Copyright is owned by the Author of the thesis. Permission is given for a copy to be downloaded by an individual for the purpose of research and private study only. The thesis may not be reproduced elsewhere without the permission of the Author. Integrating the green consumption dimension: Consumer Styles Inventory scale development and validation Fred Angels Amulike Musika Thesis submitted in fulfilment of the requirements for the Degree of Doctor of Philosophy School of Communication Marketing and Journalism Massey University Business School May 2018 | ii ABSTRACT Organisations are increasingly seeking to understand green consumer decision-making and cater for these consumers accordingly. Despite significant practitioner interest, scholarly inquiry into the Green Consumption Styles (i.e., GCS) concept has transpired only relatively recently, resulting in a limited understanding of the concept, and its measurement to-date. Employing an integrative multimethod approach, this thesis addresses this literature gap by developing a measurement instrument for the ‘green consumption scale’ (i.e., GCS) in the context of Tanzania and New Zealand. This thesis is presented in three parts. Part I reports on a literature review and preliminary qualitative research (see Chapters 1-2) conducted to explore/define GCS, and develop an initial GCS item pool. GCS is looked at as “the ways consumers steer their green buying-decision process regarding information searching, evaluation, selection, and purchases.” Part II (Chapter 3-4) provides a theoretical rationale for adopting scale development research in this thesis as well as an overview of the proposed mixed methods research methodology (Chapter 3). It further provides specifications for data-analytical techniques and procedures adopted in this research. Key qualitative research findings were documented in section 3.6, which included the development of the proposed GCS definition, antecedents, and consequences. Chapter 4 dealt with the quantitative analysis of the thesis. A series of EFA and CFA procedures were consecutively undertaken to further assess the GCS scale in study 1 and 2. To explore the scale’s dimensionality, Study 1 an exploratory factor analysis (EFA) results revealed and substantiated a nine-factor, 31-item GCS structure (i.e., green consumption, brand conscious, Recreational, Perfectionistic, Impulsiveness, confused by over-choice, Habitual/brand-loyal, Novelty-fashion-conscious, and Price Conscious) (Table 4.12) using a sample of n=448. Finally, the results suggest a combined (original CSI scale by Sproles and Kendall (1986) plus green scale 9-factor solution with 31-items (see Chapter 4). Using the reduced, 31- item scale and a new sample of n=225 Tanzania and New Zealand-based consumers, confirmatory factor analysis (CFA) is undertaken in study 2 to confirm the nine-factor, 31-item GCS scale (section 4.3). This analysis also facilitated the assessments for the model construct validity (Chapter 4). CFA was also conducted, which served to confirm the nine-factor, 31-item GCS scale. Further, regression analyses have been done to provide predictive validity of the newly developed GCS measure was undertaken. The findings indicated the attainment of high GCS items scores across the two samples; thus, providing evidence for the robustness of the GCS scale across samples and cultures. Furthermore, adequate Cronbach’s alphas were reported for each of the proposed GCS factors, in addition to the overall GCS scale. Part III provides the contributions, limitations and future research directions arising from this thesis (Chapter 5). The chapter commenced with an overview of key contributions of this research, followed by an overview of the key research limitations and directions for future research. Keywords: Green consumption scale, structural equation modelling, scale development. | iii Acknowledgments First and foremost, I thank my supervisors Professor Valentyna Melnyk, Dr. Andrew Murphy, and Dr. Alexandra Hess for their support on the journey that this project has represented over the last few years. Your contribution is beyond the words of mouth can comprehend. I am very thankful for invaluable support you rendered me. I really Appreciate. Indeed, I am grateful for the invaluable lessons and opportunities, and hope to make you proud. I am also thankful for the fun, intellectual stimulation and ongoing encouragement, which have helped me see this project through to completion. Thanks also to Florida and Frida for the good times, and for helping me keep a good balance during my candidacy; for providing invaluable friendship and support over the years, each in your own unique way. I hope our friendship will last for many years to come. Thank you to my family for being part of my journey. | iv Table of Contents ABSTRACT ................................................................................................................... ii Acknowledgments ...................................................................................................... iii List of Tables ............................................................................................................. vii List of Figures .............................................................................................................. x List of Appendices ...................................................................................................... xi PART I: PREFACE, LITERATURE REVIEW & CONCEPTUAL DEVELOPMENT ....................... 1 INTRODUCTION ................................................................................... 2 1.1 Background of the Study ........................................................................................ 2 1.2 Consumer Decision-Making Styles (CDMS) Concept ................................................ 2 1.3 Culture and CSI ...................................................................................................... 4 1.4 Green Consumption ............................................................................................... 4 1.5 Statement of the Problem and Research Questions ................................................ 5 1.6 Objectives of the Study .......................................................................................... 5 1.7 Significance of the Study ........................................................................................ 6 1.8 Theoretical Framework .......................................................................................... 6 1.9 Research design ..................................................................................................... 7 1.10 Thesis Structure ................................................................................................... 7 LITERATURE REVIEW ............................................................................ 8 2.1 Introduction and overview ..................................................................................... 8 2.2 Consumer Decision-Making Styles (CDMS): Definition and an overview .................. 8 2.3 CSI validity, reliability, generalizability, and applicability ........................................ 9 2.4 CSI profiles and dimensions ................................................................................. 17 2.4.1 Perfectionism ..................................................................................................... 17 2.4.2 Brand consciousness.......................................................................................... 20 2.4.3 Novelty-fashion consciousness ......................................................................... 21 2.4.4 Recreational, hedonistic-shopping consciousness ............................................ 23 2.4.5 Price and value-for-money consciousness ........................................................ 26 2.4.6 Impulsiveness .................................................................................................... 28 2.4.7 Confusion from over-choice .............................................................................. 30 | v 2.4.8 Habitual, Brand Loyal ......................................................................................... 34 2.5 Profiles, dimensions, and features of the CSI Scale ............................................... 36 2.6 The CSI and its Applicability in Different Cultures ................................................. 37 2.6.1 The CSI in Asia (China) ....................................................................................... 41 2.6.2 The CSI in Germany............................................................................................ 42 2.6.4 The CSI in United Kingdom ................................................................................ 43 2.6.5 The CSI in Australasia ......................................................................................... 44 2.6.6 CSI trends among student consumers ............................................................... 45 2.6.7 CSI trends in general public samples ................................................................. 49 2. 7 Newly identified CSI factors ................................................................................ 54 2.8 Green Consumption – An Emerging Trend............................................................ 63 2.8.1 Related Scales to Green Consumption .............................................................. 67 2.10 Summary and Conclusion of the Literature Review and Directions for the Present Study ........................................................................................................................ 69 PART II: RESEARCH METHODOLOGY AND RESULTS .................................................... 70 EXPLORATORY QUALITATIVE RESEARCH ............................................ 71 3.1 Chapter Overview ............................................................................................... 71 3.2 Research Methodology Overview, Purpose, and Rationale................................... 71 3.3 Development of the Green Consumption Scale: Overview ................................... 72 3.4 Step 1: Specify domain of construct ..................................................................... 