking fit the respondents would reach better self-esteem, which is considered as a value. This is well in line with Gutman (1984), who indentified rather similar ladder in his work. The value of self–esteem can be connected to the age group targeted in this study. It is possible that possible future studies which could include different age group would not have this value as strongly as it is the case here. The last attribute identified for fruit juices are the vitamins in the fruit juice. The respondents believe that consuming fruit juice can add or substitute the vitamins from eating fruits and in turn it results in better health. Consequently this is connected to the value of health which is the second value that is identified on this cut off level. _____________________________________________________________________________ Page 50 of 72 Nino Jakupovic NJ64041 ---------------------------------------------------------------------------------------------------------------------------------------------------------- CutOff=4 Self esteem Health Satisfied Can do more Better looks Feel Healthy As filling but has more taste Water substitute Colour Satisfying Can be social Feel better/Less warm Coffee alternative Cold Keeping weight Natural fruits/ no sugar Vitamins Abstraction level Attributes Functional consequences Psychological consequences Values Figure 7: Hierarchical value map for fruit juice Source: MEC analyst results from laddering interviews. Devised by author 5.4.1 Involvement with fruit juice The total number of ladders elicited for juice is 72 or 2,4 by respondent, which is a slightly higher number than coffee. Furthermore, the number of ladders is ranging from the lowest total of 1, by 6,67% of respondents up to 4. However, the ladders elicited are producing a simpler HVM then the ones for coffee. Thus, involvement for fruit juice is _____________________________________________________________________________ Page 51 of 72 Nino Jakupovic NJ64041 ---------------------------------------------------------------------------------------------------------------------------------------------------------- depicted by lower complexity of HVM than coffee. This suggests that fruit juice can be described as low involvement product. In contrast to coffee, the fruit juice laddering interviews provided a larger number of attributes and functional consequences but a lower number of psychological consequences and values. Furthermore, the ladders for fruit juice have a lower number of complete links from attribute to the value level. A total of 12,8% of ladders had a direct link between an attribute and a value, compared to 59,7% for coffee. Having all this in mind one can say that choice of consumption of fruit juice is connected more to the functional values of fruit juice than its symbolic value. Thus, fruit juice can be defined as a think product. 5.4.2 MECCAS for fruit juice The analyses of fruit juice interviews have shown that values of health and self-esteem are the strongest values identified. Most fruit juice competitors are also concentrating on these values. One of the stronger market players in BiH Nectar (www.nectar.rs) is targeting young people and the value of high self-esteem when consuming their juice, because their juice keeps the weight down. The other player Cappy6 is mostly concentrating on having the juice as powerful drink to start the day with vitamins. Having in mind the similarities between results obtained from laddering and, the actual messages sent by players on the market one can assume that either the consumers are very much influenced by the advertising of the or the companies are rather good to target the cognitive structure of the consumers. However, this is out of scope of this dissertation as answer can only be based on pure speculation. It is advisable to perform a new research to clarify this relationship. 6 Cappy is Minute Maid juice using another brand name in BIH _____________________________________________________________________________ Page 52 of 72 Nino Jakupovic NJ64041 ---------------------------------------------------------------------------------------------------------------------------------------------------------- Vitamins →Feel healthy → Health and Natural fruits → Keeping Weight →Better looks → Self-esteem ladders in the single message. Marketing would be more efficient in this way and would target both cognitive structures of consumers. Message elements: The message should include information that there is no sugar in the fruit juice and enough vitamins to fulfill a Recommended Daily Allowance (RDA). Consumer benefit: Get enough fresh fruit and vitamins, keep weight down, and feel good about yourself. Leverage point: The leverage point should be to stay healthy and by drinking the fruit juice one can invest in healthier future by remaining thin. Driving force: The driving force should include healthy and long life. 5.5 Bottled water Bottled water is a beverage category which was picked as the second most frequent choice by the respondents in this research. The accumulated number of points gathered was 84 and 10/30 people regard bottled water as their first choice when choosing non alcoholic beverage in their spare time. The first ladder far to the left gives bottled water the abstract attribute of being the cheap refreshment. As BiH