73 3.5 Step 2: Generate sample of items ........................................................................ 74 3.5.1 Interviews .......................................................................................................... 74 3.5.2 The Focus Group ............................................................................................... 77 3.5.3 Expert judges’ item assessment and refinement ............................................... 79 3.6 Qualitative Research: Findings ............................................................................. 80 3.6.1 Green Product Concept Definition .................................................................... 81 3.6.2 Key Green Products Themes ............................................................................. 81 3.7 Chapter Summary ............................................................................................... 82 CHAPTER 4: QUANTITATIVE ANALYSIS ....................................................................... 83 4.1 Introduction ........................................................................................................ 83 4.2 Study 1: Green Scale Development – Exploratory Factor Analysis ......................... 83 | vi 4.2.1 Step 3: Data Collection: research design ........................................................... 83 4.3 Step 4: Purify measure ......................................................................................... 85 4.3.1 Results Study 1A (New Zealand participants – Green Dimension only) ............ 85 4.3.2 Results Study 1B (Tanzania participants – Green Dimension only)................... 88 4.3.3 Results Study 1C ............................................................................................... 90 4.4 Study 1 Summary ................................................................................................. 93 4.5 Study 2: Confirmatory Factor Analysis (CFA) ......................................................... 94 4.5.1 Study 2 introduction (Churchill’s steps 5-8) ...................................................... 94 4.5.2 Step 5: Survey Design and Structure ................................................................. 95 4.5.3 Results: Confirmatory factor analysis ................................................................ 97 4.5.4 Step 6: Construct Validity ................................................................................ 100 4.5.5 Step 7: Predictive Validity ................................................................................ 105 4.5.6 Step 8: Norms .................................................................................................. 109 4.6 Study2 Summary ................................................................................................ 110 4.7 Summary of Chapter 4: Green Scale and CSI Analysis .......................................... 110 4.8 Summary of Part II – Green Scale Development and Validation .......................... 111 4.8.1 Key findings .................................................................................................... 112 PART III: CONTRIBUTIONS AND FUTURE DIRECTIONS ............................................... 114 CHAPTER 5: CONTRIBUTIONS, LIMITATIONS AND FUTURE DIRECTIONS ................... 115 5.1 Introduction ...................................................................................................... 115 5.2. An Overview ..................................................................................................... 115 5.3 Contributions ..................................................................................................... 120 5.3.1 Academic Contributions .................................................................................. 121 5.3.2 Managerial Contributions ................................................................................ 121 5.4 Limitations and Future research ......................................................................... 122 5.4.1 Limitations ....................................................................................................... 122 5.4.2 Areas for future research ................................................................................ 123 5.5 Chapter Summary .............................................................................................. 124 REFERENCES ............................................................................................................ 125 APPENDICES ............................................................................................................ 138 | vii List of Tables Table 1.1: Consumer Decision-Making Styles: dimension description ....................................................... 3 Table 2.1 CSI validity, reliability, generalizability and applicability results (Cronbach Alpha) .................. 10 Table 2.2a Sample, instrument, analysis, and Cronbach Alpha ................................................................ 12 Table 2.2b: Researchers concluding remarks on the CSI’s validity, reliability, applicability, and generalisability .............................................................................................................................. 14 Table 2.3 Studies that do not support Perfectionism .............................................................................. 18 Table 2.4 The Wickliffe study summary................................................................................................... 19 Table 2.5 Studies that do not support Novelty-Fashion consciousness ................................................... 23 Table 2.6 Studies that do not support the recreational, hedonistic shopping consciousness................... 25 Table 2.7: Studies that do not support the price-value consciousness factor .......................................... 28 Table 2.8: Studies that do not support impulsiveness factor ................................................................... 30 Table 2.9 Studies that do not support confusion from overchoice .......................................................... 34 Table 2.10. Studies that do not support the habitual, brand loyal factor ................................................ 36 Table 2.11 The CSI trend in the US .......................................................................................................... 37 Table 2.12: Culture and sample structure influence ................................................................................ 38 Table 2.13 Studies with Multi-cultural samples ....................................................................................... 40 Table 2.14 The CSI trend in China ............................................................................................................ 41 Table 2.15 The German CSI ..................................................................................................................... 43 Table 2.16 The CSI in the United Kingdom ............................................................................................... 44 Table 2.17 CSI trends among student consumers .................................................................................... 47 Table 2.18 CSI trends in general public samples ...................................................................................... 50 Table 2.19: CSI Trends Influencers ........................................................................................................... 52 Table: 2.20 Time-energy conserving ........................................................................................................ 54 Table 2.21 Time conscious ...................................................................................................................... 55 Table 2.22 Time restricted ...................................................................................................................... 55 | viii Table 2.23 Quality Conscious .................................................................................................................. 56 Table 2. 24 Price Conscious ..................................................................................................................... 57 Table 2.25 Economy seeking ................................................................................................................... 58 Table. 2.26 Variety Seeking ..................................................................................................................... 58 Table 2.27. Enjoyment-Variety Seeking................................................................................................... 59 Table 2.28 Recreational, hedonistic conscious ........................................................................................ 60 Table 2.29 Information Utilization .......................................................................................................... 60 Table 2.30 Brand loyal ............................................................................................................................ 61 Table 2.31 Store Loyal ............................................................................................................................. 61 Table 2.32 Satisfying ............................................................................................................................... 62 Table 2. 33 Fashion-sale seeking ............................................................................................................. 62 Table 2.34: Green consumption dimension items ................................................................................... 68 Table 3.1: Interview Results: Consumer decision-making styles ............................................................. 75 Table 3.2. Interview Results: Green consumption ................................................................................... 76 Table 3.3 Focus Group Participants* ....................................................................................................... 77 Table 3.4. Items to be presented to the expert Judges (For scale items purification) ............................. 79 Table 3.5. Expert judges’ item assessment and refinement .................................................................... 80 Table 4.2. KMO and Bartlett's Test- New Zealand– 10 items ................................................................... 86 Table 4.3. Communalities- New Zealand – 10 items ................................................................................ 86 Table 4.4 Communalities – New Zealand – 9 items ................................................................................. 87 Table 4.5. KMO and Bartlett's Test – New Zealand - 9 items ................................................................... 87 Table 4.6 Total Variance Explained - New Zealand – 9 items ................................................................... 87 Table 4.7 KMO and Bartlett's Test – Tanzania – 9 items .......................................................................... 89 Table 4.8 Total Variance Explained – Tanzania – 9 items ........................................................................ 89 Table 4.9 KMO and Bartlett's Test – Whole sample (New Zealand and Tanzania) with green and CSI ..... 91 Table 4.10 Communalities – Whole sample (New Zealand and Tanzania) with green and CSI ................. 91 Table 4.11 Total Variance Explained– Whole sample (New Zealand and Tanzania) with green and CSI .. 91 Table 4.12 Rotated Component Matrix ................................................................................................... 93 Table 4.13 Frequency Table- Country-wise ............................................................................................. 96 | ix Table 4.14 Age Distribution Frequency Table .......................................................................................... 96 Figure 4.5 Age Distribution ..................................................................................................................... 97 Table 4.15 SEM Fit Index ....................................................................................................................... 103 Table 4.16 Standardised Regression Weights ........................................................................................ 105 Table 4.17 Predictive Validity ................................................................................................................ 108 Table 4.18 One Sample Statistics .......................................................................................................... 109 Table 4.19. Group Sample Statistics: Two-sample t-test NZ vs Tanzania ............................................... 109 Table 4.20: Independent Samples Test .................................................................................................. 110 | x List of Figures Figure 1.1: Consumer Decision-Making Styles ........................................................................................... 3 Figure 1.2 Green - Consumption CSI Framework ....................................................................................... 6 Figure: 3.1 Churchill’s Procedure for Scale Development ........................................................................ 73 Figure 4.1: Age distribution ..................................................................................................................... 84 Figure 4.2. Scree Plot - New Zealand – 9 items ........................................................................................ 88 Figure 4.3. Scree Plot– New Zealand – 9 items ........................................................................................ 89 Figure 4.4. Scree Plot– 33 items .............................................................................................................. 92 Figure 4.6: Structural Model ................................................................................................................. 102 Figure 4.7. Standardised regression weights ......................................................................................... 104 | xi List of Appendices Appendix 1. Summary table for this study’s CSI literature ...................................................................... 138 Appendix 3: Tanzania & Tanzania Survey - II ......................................................................................... 144 Appendix 3: Regression Summary ......................................................................................................... 150 Appendix 4: ANOVA .............................................................................................................................. 151 Appendix 5: Independent Variables that have impact on two or more variables .................................. 152 Appendix 6: Coefficients ....................................................................................................................... 153 Appendix 7: Independent Variables that affected only one DV ............................................................. 154 Appendix 8: GREEN Independent Variables .......................................................................................... 155 | 1 PART I: PREFACE, LITERATURE REVIEW & CONCEPTUAL DEVELOPMENT | 2 INTRODUCTION 1.1 Background of the Study Marketers believe that consumers’ consumption style has an influence on the purchasing decision of a consumer. Yet, the exact structure of the factors influencing the decision is still debated. Specifically, a recent trend in consumer decision making is a growing demand for green and sustainable products. For example organic food demand in China has quadrupled between 2010-2015 and is expected to continue to rise (Li, Ge, & Bai, 2013; McCarthy, 2015). Further, organic food demand has increased by double-digits since 1990 (USDA ERS, 2016; Trauger and Murphy, 2013), growing faster than all other food sectors (Nie & Zepeda, 2011) with demand growth projected at 14% per annum until 2018 (Daniells, 2014; Mosier & Thilman, 2016). Further, certified organic cropland in the United States increased from 163,250 to 1,248,000 hectares between 1992 and 2011 (USDA ERS, 2013). Moreover, organic food sales have experienced tremendous growth in the last decade reaching a USD 43 billion mark in 2016 (Statista, 2017). There is also an increase in organic food research activity and funding (USDA ERS, 2015). This growing demand for green products reflects changes in consumer decision making-styles and attitudes, as consumers set up a kind of attitude towards green consumption (Yoon, 2013). Yet, measures to capture this green orientation in the context of consumer decision making are still virtually non-existent. This situation therefore calls for researchers and marketers to identify the types of consumption factors that influence consumer’s decision-making. This is a gap that this study aims to fill. 1.2 Consumer Decision-Making Styles (CDMS) Concept The concept of Consumer Decision-Making Styles (CDMS) refers to the ways consumers steer their buying- decision process regarding information searching, evaluation, selection, and purchases (Sproles & Kendall, 1986). These styles may differ depending on products and the market (Bauer, Sauer, & Becker, 2006). Marketers use CDMS to evaluate market segments and for developing effective positioning strategies (Walsh, Thurau, & Mitchell, 2001; Wang, Siu, & Hui, 2004), and for understanding cultural differences in buying, decision-making styles, and product adoption (Walsh, Mitchell, & Hennig‐Thurau, 2001). The most generally recognised approach to general Consumer Decision-Making Styles is the study of “Consumers Styles Inventory” (Sproles & Kendall, 1986). The CSI’s key assumption is that each consumer has a specific decision-making style, involving individual decision-making dimensions (Wesley, LeHew, & Woodside, 2006). While this CSI inventory undoubtedly represents a systematic measure of buying orientations using decision-making coordination, it does not take the most recent consumer developments, such as emerging changes towards green orientation. Specifically, the CSI construct is comprised of eight dimensions as summarised in Figure 1.1 and Table 1.1 respectively. http://www.tandfonline.com/doi/full/10.1080/08974438.2016.1266565 http://www.tandfonline.com/doi/full/10.1080/08974438.2016.1266565 http://www.tandfonline.com/doi/full/10.1080/08974438.2016.1266565 http://www.tandfonline.com/doi/full/10.1080/08974438.2016.1266565 http://www.tandfonline.com/doi/full/10.1080/08974438.2016.1266565 | 3 Figure 1.1: Consumer Decision-Making Styles The Consumer Styles Inventory (Li et al., 2013; McCarthy, 2015; Zsóka et al., 2013) is defined as a consumer characteristics approach with emphasis on cognitive and affective consumer decision- making styles (CDMS) (Sproles & Kendall, 1986). It also measures the type of mental characteristics that are present when consumers make purchasing decisions (Sinkovics, Leelapanyalert, & Yamin, 2010). Table 1.1: Consumer Decision-Making Styles: dimension description N o Decision-making style Description 1 Perfectionism, high-quality conscious Search for the best quality in products; shop more carefully, more systematically, or by comparison; not satisfied with the “good enough” product. 2 Brand consciousness / Price Equals Quality Oriented toward buying more expensive well-known brands. Likely to believe that higher price equals higher quality; have a positive attitude towards stores with brand names and higher prices; prefer bestselling advertised brands. 3 Novelty-fashion conscious Fashion and novelty conscious; gain excitement and pleasure from seeking out new things; keep up-to-date with styles; like being in style; seek variety. 4 Recreational, hedonistic Feel pleasure to shop; shop just for fun of it; shop for recreation and entertainment. 5 Price and “value for money” Look for sale prices; conscious of lower prices in general; concerned with getting the best value for money; comparison shoppers. 