is rich with rivers the production of domestic bottled water is relatively high and therefore the price is low compared to other foreign refreshment. It can be discussed whether this particular attribute influences the ranking method of the beverages. However, there can be no clear answer for this from this research. The cheapness of the water leads to the consequence of having more money to buy other things. Another attribute of bottled water is that it is domestic/local product which is a way to support the employment in the country, which is considered to be one of the issues concerning the country now. _____________________________________________________________________________ Page 53 of 72 Nino Jakupovic NJ64041 ---------------------------------------------------------------------------------------------------------------------------------------------------------- The attribute-consequence ladder of being cold corresponds to the previous discussion concerning fruit juice, where cold water is used to cool the body down in the warm weather. The functional consequence of water being satisfying, as well as feel healthy, do not have direct connection to the attribute. The reason for this can be that they are connected to the different attributes which have not been strong enough to be included at this cut off level. The psychological consequence of feel healthy is connected to the value of health. The last attribute on the right side of the HVM is no sugar no calories attribute which s connected to the consequence of keeping weight and have better self esteem. This ladder is again similar to the fruit juice ladder. It is therefore possible that the choice of fruit juice and bottled water are based on the same cognitive structure. People choose these beverages in order to avoid excessive sugar, keep their weight down and at the end feel better about themselves. _____________________________________________________________________________ Page 54 of 72 Nino Jakupovic NJ64041 ---------------------------------------------------------------------------------------------------------------------------------------------------------- CutOff=4 Health Better selfesteem Accomplishment Can do more Security Feel healthy Keeping weight Cool the body down Have more money for other things Thirst quenching Cheap Colour Higher employment Healthy Domestic/local Basic need Satisfying Cold No sugar/no calories Abstraction level Attributes Functional consequences Psychological consequences Values Figure 8: Hierarchical Value Map for bottled water Source: MEC Analyst – devised by author 5.5.1 Involvement with bottled water The total number of ladders for bottled water is 80 or 2,66 ladders per single respondent. This is slightly more than coffee and fruit juice and also the highest total number of ladders for all four types of beverages. The most ladders elicited by a single respondent is four and the least number elicited is two ladders. _____________________________________________________________________________ Page 55 of 72 Nino Jakupovic NJ64041 ---------------------------------------------------------------------------------------------------------------------------------------------------------- One can argue that bottled water HVM is characterized by simple HVM which indicates a lower degree of involvement than one created for coffee. Even though water has gathered highest number of ladders it can be assumed that bottled water is still a low involvement product due to the simplicity of the ladders. Similar to the fruit juice bottled water elicited more attributes and functional consequences than psychological consequences and values. Furthermore, a larger number of direct connections between attribute and value have been identified (27,5%) than with fruit juice, however a lower number than with coffee. Taking both number of direct A→V links and types of elements into consideration one can say that bottled water is mainly chosen for its functional values and therefore a think product. 5.5.2 MECCAS for bottled water The results of this study suggest that bottled water is a think product; present advertising is predominantly focused on the values of health and self-esteem. ˝Ilidzanski dijamant˝ is one of market leaders for bottled water in BiH. Their marketing campaign contains pictures of beautiful, fit young woman. The campaign is mainly based on fitness, beauty and health as the main massage for their campaign (www.ilidzanskidijamant.ba). ˝Jana˝ is the another competitor on the market. Their marketing efforts concentrate on the purity of water combined with young, fit business people and consuming water as a natural part of their everyday. The campaign slogan is “bring freshness into your work day”. This strategy is well in line with the ladder of Satisfying → Can do more → Accomplishment, elicited from the study. In order to deliver MECCAS for hypothetical water producer in the area, one can choose a route of No sugar/No calories →Keeping weight→ Better self-esteem, which shows rather strong links in the consumers’ cognitive structure. However, similar to the fruit juice, this ladder is already covered by the competitors in the market as discussed _____________________________________________________________________________ Page 56 of 72 Nino Jakupovic NJ64041 ---------------------------------------------------------------------------------------------------------------------------------------------------------- in earlier examples. If any differentiation is to be achieved it could be interesting to concentrate on the domestic attribute of bottled water which was mentioned by 33% of the respondents in the study, and possibly combine it with the ladder of No sugar/No calories →Keeping weight→ Better self-esteem in order to produce a synergy effect. Message elements: The message should include origins of domestic Bosnian water, possibly featuring untouched rivers of the country. Moreover, the message should also include information on how healthy and fit one can become by having water, and promote high self-esteem lifestyle. Consumer benefit: Consuming bottled water will help consumer to stay fit as well as help the country by consuming a healthy domestic product. Leverage point: The leverage point is to invest in the body and invest in the country. Driving force: The end level to be focused on is to improve the self-esteem by drinking water and helping the homeland. 5.6 Carbonated soft drinks Carbonated soft drinks are the least frequently chosen beverage according to the ranking technique. It has gathered 53 points in the list of most frequent choices and only 4/30 (13,3%) respondents choose carbonated soft drinks as their first option when having a refreshment in their spare time. This result was rather surprising having in mind the popularity of carbonated soft drinks in the World. However, one can only guess the origin of this result. It may have occurred due to the “pressure” of the interview on the consumers wanting to present the “more healthy choice” or due to the context of the interview. However, it would be interesting to see whether different result will transpire if “paper and pencil” technique was used. The first ladder in the HVM for carbonated soft drink is similar to many other beverages with having a cold carbonated soft drink in the warm weather is refreshing and helps the body cool down. _____________________________________________________________________________ Page 57 of 72 Nino Jakupovic NJ64041 ---------------------------------------------------------------------------------------------------------------------------------------------------------- The second attribute elicited is unique for the carbonated soft drinks. Some varieties of carbonated soft drinks are considered exciting. Thus, there are many different flavors, brands and shapes of carbonated soft drinks to choose from. However, this attribute could be slightly outside of the frame of this research. The variety can be guided by the occasion. Soda (carbonization) is category used to include the carbonization and tingling feeling of carbonated soft drinks when consumed. This attribute has a more complex connections compared to others in this research. The connections are ranging from negatively driven consequences of being bad for stomach, to more positively laden consequences of soda in the carbonated soft drinks complementing food, and reaching more hedonistic psychological consequence of being satisfied. The other consequence of soda is fulfilling a habit one has developed of consuming carbonated soft drinks on daily basis. The strongest attribute for carbonated soft drinks is sugar. Sugar is associated negatively with the consequences of having the weight and teeth hygiene issues because of percept high content of sugar. The weight issue and bad for teeth consequences both result in lower self-esteem. On the other hand sugar is also negatively connected to the value of health. Comparing this study to Gutman (1984) study one notices that consumers’ cognitive structure differentiates on the perception of carbonated soft drinks. Gutman (1984) claims that carbonated drinks are filling (attribute) which helps people avoid eating other foods thus contributing to weight control, ultimately leading to self confidence and self-esteem. This study, on the other hand, shows that avoiding sugar in carbonated soft drinks helps to control weight issues and in the long run should help the selfesteem. Therefore, it can be said that since Gutmans study up till now the cognitive structure of people towards carbonated soft drinks has changed in more negative manner. The reasons for this could be more focus from the media and corporations on the health and self-esteem issues in the modern world. Having this in mind and by using information processing models described in Chapter 2.1 one can say that the cognitive structure of consumers has shifted in the 25 years _____________________________________________________________________________ Page 58 of 72 Nino Jakupovic NJ64041 ---------------------------------------------------------------------------------------------------------------------------------------------------------- period. It would be interesting to conduct future studies to diversify carbonated soft drinks e.g. include diet drinks, to get a more nuanced picture of carbonated soft drinks. CutOff=4 Health Lower selfesteem Not healthy Satisfying Bad for health Habit Good in warm weather Exciting Bad for teeth Weight issues Bad for stomach Compliments other food Cold Color Variety Soda Sugar Abstraction level Attributes Functional consequences Psychological consequences Values Figure 9: Hierarchical