6 Impulsiveness, Careless Do not plan their shopping; unconcerned about how much they spend or about the “best buys.” 7 Confused by over-choice Perceive many stores and brands from which to choose, and have problem in making choices; experience information overload. Proliferation of brands, stores, and consumer information 8 Habitual, brand-loyal Likely to have favourite brands, stores and form habit to choosing these. Developed from Sproles and Kendall (1986). For the past few decades, Consumer Decision-Making Styles (CDMS) outcomes for different cultures have been studied widely using Sproles and Kendall’s (1986) CSI framework. The results of those studies (e.g. Chen et al., 2012; Durvasula et al., 1993; Kwan et al., 2008; Sinkovics et al., 2010; Sproles & Kendall, 1986, Sproles and Sproles, 1990; Zhou et al., 2010) generally supported the framework. Hence, Sproles and Kendall’s CSI model has been taken as consistent and universal (Chen et al., 2012; Sinkovics et al., 2010; Zhou et al., 2010). Yet, while CSI has been considered the most superior, stable, and widely used CDMS scale across the globe (Hafstrom et al., 1992; Lysonski et al., 1996; Sinkovics et al., 2010; Wickliffe, 2004), there are several important issues raised with this model. Perfectionism Brand consciousness Novelty-fashion consciousness Recreational, hedonistic consciousness Price consciousness Impulsiveness Confusion from over- choice Habitual Consumer Styles Inventory (Zsóka et al.) | 4 1.3 Culture and CSI Although the CSI framework has been recognized as universal, prior research suggests that culture is an important moderator of this framework. For example, researchers (Chen et al., 2012; Cowart & Goldsmith, 2007; Wickliffe, 2004; Yang & Wu, 2007) argued that the interpretation of CSI could only be fully understood in the consumption context of a given culture. This means, one CSI style scale that is seen suitable in one culture may be regarded unsuitable in another culture (McCarthy, 2015; Wickliffe, 2004; Yasin, 2009), because among the most important factors that influence consumers’ decision-making styles is the contexts they are in. Consistently, various studies reported that consumers differ in their consumption behaviour patterns due to differences in their cultural background (Solka et al., 2011). For example, when consumers are frequently exposed to a given culture, they become affected by the norms and values of that particular culture. Subsequently, the learned norms and values offer criteria that consumers will use to direct their own consumption decisions. Therefore, these different cultural orientations lead to different consumer decision- making orientation and the meanings given for such interactions may also vary among cultures (Wicklife, 2004). Further, Wickliffe (2004) argued that Sproles and Kendall’s American-based CSI might not be culturally relevant and meaningful to consumers in Asia, Africa, or Latin America as they attribute different meanings to consumption values as for Europeans and North Americans (Melnyk, Giarratana & Torres, 2013). This is because CDMS have been extensively studied in the west, and to a much lesser extent in developing countries (Wang et al., 2004; Byrne, 2010). Consistently, studies (e.g. Kacen & Lee, 2002; Kim et al., 2009) carried out in emerging Asian cultures like China, Taiwan, and Korea have provided evidence that CSI in emerging countries may reveal somewhat different results from those shown in Western cultures; as some CSI factors and dimensions were rejected and new ones emerged (Zhou, Arnold, Pereira, & Yu, 2010). In additional, there has been variations in the identified CSI dimensions among different studies from different countries (Solka et al., 2011). Many studies have found that the CSI differs between intra-cultural cohorts (Hafstrom et al., 1992; Lysonski et al., 1996). For instance, Kwan, Yeung, and Au (2008) reported that Chinese consumers in different locations display different CDMS. Correspondingly, Kamaruddin and Mokhlis (2003) showed the presence of intra-cultural differences in Chinese Malay. The presence of intracultural differences was as well reported in studies carried out in Taiwan and the US (Chen et al., 2012; Cowart & Goldsmith, 2007; Yang & Wu, 2007). Yet, the nature of the differences remain unclear and there was a call in the literature for extending CSI for emerging countries (Eun Park, Yu, & Xin Zhou, 2010; Kavas & Yesilada, 2007; Sinkovics, ‘Mink’ Leelapanyalert, & Yamin, 2010) . 1.4 Green Consumption While CSI has been an established instrument, focusing on culture, age, gender, regular products, and services over the last decade (Cowart & Goldsmith, 2007; Hafstrom, Chae, & Chung, 1992; Kasper, Bloemer, & Driessen, 2010; Lysonski et al., 1996; Solka, Jackson, & Lee, 2011), however, the instrument did not incorporate recent consumer trends, such as the emergence of organic food consumption (Wang et al., | 5 2004; Yasin, 2009; Dumortier et al., 2017). Furthermore, the CSI has not captured new CDMS dimensions and traits such as fair trade and green consumption. Green Consumption is taken here to refer to the consumption of goods and services that are: biodegradable, recyclable, fair traded, organic, non-toxic, eco-friendly, or renewable (Autio, Heiskanen, & Heinonen, 2009; Murphy & Jenner-Leuthart, 2011; Ibok & Etuk, 2014; Trauger & Murphy, 2013; Wu & Chen, 2014). As a result, little is known about potential green consumption aspects of the CDMS scale’s characteristics, including conceptualisation, profiling, and operationalisation, which are critical for marketers, policy makers, consumer counsellors, and researchers. 1.5 Statement of the Problem and Research Questions The general goal of this study is to address the two major gaps with regards to the Consumer Decision- Making Styles (CDMS), i.e., 1) to develop a measure that captures the recent developments in the green consumption domain and 2) to test the generalisability of the measure across both developed and developing countries. In particular, this research aims to 1) develop and comprehensively validate green CSI scale (instrument) in the context of Consumer Decision-Making Styles (CDMS) measure and 2) examine consumers’ purchasing decision-making of different products using the newly developed green CSI scale separately and as a part of the CSI measure 3) across both emerging (Tanzania) and developed (Marshall, Baldwin, Peach, 2008) country contexts. Therefore, the present study was designed to answer the following research questions. Research Question1. What are the types of green Consumer Decision-Making Styles (CDMS) exercised in New Zealand and Tanzania? Research Question 2. Which items should constitute the green consumption scale? Research Question 3. To what extent does the newly developed green CSI scale measure consumers’ consumption style and is generalizable across both developed and emerging countries? 1.6 Objectives of the Study The general objectives of this study were (a) to develop and validate green CSI scale, and then (b) to investigate green Consumer Decision-Making Styles (CDMS) with consumers’ buying and consumption style. Specifically, the study intends to: • identify the types of green Consumer Decision-Making Styles (CDMS) exercised in New Zealand and Tanzania. • Develop and validate a new green CSI scale. • investigate the extent to which the newly developed green CSI relates to consumers’ consumption styles across both developed (Marshall et al., 2008) and emerging (Tanzania) countries. | 6 1.7 Significance of the Study This study contributes to consumer decision-making research (Chaudhary & Dey, 2016; Merriam, 2017; Sproles & Kendall, 1986; Frimpong, Nwankwo, & Omar, 2015) by developing and comprehensively validating green consumption CSI scale as perceived by consumers about their own consumption styles (Diamantopoulos, Schlegelmilch, Sinkovics, & Bohlen, 2003; Joshi & Rahman, 2015; Lee, 2009; Sehgal, Landran, & Singh). Likewise, this study is significant in developing a green consumption scale that is validated within both developed and emerging economies (New Zealand and Tanzanian respectively) green consumption contexts. There is a scarcity of studies on CDMS in Tanzania and other African countries, and so one cannot draw a conclusion regarding the predominant consumption style exercised in Tanzania. The newly developed consumption instrument will be of a great importance for researchers, consumers, marketers and other organizations that work with consumers. It is hoped that it will provide researchers with a valid and reliable instrument for measuring green consumption style objectively. Moreover, it provides insight for consumers and other interested stakeholders who work in the area by identifying the most dominant green- consumption style in the said countries. Finally, developing and validating a suitable green-consumption CSI scale in the context of developing country like Tanzania contributes to the research on environmental issues in emerging countries because such studies are very scarce in the context of the emerging economies (Biswas & Roy, 2015; Saxena & Khandelwal, 2010) 1.8 Theoretical Framework This research uses Consumer Styles Inventory (Zsóka et al., 2013) theoretical model and concept as its theoretical framework; and the two countries’ consumer market providing a worthy context to explore the CSI concept more closely. Furthermore, CSI growth and popularity deem it a worthy scale development research context in which to study consumer decision-making styles (DMS). From different studies, it appears that marketers and academics have not yet identified the predictors, roles, processes, and effects associated with the CSI green consumption dimension (Wesley et al., 2006; Sinkovics et al., 2010). As such, this study proposes a new green dimension CSI framework namely “green consumption”. The green consumption dimension is the newly proposed dimension to the original CSI scale as shown in the proposed new framework in Figure 1.2 below. Figure 1.2 Green - Consumption CSI Framework Perfectionis m Brand consciousnes s Novelty- fashion consciousness Recreational, hedonistic consciousness Price Conscious Impulsiveness Confusion from over- choice Brand loyalty Green Consumption Green consumption–integrated CSI Decision Consumer Styles Inventory (Zsóka, Szerényi, Széchy, & Kocsis) | 7 1.9 Research design This study employs multi-method approach, making use of both qualitative and quantitative techniques. First, exploratory research design was employed to explore the types of green Consumer Decision-Making Styles (CDMS) and their indicators qualitatively, then followed by quantitative technique to examine the psychometric properties of the scale. Samples were drawn from the general public in New Zealand and Tanzania using simple random sampling and stratified random sampling. Data for qualitative analysis was collected using a) focus groups, b) interviews, and c) discussions with expert judges. The data for quantitative analysis were gathered using on-line survey across 2 studies, involving New Zealand and Tanzanian samples. Based on the qualitative research, a preliminary item pool of 10 Green Consumption Scale (GCS) items are identified and proposed (see Chapters 2 and 3). In order to explore the scale’s dimensionality, Study 1 an exploratory factor analysis revealed a 9 Item Solution for the Green Factor using a sample of n=448. Finally, the results suggest a combined (original CSI scale by Sproles and Kendall (1986) plus green scale) nine-factor solution with , 31-item (see Chapter 4). Using the reduced, 31- item scale and a new sample of n=225 Tanzania and New Zealand-based consumers, confirmatory factor analysis is undertaken in study 2 to confirm the nine-factor, 31-item GCS scale (Chapter 4). This analysis also facilitated the assessments for the model construct validity (Chapter 4). 1.10 Thesis Structure This thesis is structured into three parts representing the stages undertaken in this research process. Part I provides an introduction, literature review and conceptual development for this research. This chapter has presented an introduction and thesis overview by identifying a key literature gap, and addressing how this research attempts to remedy this gap. This chapter has also introduced the conceptual foundations underlying this research, and provided an overview of the research purpose and methodology, as well as the expected contributions. The next chapter provides a review of recent CDMS research in Consumer Behaviour, and addresses the preliminary GCS conceptual development procedures undertaken. Further, Chapter 2 reviews key literature addressing the application of CDMS in green consumption in emerging market settings. Part II provides an overview of the adopted mixed method research approach and its relevance. Qualitative research using in-depth-interviews and focus group was first conducted to explore the nature and features of GCS (Chapter 3). The GCS scale development procedures and model validity assessments are reported in Chapter 4. Finally, Part III addresses the research contributions, limitations, and future directions arising from this research. Specifically, Chapter 5 identifies key contributions and implications arising from the research, and provides an overview of selected limitations inherent in this research. The thesis concludes by proposing future research directions. | 8 LITERATURE REVIEW 2.1 Introduction and overview This chapter presents the review of previous research on CDMS, CSI, and scale development and validation. In line with this, a review of relevant consumption research in different countries is discussed as well as the concept of green consumption dimension in CSI scale. This chapter also discusses the scale development and validation models and procedures relevant to this research. Finally, the results of the literature review were summarized, and the directions for the current study are also underscored. 2.2 Consumer Decision-Making Styles (CDMS): Definition and an overview The concept of CDMS refers to the ways consumers steer their buying-decision process regarding information searching, evaluation, selection, and purchases (Sproles & Kendall, 1986). These styles may differ depending on products and the market (Bauer, Sauer, & Becker, 2006). For example, consumers tend to be more price-, brand-, and quality-conscious for luxury products than convenience goods in their decision-making styles (Leo, Bennett, & Härtel, 2005b). Marketers use CDMS to evaluate market segments and for developing effective positioning strategies (Walsh, Thurau, & Mitchell, 2001; Wang, Siu, & Hui, 2004), and for understanding cultural differences in buying, decision-making styles, and product adoption (Walsh, Mitchell, & Hennig‐Thurau, 2001). However, substantial academic research has focused on traditional products in evaluating CDMS (Walsh, Thurau, Mitchell, 2001) paying little attention to green consumption. Generally, from the above given explanations, it can be said that CDMS involves behavioural, attitudinal, and emotional interactions in which consumer meet consumption needs and wants CDMS have been popular since the 1950s and used in numerous studies from local to cross-country comparison studies during the late 1980s (Yasin, 2009). Among the three CDMS instruments (the CSI, Consumer Typology, and Consumer lifestyle), this study will concentrate on the CSI for the reasons explained in Chapter 1. The CSI was developed by Sproles and Kendall in 1986. The main theoretical assumption behind Sproles and Kendall (1986) ideas about CDMS is that consumers have eight different decision-making dimensions that determine the shopping decisions they make. As people continue to buy, it is imperative to understand how they make decisions as consumers, which calls for a better understanding of CDMS because it is linked to their purchase behaviours (Mitchell & Bates, 1998; Yasin, 2009). The CSI is used to profile, understand, and predict consumers’ buying behaviour and loyalty (Zhou et al., 2010). It can also be applied as a consumer education tool and as a counselling device; for market segmentation, positioning, and marketing-mix adjustment strategies for goods and services (Mitchell & Bates, 1998; Yasin, 2009); as a quantitative technique for categorising consumers’ heterogeneous decision- making styles into discrete categories (Lysonski et al., 1996). This last feature is one of the factors that informed the choice of methodology approach for this study. | 9 Despite the usefulness of the CSI, it has been observed that some goods and services are not general, some CSI respondents may be encouraged to take one product as their primary point of reference, which limits its ability to measure consumers' decision-making styles (Mitchell & Bates, 1998). Also, the use of the US- based CSI might disguise the richness of country-specific CDMS, which an ethnographically grounded instrument might uncover (Mitchell & Bates, 1998). This is because some cultures differ in the extent to which the CSI dimensions were confirmed (Lysonski et al., 1996). Therefore, in order to have a better understanding of the CSI concept, a discussion of its profile, reliability, validity, applicability, and the newly proposed green consumption dimension is discussed hereunder. I commence with a discussion of the validity, reliability, generalisability, and applicability of the CSI across cultures, demographics, and social classes. This is essential as it is important to establish how the CSI scale performed in past studies in different nations and times. 2.3 CSI validity, reliability, generalizability, and applicability Debate over the validity, reliability, applicability, and generalisability of the CSI continues. Proponents of the CSI argue that most of its variables have satisfactory or higher reliability (Zhou et al., 2010). This position is also supported by Sproles and Sproles (1990), who found statistically significant relationships between learning and the CSI characteristics. Hafstrom et al. (1992) examined and compared the consumer decision- making styles of US and Korean youth, and found that they shared seven out of eight styles. This may mean that youth across the globe can be marketed to in an almost similar fashion by international marketers, hence reducing marketing costs and improving marketing results. Further, these researchers concluded that the CSI has elements of construct validity and usage potential across nations. In a New Zealand study (Durvasula et al. 1993) the generalisability of the CSI was compared to the US original, the results showed that the similarities between these two countries outweighed the differences, hence providing general support for the CSI scale. Ten years after the introduction of the CSI, Lysonski et al. (1996) researched the CSI in four diverse countries (Greece, India, New Zealand, and the US) in a major study, and they confirmed seven out of the eight factors with 34 items. Similarly, Mitchell and Bates (1998) found evidence for generalisability of CSI styles and showed that most of the original US traits were found in the UK. Later, Mitchell and Walsh (2004) reported that German consumers found that seven out of eight CSI characteristics were valid for female consumers. Twenty years after the introduction of CSI, Bakewell and Mitchell (2006), confirmed all of the eight original US CSI factors, compared to the previous study in the UK by Mitchell and Bates (1998), which confirmed six factors only. Hence, this shows a good progress on the validity and reliability of the CSI model with time. This may imply that as time goes on so does the confirmation of the CSI instrument. Similarly, Wesley et al. (2006) findings supported the CSI’s applicability among adult shoppers in different mall contexts in America. Further, Yang and Wu (2007) confirmed six characteristics of the CSI for Taiwanese internet shoppers, and five years later all eight CSI factors were confirmed in Taiwan (Chen et al., 2012). Likewise, Sinkovics et al. (2010) confirmed the CSI with results that are congruent with findings from earlier studies using student | 10 samples. This confirmatory-oriented progress regarding the CSI could be a good sign of acceptability and reliability of the CSI instrument. Even though there is some evidence for the cross-cultural validity of the CSI, several face-validity problems became evident when validating the scale by respondents from outside the US, hindering meaningful validation (Mitchell & Bates, 1998). Some studies have shown mixed results in the validity and reliability of the CSI. For example, Lysonski et al. (1996) showed that the CSI requires additional psychometric work before it can be applied to other countries, mainly less-developed countries. Also, some researchers have confirmed less than six factors in their CSI studies (Hiu et al., 2001; Walsh, Thurau, et al., 2001). Mitchell and Walsh (2004) demonstrated that the CSI is gender-biased and has constructed validity for females, but not for males. If one group of consumers is omitted with regard to the CSI, then this is a sign of a problem. In