Value map for Carbonated soft drinks Source: MEC Analyst – devised by author 5.6.1 Involvement with carbonated soft drinks One very interesting occurrence with carbonated soft drink is that even though it is the least frequently chosen beverage it has gathered most chains among respondents. The _____________________________________________________________________________ Page 59 of 72 Nino Jakupovic NJ64041 ---------------------------------------------------------------------------------------------------------------------------------------------------------- total number of chains is 86 or 2,86 per respondent. The number of ladders by a single respondent ranges from 0 where a respondent claimed not to have any knowledge or reasons his/her choice up to 6. One can argue that even though respondents do not frequently choose carbonated soft drinks they do have established structures of knowledge regarding them. However, the ladders are rather short and mostly not complete. Combining the number of ladders and complexity of the HVM one can say that involvement for carbonated soft drinks is higher than with water and juice and rather similar to coffee or medium level of involvement. Similar to the fruit juice and bottled water carbonated soft drinks produced more attributes and functional consequences than psychological consequences and values. Furthermore, 12,8 % of the links had a direct connection between attribute and value, which suggests that carbonated soft drinks are regarded as think products. 5.6.2 MECCAS for carbonated soft drinks Market for carbonated soft drinks is mostly similar to the other markets in the World with larger companies like The Coca-Cola Company and PepsiCo and a number of local competitors. Coca-Cola is targeting young adults and teenagers by using ads which present fun, cold refreshments, and enjoying life with Coca-Cola (www.coa-cola.ba). The other producers in the area are using similar tactics to attract young people to their products. A part of this strategy is depicted in the ladder Cold→Good in warm weather ladder or Variety → Exciting ladder. Moreover the competitors are targeting consequences of fun and values of belonging/affiliation which were not strongly present in this study. In order to differentiate a hypothetical carbonated soft drink product one can possibly use another ladder which showed strong links and that is Soda (Carbonization) → Compliments other Food → Satisfying. _____________________________________________________________________________ Page 60 of 72 Nino Jakupovic NJ64041 ---------------------------------------------------------------------------------------------------------------------------------------------------------- Message elements: The message elements should contain tingling feeling of sodium in the drink and satisfactory/inviting look of blend of meal and carbonated soft drink. Possibly, having dinner with family and/or friends. Consumer benefit: It should be visually and verbally communicated the feeling of ultimate hedonistic satisfaction consumer will experience when having carbonated soft drink and a meal. Leverage point: The marketing should concentrate on the satisfaction when having a meal complemented with a drink in a relaxed atmosphere. Driving force: Is to have an ultimate hedonistic pleasure with carbonated soft drink. _____________________________________________________________________________ Page 61 of 72 Nino Jakupovic NJ64041 ---------------------------------------------------------------------------------------------------------------------------------------------------------- Chapter 6 Validity, reliability and generalization of the study Validity in qualitative study corresponds to the question if the study reflects the phenomena of interest to the reader (Kvale, 1997). Validity of this study is concerned with discussions regarding the success and application of the method concerned in representing the choice of beverages in BiH. Moreover, validity is also concerned with truthfulness of what is meant to be investigated. In this study the author assumes that HVMs and later analysis describe the beverage choices by consumers in BiH. The validity of the study could have been improved if there were two persons conducting the interviews and consequently coding the interviews, as suggested by Grunert (1995). This was not possible to achieve due to the form of the dissertation. However, in order to ensure the largest possible degree of validity each informant has been asked to validate his/her answers as well as researchers’ interpretations during the course of each interview. Moreover, one could have formed a focus group with some of the participants after all the interviews were conducted. The focus group would aim to further validate the answers. This was not possible due to time and scope restraints. This can be utilized if and when future studies are conducted. The other question concerned with validity of the study using MEC and laddering is if the interviewing process has allowed the respondents to use their own frame of reference while providing their reasons for choice of different beverages. The author assumes that this is fulfilled to a high degree as Attributes, Consequences and Values have been named with the departing point in respondents own frame of reference. However, the translation of