another study, Wickliffe (2004) showed that the CSI is not a reliable or valid measure of CDMS in both Korea and the US. However, regarding Wickliffe (2004) study, it is unclear whether these differences are caused by population variances, or analytical and interpretation problems (Mitchell & Bates, 1998). Some researchers suggest that the differences among economies may affect the generalisability of the CSI as a consumer decision-making gauge (Lysonski et al., 1996). Following the above argument, there are some questions that will require answers, such as: what type and level of the economy has a positive or negative impact on the CSI, as well as what effect an economy has on CSI and consumer decision-making styles, and many other questions. Adding to that, some researchers have the opinion that the differences seen in CSI factor loadings may be due to chance variation, researcher bias, recording, coding, data analysis errors, change in the phenomenon over time, difficulty in interpreting the CSI in other countries, and cultural differences in decision-making styles (Lysonski et al., 1996 ; Mitchell & Bates, 1998; Yasin, 2009). However, it seems most studies followed a similar CSI construct and approach to that used in the original study, and still produced different results, as shown in Table 2.1 below. These inconsistent results raise some concern about the CSI scale. Table 2.1 CSI validity, reliability, generalizability and applicability results (Cronbach Alpha) No . Reference Country Factors P er fe ct io n is t B ra n d c o n sc io u s N o ve lt y- fa sh io n co n sc io u s R ec re at io n al Sh o p p in g co n sc io u s P ri ce co n sc io u s/ V al u e fo r th e m o n ey Im p u ls iv e C o n fu se d b y o ve r- ch o ic e H ab it u al b ra n d lo ya l 1 (Sproles & Kendall, 1986) US 0.74 0.75 0.74 0.76 0.48 0.48 0.55 0.53 2 (Hafstrom et al., 1992) US & KR 0.77 0.84 0.70 0.31 0.54 0.54 0.34 3 (Durvasula et al., 1993) US 0.74 0.75 0.74 0.76 0.48 0.48 0.55 0.53 NZ 0.75 0.59 0.70 0.82 0.50 0.71 0.66 0.58 4 (Lysonski et al., 1996) NZ 0.8 0.59 0.75 0.82 0.71 0.66 0.54 GR 0.6 0.68 0.63 0.61 0.64 0.55 0.34 US 0.72 0.63 0.75 0.85 0.68 0.69 0.62 IN 0.61 0.71 0.72 0.45 0.41 0.64 0.51 5 (Fan & Xio, 1998) CN 0.61 0.59 6 (Hiu et al., 2001) CN 0.68 0.37 0.65 0.72 0.51 0.62 0.40 7 (Walsh, Thurau, et al., 2001) DE 0.77 0.48 0.71 0.42 0.61 0.76 8 (Bakewell & Mitchell, 2004) UK 0.27 0.76 0.73 0.36 0.36 0.26 0.64 0.09 | 11 9 (Wickliffe, 2004) US 0.65 9 0.84 2 0.71 8 KR 0.83 9 0.563 0.62 2 10 (Mitchell & Walsh, 2004) DE Male 0.76 0.76 0.69 0.71 Femal e 0.77 0.79 0.73 0.69 0.71 0.79 11 (Tai, 2005) CN 0.67 0.66 0.68 0.63 12 (Bakewell & Mitchell, 2006) U K Male 0.47 0.76 0.73 0. 56 0.36 0.26 0.64 0.09 Femal e 0.64 0.76 0.79 0.38 0.39 0.48 0.71 0.43 UK 0.74 0.75 0.74 0.76 0.48 0.55 0.53 ? 13 (Wesley et al., 2006) US 0.80 0.70 0.45 0.77 0.77 0.69 0.62 14 (Yang & Wu, 2007) TW 0.83 0.74 0.79 0.74 0.71 0.76 15 (Hanzaee & Aghasibeig, 2008) IR 0.57 0.79 0.80 0.72 0.62 0.25 0.71 0.42 16 (Yasin, 2009) TR 0.77 5 0.82 1 0.84 0 0.849 0.720 0.69 9 0.84 5 0.684 17 (Mokhlis & Salleh, 2009) MY 0.77 0.67 0.3 0.61 18 (Zhou et al., 2010) CN 0.86 0.81 0.77 0.76 0.69 0.68 0.73 0.70 19 (Kasper et al., 2010) NL 0.85 5 0.85 3 0.812 0.687 0.610 20 (Nayeem, 2012) AU 0.64 9 0.67 1 0.438 0.689 0.73 8 0.731 Key: CN – China, DE- Germany, GR- Greece, IN -India, IR-Iran, KR- Korea, MY- Malaysia, NL- Netherlands, NZ-New Zealand, TR-Turkey, TW-Taiwan, UK- The United Kingdom, US-United States of America In addition, researchers have used different Cronbach alpha cut-off points for testing CSI reliability. Some adopted the Sproles and Kendall (1986) cut-off point of 0.4, while others have used 0.7 (Hair, Black, and Babin, 2006), Nunnally and Bernstein (1994), and Robinson, Shaver, and Wrightsman (1991) (Table 2.2a). However, no reasons were given for researchers choosing a particular cut-off point rather than just indicating that it is recommended. This could be why some researchers decided to have their own cut-off points. For example, some chose a 0.5 cut-off point (Bakewell & Mitchell, 2006; Fan & Xio, 1998 ; Hanzaee & Aghasibeig, 2008; Hiu et al., 2001; Mitchell & Bates, 1998; Mokhlis & Salleh, 2009; Tai, 2005; Zhou et al., 2010), while others 0.6 (Bakewell & Mitchell, 2004; Kasper et al., 2010; Kwan et al., 2008; Sinkovics et al., 2010; Solka et al., 2011). Although this research generally aimed for Nunnally’s cut-off point, it is important to point out that due to the multidimensional nature of the CSI scale some of the alphas are below Nunnally’s cut-off point. Regarding sample size, Hair, Anderson, Tatham, and Black (1998) suggest a sample of 100 or more. To validate Hair at al’s (1998) findings, Reise, Waller, and Comrey (2000) generated a sampling guide indicating the following size quality: 100 being poor, 200 as reasonable, 300 as decent, 500 being very good; and 1000 and above as excellent. Yet, Sapnas and Zeller (2002) found that even with a sample size of 50 it is sufficient to conduct factor analysis. As seen above, the sample size required to complete a factor analysis varies significantly, which is not helpful for researchers (Williams, Brown, & Onsman, 2012). It was further observed that differences in content, language, the number of items, and factors in the data collection instrument have also contributed to the different results regarding CSI reliability (from 17 to 44 per instrument; from 4 to 15 factors), see Table 2.2a below. Another concern is the mismatch between the objectives of some of these studies, and the methodological analysis approaches, causing challenged research results. Mitchell and Walsh (2004) used exploratory factor analysis and did not consider confirmatory factor analysis to examine the validity of the CSI as an instrument designed to measure CDMS with German male and female consumers as participants. As a result, only six factors were confirmed, despite the lowest cut-off point of 0.4. And in turn nine extra factors were formed. However, in later studies among these fifteen factors, only six studies considered these new factors and two studies did not confirm them (Table 2.2a and Appendix I). What is required is a better CSI confirmation test, with consistency in sample sizes, instruments, analysis, and a reliable/consistent Cronbach alpha cut-off point. | 12 Table 2.2a Sample, instrument, analysis, and Cronbach Alpha In addition, despite its popularity, few studies confirmed all eight CSI factors, such as in the US, Tanzania, China, and Taiwan (Chen et al., 2012; Durvasula et al., 1993 ; Sproles & Kendall, 1986; Zhou et al., 2010), while the majority of studies supported between six and seven factors (Kwan et al., 2008; Sinkovics et al., 2010), and only one study rejected all eight CSI factors (Wickliffe, 2004). Those studies which confirmed all 3 KR (Hafstrom et al., 1992) 310 college students CSI scale, 44 items, 5-point Likert scale As Sproles and Kendall (1986) 0.4 7 1 4 NZ (Durvasula et al., 1993) 210 undergrad students Sproles and Kendall (1986) As Sproles and Kendall (1986) 0.4 8 0 5 GR IN NZ US (Lysonski et al., 1996) 486 undergrad students CSI scale As Sproles and Kendall (1986) 0.7 7 1 6 CN (Fan & Xio, 1998) 271 undergrad students CSI scale, 7-factor, 40 items, 5-point Likert scale As Sproles and Kendall (1986) 0.5 1 7 7 UK (Mitchell & Bates, 1998) 401 undergrad students CSI scale, 10-factor 38 items, 5-point Likert scale As Sproles and Kendall (1986) 0.5 8 0 8 CN (Hiu et al., 2001) 381 adult consumers CSI scale, 8-factor, 40 items, 5- point Likert scale Exploratory and confirmatory factor analysis. 0.5 7 1 9 DE (Walsh, Thurau, et al., 2001) 455 male and female CSI scale As Sproles and Kendall (1986) 0.4 6 2 10 UK (Bakewell & Mitchell, 2004) 244 female undergrads CSI scale As Sproles and Kendall (1986) 0.6 3 5 11 DE (Mitchell & Walsh, 2004) 358 German shoppers CSI scale, 4 common factors, 22 items, 5 Male factors of 19 items, 5- Female factors of 17 items As Sproles and Kendall (1986) 0.4 6 2 12 CN (Wang et al., 2004) 431 adults in Guangzhou CSI scale, 7-factor, 18-items, 5- point Likert scale MANOVA, then canonical discriminant analysis 8 0 13 KR, US (Wickliffe, 2004) 126 American factory workers and students, 156 Korean factory workers and students CSI scale As Sproles and Kendall (1986) 0.7 0 8 14 CN (Tai, 2005) 148 Hong Kong, 126 Shanghai CSI scale As Sproles and Kendall (1986) 0.5 4 4 15 UK (Bakewell & Mitchell, 2006) 480 undergraduate students aged 18-22 years CSI scale, 38-items, 8- factors, new 4-male factors, 3-female factors, 5- point Likert scale Principal component analysis with an orthogonal rotation 0.5 8 0 16 US (Wesley et al., 2006) 527 adult consumers CSI scale, 8-factor, 39-items, 5- point Likert scale Exploratory Data Analysis (EDA) 0.4 8 0 17 TW (Yang & Wu, 2007) 472; 240 females, 232 males. about 20–30 years old, with college education 40-item CSI, 5 five-point scale EFA, principal components analysis with varimax rotation and eigenvalue 0.7 6 2 18 IR (Hanzaee & Aghasibeig, 2008) 354 female and 338 male undergraduate students CSI scale, 40-items Principal component analysis with varimax (orthogonal) rotation. Kaiser– Mayer–Oklin (KMO) used to measure sampling adequacy Factor analysis. 0.5 7 1 19 CN (Kwan et al., 2008) 264 undergrads in Beijing, Shanghai, Guangzhou, Hong Kong and Taipei CSI scale, 8-factor, 40 items, 5- point Likert scale Confirmatory and exploratory factor analyses, were employed. 0.6 6 2 20 MY (Mokhlis & Salleh, 2009) 419 undergrad students CSI scale, 8-factor, 40, 5- point Likert scale Factor analysis with principal component 0.5 8 0 21 NL (Kasper et al., 2010) 203 Dutch mobile phone users CSI scale, 8-factor, 41 items. 