the coding to English could have slightly adjusted some answers as a direct translation was not possible (i.e. psychological consequence of being satisfied with carbonated soft drinks has a special word in Bosnian language which does _____________________________________________________________________________ Page 62 of 72 Nino Jakupovic NJ64041 ---------------------------------------------------------------------------------------------------------------------------------------------------------- not have a direct English translation). The word covers hedonistic and cultural specific satisfaction feeling, but the best translation appeared to be satisfied. Reliability refers to another study under similar circumstances and its ability to produce similar findings (Rekom, Riel et al. 2000). It is difficult to ensure reliability of a qualitative study due to its context and current time of the study. Comparing this study to that of Gutman (1984), one can say that cognitive structure of consumers have changed leading to different results in many areas. Previous discussion of change in cognitive structure due to the new information also confirms this. Therefore it can be anticipated that some of the information gathered could be changed if another researcher does the same study in future. It will not be possible to generalize on the basis of the results obtained in this study for the reason of that it is the first study of this character in this geographical area with rather small number of informants. Within qualitative studies, it is, however, possible to indicate some tendencies and some patterns based on what is termed analytical generalization (Maaløe 2002). In order to improve analytical generalization it will be necessary to conduct smaller comparison studies, which will contribute to further elaboration of the topic. _____________________________________________________________________________ Page 63 of 72 Nino Jakupovic NJ64041 ---------------------------------------------------------------------------------------------------------------------------------------------------------- Chapter 7 Conclusions and further research 7.0 Conclusion This dissertation was one of first studies concerning beverage choices and consumers in Bosnia and Herzegovina. The study has shown that MEC can be a valuable tool in modeling cognitive structure of consumers in BIH concerning choice of non alcoholic beverages. The analysis of laddering interviews in respect to the main question showed that there are different reasons for choice of a particular beverage category in BIH. The reasons however, are more or less abstract depending on beverage chosen. Coffee was a product most frequently chosen by the interviewed consumers and provided most abstract reasons for choice of all beverage categories. On the other hand all other categories less frequently chosen have also shown that consumers in BiH link personal consequences and values to the choice of beverage, though not to the same degree as for coffee. The analysis of involvement showed that also here the interviewed consumers seem to be more involved in the choice of coffee compared to the other categories. Also here coffee serves as an alternative to the low involvement products that food and beverages are considered to be. Furthermore, HVM’s for beverages show that while carbonated soft drinks, fruit juice and bottled water are mostly chosen for their functionality, coffee is a beverage that tend to be chosen more for its symbolic value. Moreover, the study and analysis of the results have indicated that MEC approach in conjunction with the MECCAS model can be used to compare cognitive structure of the consumers and marketing activities of the different companies at the market. It is clear from the study that current marketing activities of the companies to high degree follow the cognitive structure of the consumers in the area. However, analysis have also indicated that it is possible to identify various alternative routes a hypothetical beverage company can use to influence and target consumers cognitive structure. _____________________________________________________________________________ Page 64 of 72 Nino Jakupovic NJ64041 ---------------------------------------------------------------------------------------------------------------------------------------------------------- 7.1 Implications for further research This study can be regarded as a starting point for more future studies of the products and the area in question. During the process of conducting this dissertation different areas for further research have been identified. First of all the question of consumption context should be further investigated. It would be interesting to investigate whether the results of this study would be different if another context was chosen. Moreover, if different context were used simultaneously it the study could elicit different results. Furthermore, the future research should include use of “hard” laddering approach in order to uncover possible similarities/dissimilarities in cognitive structure using different approaches. As this study covered four categories it can be argued that more narrow approach should be taken in the future studies. They can include only a single beverage which would allow deeper analysis of the consumers and the information