5- point Likert scale Cluster analysis using Mancova 0.6 5 3 22 AT (Sinkovics et al., 2010) 225 general public Austrian consumers CSI, scale 6-factor, 54-items, 5- point Likert scale Descriptive analyses, Factor analysis 0.6 8 0 23 CN (Zhou et al., 2010) coastal 195 students (114 females and 81 males), inland, 245 students (152 females and 90 males) 7-point Likert scale, 39 items from Sproles and Kendall correlation analysis, 39 items, confirmatory factor analysis 0.5 8 0 24 PL (Solka et al., 2011) 188 Polish students and 208 Americans 5-factor model of 41 items, 5- point Likert scale principal component factor analysis” 0.6 2 6 Key: CN – China, DE- Germany, GR- Greece, IN -India, IR-Iran, KR- Korea, MY- Malaysia, NZ-Tanzania, TR-Turkey, TW-Taiwan, UK- United Kingdom, US-United States of America | 13 of the eight factors came from different countries and cultures such as China, the US, and Taiwan. This could mean that culture may have a limited effect on consumer decision-making styles. Also it has been observed that within the same country and culture, the CSI has given different results, as exhibited with some studies done in Taiwan and the US (Chen et al., 2012; Cowart & Goldsmith, 2007; Yang & Wu, 2007). These mixed findings concerning the CSI may be a sign that there is a set of mixed factors that influence the diverse results regarding CDMS applicability and generalisability, or CSI sensitivity to sampling and methodological approaches. Most studies were carried out on students, for the reason that it is good to test the instrument using comparable or matched samples to demonstrate whether the CSI can be applied across nations on a similar demographic category and give similar results (Lysonski et al., 1996). However, the CSI model should not only be limited to students. Thus, it is necessary that the CSI be tested on non-student samples if the instrument is to be applied to the general population. Moreover, in order to achieve generalisability, more refinement and development of the CSI scale is needed, rather than developing a new scale altogether (Mitchell & Bates, 1998). This is because it is vital to establish the applicability of the CSI to different contexts and societies for it to achieve international validity and reliability (Yasin, 2009), as there is no such instrument yet in place (Yasin, 2009). That is why further research on application and validation of the CSI scale across cultures and populations is encouraged (Sproles & Kendall, 1986). Studies discussed so far have explored the dynamics of validity, applicability, reliability, and generalisability of the CSI as a consumer decision-making styles instrument. The discussions above have shown that CSI validity, reliability, applicability, and generalisability is dynamic and complex. Despite the non-research factors affecting the CSI, some research factors, such as the Cronbach alpha cut-off point differences (i.e. 0.4, or 0.5, or 0.6, or 0.7), have resulted in different outcomes, as discussed above. This has an effect on validity and reliability of the CSI scale. Therefore, researchers need to develop one standard Cronbach alpha for testing a scale. In the concluding remarks of some studies, researchers have commented on the CSI’s validity, reliability, applicability, and generalisability. These remarks are summarised in Table 2.2b below: | 14 Table 2.2b: Researchers concluding remarks on the CSI’s validity, reliability, applicability, and generalisability R EL IA B IL IT Y T A B LE C o u n tr y R ef er en ce P er fe ct io n is t B ra n d C o n sc io u s N o ve lt y- Fa sh io n C o n sc io u s R ec re at io n al , h ed o n is ti c co n su m er P ri ce -V al u e C o n sc io u s Im p u ls iv en es s C o n fu se d b y O ve r- ch o ic e H ab it u al , B ra n d L o ya l Q u al it y C o n sc io u s V ar ie ty -S ee ki n g En jo ym en t- V ar ie ty S ee ki n g R ec re at io n al -H ed o n is ti c P ri ce C o n sc io u s Ti m e- En er gy C o n se rv in g Ti m e c o n sc io u s B ra n d L o ya l St o re L o ya l In fo rm at io n U ti liz at io n Sa ti sf yi n g Fa sh io n -s al e se e ki n g Ti m e r es tr ic te d Ec o n o m y se e ki n g 1 US (Sproles & Kendall, 1986) * * * * * * * * 2 US (Sproles & Sproles, 1990) * * * * * * * * 3 KR US (Hafstrom et al., 1992) * * x * * * * * * 4 NZ (Durvasula et al., 1993) * * * * * * * * 5 GR IN NZ US (Lysonski et al., 1996) * * * * x * * * 6 CN (Fan & Xio, 1998) x * x x x x x x * * * * 7 UK (Mitchell & Bates, 1998) * * * * * * * * * * 8 CN (Hiu et al., 2001) * * * * * x * * 9 DE (Walsh, Thurau, et al., 2001) * * * * x * * x 10 UK (Bakewell & Mitchell, 2004) * * * * * * * * * * 11 DE (Mitchell & Walsh, 2004) * * * * x * * x * * * * * * * * * (Wickliffe, 2004) x x x x x x x x 12 UK (Bakewell & Mitchell, 2006) * * * * * * * * * x 13 US (Wesley et al., 2006) * * * * * * * * 14 TW (Yang & Wu, 2007) * * * x x * * * 15 IR (Hanzaee & Aghasibeig, 2008) * * * * * * * * * * 16 CN (Kwan et al., 2008) * * x * * * * * 17 MY (Mokhlis & Salleh, 2009) * * * * * * * × x * x x 18 NL (Kasper et al., 2010) * * x * * x x * * * 19 AT (Sinkovics et al., 2010) * * * * * * 20 CN (Zhou et al., 2010) * * * * * * * * 21 PL US (Solka et al., 2011) x * x x * x x x * * * No . Count ry Reference Study Objective Sample Instrument Analysis Results-Conclusion 1 US Sproles, G.B., & Kendall. 1986. A methodology for profiling consumers’ decision-making styles. Journal of Consumer Affairs 20 (2). 267-279 method for measuring characteristics of CDMS 501 US high school students 8-factor method of 48 items, 5-point Likert scale The principal component method with varimax rotation of factors, communality estimates of 1.0. A constrained 8- factor solution was extracted to test the 8 characteristics model CSI is useful for consumer-interest professionals. Further application and validation of the CSI across the population is encouraged. 8 factors of 40 items 2 US E.K. Sproles and Sproles (1990) To explore the relationships between individuals’ learning styles and their CDMS 501 US high school students Sproles and Kendall (1986) Sproles and Kendall (1986) Found statistically significant relationships between learning and decision-making characteristics 3 KR Hafstrom, J.L., Chae, J.S., & Chung, Y.S. (1992). Consumer decision-making styles: Comparison between the United States and Korean young consumers. The Journal of Consumer Affairs, 26(1), 114- 122. To identify CDMS of young Koreans and find if they are similar to those of US consumer 310 college students in Korea Sproles and Kendall (1986). 44 items, 5- point Likert scale Factor analysis, principal component method varimax rotation, 8-factor solution (for comparison) The observed generality of several CDMS of young US. and Korean consumers. CSI has elements of construct validity and usage potential across nations 4 NZ Durvasula, Srinivas; Lysonski, Steven; Andrews, J. Craig (1993). Cross-cultural generalizability of a scale for profiling consumers' decision-making styles. Journal of Consumer affairs, 27(1). 55- 65 To test the generalisability of CSI in Tanzania 210 undergrad students at a large university in Tanzania Sproles and Kendall (1986). Sproles and Kendall (1986). Similarities outweigh the differences hence provided general support for CSI | 15 5 GR, IN, NZ US Lysonski, S., Durvasula, S., & Zotos, Y. (1996). Consumer decision-making styles: Multi- country investigation. European Journal of Marketing, 30(12), 10- 21. To investigate the Consumers decision- making profiles of four diverse countries 486 Undergrad . students from GR, IN, NZ, US Sproles and Kendall (1986) 40 items, 5- point scale same as Sproles & Kendall Same method as Sproles and Kendall’s (1986) Confirmed 7 factors out of 8 with 34 items. CSI requires additional psychometric work before it can be applied to other countries, mainly the less developed. 6 CN Fan, J. X. & Xio, J. J. (1998). Consumer decision-making styles of young-adult Chinese. Journal of Consumer Affairs, 32(2), 275- 294. To examine dimensions and profiles of Chinese CDMS compared to American and Korean 271 undergrad. students in China Sproles and Kendall (1986). 7-factor model of 40 items 5-point Likert scale Same method as Sproles and Kendall’s (1986) The consumer decision- making styles are similar in the three countries, but the maturity of the consumer market may impact the differences in CDMS. 5 factors of 31 items 7 UK Mitchell, V.W. & Bates, L. (1998). UK consumer decision-making styles. Journal of Marketing Management, 14(1-3), 199-225. To examine the generalisabity of Sproles and Kendall’s (1986) CSI in an extension work in the UK 401 undergrad students in the UK Sproles and Kendall(1986) 10-factor model 38 items 5-point Likert scale Same method as Sproles & Kendall’s (1986) Most of the original US traits were found in the UK, the addition of new store-loyalty and time- energy saving traits. The CSI is sensitive enough and able to assess cultural differences and produce sensible results. 8 CN Hiu, A. S. Y., Siu, N. Y. M., Wang, C. C. L. and Chang, L. M. K. (2001), “An Investigation of Decision-Making Styles of Consumers in China,” Journal of Consumer Affairs, Vol. 35, No. 2, pp. 326-345. To investigate Chinese CDMS 381 adult consumers in China Sproles and Kendall (1986). Double analysis method, 8-factor model of 40 items 5- point Likert scale Exploratory and confirmatory factor analysis. Cluster analysis for determining market segment in the future Five CDMS are valid and reliable in Chinese culture (perfectionist, novelty- fashion conscious, recreational, price conscious, and confused by over-choice. 7 factors and 5 market segments derived 9 DE Walsh, G., Wayne-Mitchell, V., & Hennig-Thurau, T. (2001). German consumer decision- making styles. The Journal of Consumer Affairs, 35(1), 73-8 To test the generalizability of CDMS in different countries and with non- student German shoppers 455 German male and female shoppers (eighteen and older) Sproles and Kendall (1986) Sproles and Kendall (1986) supported six factors only 10 UK Bakewell, C., & Mitchell, V. (2003). Generation Y female consumer decision-making styles. International Journal of Retail & Distribution Management, 31 (2), 95-106. Examine the decision making of adult female generation Y consumers 244 Female undergrad uate students in the UK Sproles and Kendall (1986) Sproles and Kendall (1986) Shoppers change as a function of their generation membership due to macro- environmental influences and 5 decision-making groups emerged 11 DE Mitchell & Walsh. 2004. Gender differences in German consumer decision-making styles. Journal of consumer behaviour. 3 (4). 331-346 To examine the validity of an instrument designed to measure CDMS of German male and female consumers 358 German shoppers Sproles and Kendall (1986) 4 common factors model of 22 items, 5 Male factors of 19 items, 5- Female factors of 17 items Exploratory principal component method with varimax rotation of factors Five new male factors (satisfying, enjoyment- variety seeking, fashion- sale seeking, time restricted and economy seeking). CSI has constructed validity for females, but not males. 