gathered. Furthermore, one can suggest a comparative study of Bosnian and Danish consumers in order to investigate whether different cultures produce different information regarding beverages. _____________________________________________________________________________ Page 65 of 72 Nino Jakupovic NJ64041 ---------------------------------------------------------------------------------------------------------------------------------------------------------- Chapter 8 References Anderson , John R. (1983). A spreading activation theory of memory. Journal of verbal learning and verbal behavior, No. 22 pp. 261-295. Assael, Henry (1994). Consumer behavior: a strategic approach; Boston Houghton Mifflin Company. Bagozzi, R.P. and Dabholkar, P.A. (1994). Consumer recycling goals and their effect on decisions to recycle: a means-end chain analysis”, Psychology & Marketing, Vol. 11 No. 4, pp. 313- 341. Barrena, Ramo, Sánchez Mercedes (2009). Consumption frequency and degree of abstraction: A study using the laddering technique on beef consumer. Food quality and Preferences pp. 144-155. Bech – Larsen, Tino (2001). Model-based development and testing of advertising messages: a comparative study of two campaign proposals based on the MECCAS model and a conventional approach. International Journal of Advertising, 20, pp. 499– 519 Bech – Larsen, Tino, Nielsen, Niels Asger, Grunert, Klaus G., Sørensen Elin (1997). Attributes of low involvement products – A comparison of five elicitation techniques and a test of their nomological validity.Centre for market surveillance, research and strategy for the food sector. Working paper no 43. pp. 1 – 24 Bettman, R. James (1979). An information processing theory of consumer choice. Reading, Mass; Addison-Wesley Bettman, R. James, Luce, Mary Frances, Payne, W. John (1998). Constructive Consumer Processes. Journal of Consumer Research Vol. 25 No. 3 pp. 187-217. Botschen Gunther, Hemetsberger Andrea (1998). Diagnosing Means-End Structures to _____________________________________________________________________________ Page 66 of 72 Nino Jakupovic NJ64041 ---------------------------------------------------------------------------------------------------------------------------------------------------------- Determine the Degree of Potential Marketing Program Standardization. Journal of Business Research Vol. 42, pp. 151–159 Celsi L. Richard, Olson C Jerry (1988). The Role of Involvement in Attention and Comprehension Processes. The Journal of Consumer Research, Vol. 15, No. 2, pp. 210224 Claeys, C., Swinnen, A. & Vanden Abeele, P. (1995). Consumers’ means-end chains for ”think” and ”feel” products. International Journal of Research in Marketing, 12 Costa A.I.A.,Dekker M., Jongen V.M.F (2004). An overview of means-end theory: Potential application in consumer-oriented food product design. Trends in Food Science & Technology 15 pp. 403- 415 Costa Ana I. de A, Schoolmeester, Diane, Dekker Mathijs, Jongen, Wim M.F (2007). To cook or not to cook: A means-end study of motives for choice of meal solutions. Food Quality and Preference No. 18 pp. 77-88 Gengler E Charles, Reynolds J Thomas (1995). Consumer understanding and advertising strategy: analysis and Strategic translation of laddering data. Journal of Advertising Research Vol. 35 No. 4 pp. 19-33. Goldenberg, M.A. et al. (2000), “A means-end investigation of ropes course experiences”,. Journal of Leisure Research, Vol. 32 No. 2, pp. 208-24. . Grunert, K. G.; Beckmann, S. C. & Sørensen, E. (2001) “Means-end chains and laddering: An inventory of problems and an agenda for research”. In: Understanding Consumer Decision-making: The Means-end Approach to Marketing and Advertising Strategy, pp. 63-90, T. C. Reynolds & J. C. Olson (Eds.). Mahwah, NJ: Lawrence Erlbaum Associates _____________________________________________________________________________ Page 67 of 72 Nino Jakupovic NJ64041 ---------------------------------------------------------------------------------------------------------------------------------------------------------- Grunert, G. Klaus, Bech, Larsen Tino (2004). Explaining choice attractiveness by beliefs elicited by the laddering method. Journal of Economic Psychology 26 pp. 223241. Grunert, G. Klaus , Scholderer , Joachim (2005). Do Means-End Chains Exist? Experimental Tests of their Hierarchicity, Automatic Spreading Activation, Directionality, and Self-Relevance. Advances in Consumer Research Volume 32 pp. 530 Grunert, K. G., & Grunert, S. C. (1995). Measuring subjective meaning structures by the laddering method: theoretical considerations and methodological problems. International Journal of Research in Marketing, 12, pp. 209–225. Grunert, Klaus (2000) Consumer behavior In Reading list for the MSc course The Aarhus school of business – Department of Marketing. Compenidium E No. 217 pp.110-129 Grunert, K. G., & Grunert, S. C, Sørensen E (1995). Means-End Chains and laddering: an inventory of problems and an agenda for research .Centre for market surveillance, research and strategy for the food sector. Working paper no 34. pp. 1 – 23 Grunert, G. Klaus, Lahteenmaki,Liisa, Nielsen Niels Asger, Poulsen, Jacob B. Ueland Oydis, Astrom Annika (2001). Consumer perceptions of food products involving genetic modification—results from a qualitative study in four Nordic countries. Food Quality and Preference 12 pp. 527-542 Gutman, Johnatan ( 1982). A means-end chain model based on consumer categorization processes. Journal of Marketing No.46 Vol. 2 pp. 60-72. Gutman Jonathan (1984). Analyzing consumer orientations toward beverages through means-end chain analysis. Psychology & Marketing Volume 1 no. ¾ pp. 23-43. _____________________________________________________________________________ Page 68 of 72 Nino Jakupovic NJ64041 ---------------------------------------------------------------------------------------------------------------------------------------------------------- Hofstede, F, Audenaert A, Steenkamp, J.B. Wedel M (1998). An investigation into the association pattern technique as aquantitative approach to measuring means-end chains. International Journal of Research in Marketing 15 pp. 37-50 Howard J.A. (1977). Consumer behavior: Application an Theory, New York : McGaw Hill Book Company Kapferer Jean-Noel, Laurent Gilles ( 1985). Consumer Involvement Profiles: A New Practical Approach to Consumer Involvement. Journal of Advertising Research Journal of Advertising Research pp.48-56. Krauss, Steven Eric (2005). Research paradigms and meaning making. The qualitative report Vol. 10 no. 4 pp. 759-770. Krystallis A, Maglaras G, Mamalis S (2008). Motivations and cognitive structures of consumers in their purchasing of functional foods. Food Quality and Preference No. 19 pp. 525–538 Lin, Chin-Feng (2009). Systematic behavior research for understanding consumer decision making. Behavior Research Methods No. 41 Vol. 2 pp. 405-413 Littlejohn, Stephen W (2002). Theories of human communication. Seventh edition, Thompson McKerchar, Margreth (2008). Philosophical Paradigms, Inquiry Strategies and Knowledge Claims: Applying the Principles of Research Design and Conduct to Taxation. E- Journal of Tax research No.1 . Olson C. Jerry, Reynolds J. Thomas (2001). Understanding consumer decision making The Means-End approach to marketing and advertising strategy. New Jersey, Lawrence Erlbaum Associates, Inc. Peter J. Paul, Olson C. Jerry Grunert G. Klaus (1999). Consumer behaviour and marketing strategy. European edition London: McGraw-Hill. _____________________________________________________________________________ Page 69 of 72 Nino Jakupovic NJ64041 ---------------------------------------------------------------------------------------------------------------------------------------------------------- Phillips, Joan M., Reynolds Thomas J. (2009). A hard look at hard laddering- A comparison of studies examining the hierarchical structure of means-end theory Qualitative Market Research: An International Journal Vol. 12 No. 1, pp. 83 – 99. Rekom, J. v., C. B. M. v. Riel, et al. (2000). Sensemaking from Actions: Deriving organization member's means and ends from their day-to-day behavior. ERIM Report Series Research in Management. Reynolds, Thomas, Gutman Jonathan (1988). Laddering theory, method, analysis and interpretation. Journal of advertising research pp.11-31. Reynolds, T.J., Dethloff, C., Westberg, S.J. (2001). Advancements in laddering, in Reynolds, T.J., Olson, J.C. (Eds),Understanding Consumer Decision Making – The Means-end Approach to Marketing and Advertising Strategy, Lawrence Erlbaum Associates, Mahwah, NJ, pp.91-118 Reynolds, Thomas J. & Craddock, Alyce B (1988). The application of the meccas model to the development and assessment of advertising strategy. Journal of Advertising Research pp. 43 – 54. Rokeach, M.J. (1968). Beliefs, Attitudes and Values. San Francisco: Jossey Bass Rokeach, M.J. (1973). The nature of human values. New York: The Free Press Skytte, H. and K. Bove (2004). The concept of retailer value: A means-end chain analysis. Agribusiness Vol.20 pp. 323- 345 Sohlberg, Marie Britt, and Sohlberg Peter (2003). Erkendlesesformer – Videnskabsteori og forskningsmetode. 1. Edition, Århus; KLIM. Van Rekom, Johan, Wierenga Berend (2007). On the hierarchical nature of means–end relationships in laddering data. Journal of Business Research No. 60 pp. 401 – 410 . Veludo-de-Oliveira, Tania Modesto, Ikeda, Ana Akemi, Campomar Marcos Cortez (2006). Laddering in the practice of marketing research: Barriers and _____________________________________________________________________________ Page 70 of 72 Nino Jakupovic NJ64041 ---------------------------------------------------------------------------------------------------------------------------------------------------------- Solution. Qualitative Market Research: An International Journal Vol. 9 No. 3, pp. 297 – 306. Zaichkowsky, Judith L., Sood James H. (1988). A Global Look at Consumer Involvement and Use of Products. International Marketing Review No. 6,1 pp. 20 – 33 Zannoli , Raffaele.& Naspetti Simona (2002). Consumer motivations in the purchase of organic food – A means-end approach. British Food Journal Vol.104 No.8 pp. 643-653 Wagner, Tillman (2007). Shopping motivation revised. International Journal of Retail & Distribution Management Vol. 35 No. 7 pp. 569-582 Young, S & Feigin, B. (1975). Using the benefit chain for improved strategy formulation . Journal of Marketing, 39 (3) (July). Pp. 72-74. Web Sites: www.marketingpower.com www.coca-cola.ba www.ilidjanskidijamant.ba www.jana.com.hr www.grandkafa.com www.fds.ba/loc/ www.nektar.rs _____________________________________________________________________________ Page 71 of 72