12 CN Cheng-Lu Wang, Noel Y.M. Siu, Alice S.Y. Hui, (2004),"Consumer decision-making styles on domestic and imported brand clothing," European Journal of Marketing. 38 (1). 239 - 252 To investigate the relationship between Chinese CSI and their choice between domestic and imported clothing brands 431 adult Chinese in Guangzho u Sproles and Kendall (1986), 7-factor, 18- items, 5- point Likert scale Began with the multivariate analysis of variance (MANOVA), followed by canonical discriminant analysis General support for the usefulness of purified CSI in understanding Chinese CDMS in relationship to consumers’ preference for domestic or imported clothing brands. 13 KR, US Wickliffe, V.P. (2004). Refinement and reassessment of the consumer decision-making style instrument. Journal of Retailing and Consumer Services, 11, 9-17. To examine the psychometric properties of a popular the instrument used to 126 American factory workers and students Sproles and Kendall (1986) Sproles and Kendall (1986) CSI not a reliable or valid measure of CDMS for both Korea and the US. The confused impulsive consumer was the new | 16 measure CDMS and its findings were compared to earlier studies 156 Korean factory workers and students construct and in contrast with previous studies. 14 CN Tai, S. (2005). Shopping styles of working Chinese female. Journal of Retailing and Consumer Services, 12, 191-203. To create a typology of the shopping style dimensions of working female consumers aged 18- 44 in Shanghai and Hong Kong 148 Hong Kong 126 Shanghai Sproles and Kendall (1986) Sproles and Kendall (1986) Identified 10 CDMS relevant to Chinese working females and four new non-CSI dimensions (personal style consciousness, environment and health consciousness, reliance on mass media, and convenience and time consciousness) 15 UK Bakewell, C., & Mitchell, V.-W. (2006). Male versus female consumer decision-making styles. Journal of Business Research, 59(12), 1297-1300. To investigate male and female CDMS a non- probability sample of 245 male and 245 female undergrad uate students aged 18- 22 years (usable items 480) Sproles and Kendall (1986), 38-items, 8- common factors, 4-male factors, 3- female factors 5- point Likert scale Principal component analysis with an orthogonal rotation All 8 US original CSI were confirmed and largely- female, decision-making traits. 16 US Wesley, S., LeHew, M., & Woodside, A. G. (2006). Consumer decision-making styles and mall shopping behavior: Building theory using exploratory data analysis and the comparative method. Journal of Business Research, 59(5), 535- 548. To assess the relationship between CDMS and shopping malls behaviour 527 adult consumers aged 18 to 85 plus Sproles and Kendall (1986), 8-factor, 39- items, 5- point Likert scale adopted Exploratory Data Analysis (EDA) Empirical research supported CDMS existence among adult shoppers in different mall contexts. Gender is a prime antecedent associated CDMS. CDMS influence on mall shopping indirect Perfectionist consumers are ranked high in planned mall expenditures 17 AT Sinkovics, R. R., ‘Mink’ Leelapanyalert, K., & Yamin, M. (2010). A comparative examination of consumer decision styles in Austria. Journal of Marketing Management, 26(11-12), 1021-1036. To examine and compare CDMS in Austria and previous CSI studies in other countries (Replica for generalisation) To test the CSI’s explanatory power in a sample drawn from general public 225 Austrian consumers , from the general public Sproles and Kendall (1986), 6-factor, 54- items, 5- point Likert scale Descriptive analyses, Factor analysis (principal components, varimax rotation). Results are highly congruent with findings from earlier studies using student samples. 18 CN Zhou, J. X., Arnold, M. J., Pereira, A., & Yu, J. (2010). Chinese consumer decision-making styles: A comparison between the coastal and inland regions. Journal of Business Research, 63(1), 45-51. to develop a better understanding of the variations in CDMS between coastal and inland China coastal sample of 195 students (114 females and 81 males). inland sample, 245 students (152 females and 90 males) 7-point Likert scale (1=strongly disagree to 7 = strongly agree). 39 items from Sproles and Kendall (1986) “An item-total correlation analysis of the 39 items revealed that 4 Correlation and a multi-group confirmatory factor analysis to assess the measurement invariance between the two groups consumers in the two regions are similar in utilitarian shopping styles and differ in hedonic shopping styles. China is heterogeneous rather than homogeneous market | 17 Throughout the discussion, it has been observed that the CSI is not yet universally developed. This calls for a wider investigation into the matter. Therefore, this study will investigate the currency, validity, reliability, applicability, and generalisability of the CSI scale in respect of today’s consumers. Accordingly, the following section discusses the general picture of CSI profiles, features, and dimensions as a foundation upon which to develop the green consumption dimension about consumers’ decision-making styles. 2.4 CSI profiles and dimensions This section will discuss the features, profiles, and dimensions of the CSI’s eight factors in different contexts. These factors are Perfectionism; Brand consciousness; Novelty-fashion consciousness; Recreational/Hedonistic shopping consciousness; Price & value-for-money consciousness; Impulsiveness; Confusion from over-choice; and Brand Loyalty (Sproles & Kendall, 1986). 2.4.1 Perfectionism Sproles (1985), Sproles and Kendall (1986), and Wesley et al. (2006) define perfectionism in the CSI as a situation whereby a consumer searches for the best quality in products; shops more carefully, more systematically, or by comparison; and is not satisfied with the “good enough” product. Perfectionists are likely to be highly satisfied with their purchases because they tend to plan their expenses (Wesley et al., 2006). Perfectionists are neither brand nor price loyal (Kasper et al., 2010; Wesley et al., 2006); they are only loyal to quality. Perfectionist consumers identify and look for specific qualities in a product based on the information search they conducted before deciding to buy (Wesley et al., 2006). Further, perfectionists do not buy before they have comprehensively evaluated the product they want and are satisfied with it (Kasper et al., 2010). Sproles (1985); Sproles & Kendall (1986) demonstrated that perfectionist consumers also have a tendency to intensify their shopping processes and time in order to obtain the maximum utility. Mitchell and Walsh (2004); Wiedmann, Walsh, and Mitchell (2001); and Yasin (2009) report that female consumers tended to be more perfectionist in practice than males. However, this research did not explain why women are like that. Further, perfectionist consumers exhibit utilitarian buying characteristics (Kim, Yang, & Lee, 2009; Zhou et al., 2010), which makes them more functional than emotional buyers. Hence, marketers can take advantage of this perfectionist characteristic by focusing on quality, functionality, 19 Polan d, US Solka, A., Jackson, V. P., & Lee, M.-Y. (2011). The influence of gender and culture on Generation Y consumer decision making styles. The International Review of Retail, Distribution and Consumer Research, 21(4), 391- 409. To examine gender and culture as predictors of CDMS. 188 Polish students and 208 Americans Sproles and Kendall (1986), 5-factor model of 41 items, 5- point Likert scale principal component factor analysis” Found 4 out of 5 shopping characteristics to be different between Poland and the US (enjoyment, shopping aversion, price consciousness and quality consciousness) and 3 out 5 differ between genders (enjoyment, shopping aversion and brand consciousness). | 18 effectiveness, and practicality as their unique selling points towards these types of customers as well as use market experts to communicate a product of high quality, prestige, and self-esteem (Wiedmann et al., 2001). However, for marketers to take advantage of this situation, identifying and comprehending the theoretical explanation behind perfectionism as a consumer decision-making style is necessary as researchers and marketers would then better understand perfectionism in the CSI instrument. Durvasula et al. (1993) and Wickliffe (2004) have indicated that consumer perfectionism is one of the most stable CSI characteristics, and this has been confirmed in many countries such as Tanzania, Australia, the US, Korea, Greece, India, China, the UK, continental Europe, and in many other countries at different times (Canabal, 2002b; Durvasula et al., 1993 ; Hafstrom et al., 1992; Hiu et al., 2001; Lysonski et al., 1996 ; Mishra, 2010; Sinkovics et al., 2010; Sproles , 1985; Sproles & Kendall, 1986). However, investigations in countries like China, Korea, and Poland did not confirm perfectionism factor (Cowart & Goldsmith, 2007; Fan & Xio, 1998 ; Solka et al., 2011; Wickliffe, 2004). Table 2.3 Studies that do not support Perfectionism It can also be observed that some countries have confirmed perfectionism in one part, while other parts of the same country rejected perfectionism, as seen in Table 2.3 above. This may mean that people from the same country may have different orientations, impressions, attitudes, and perceptions regarding consumer perfectionism. Researchers and marketers may also be confronted by the question of what can be done to cope with such a situation in these countries. Sometimes these factor rejections may be caused by sampling issues. For example, Wickliffe (2004) exampled a sample of 286 respondents (Table 2.4), below the norm for a multi-country study, which requires a minimum of 400 usable responses on average (Lysonski et al., 1996 ; Solka et al., 2011). Also, this sample included irrelevant respondents (Korean students studying in the US), while the study indicated that the focus was on American students and American factory workers in America, rather than Korean students and Korean workers in Korea. N o . C o u n tr y R ef er en ce P er fe ct io n is t B ra n d C o n sc io u s N o ve lt y- Fa sh io n C o n sc io u s R ec re at io n al , h ed o n is ti c co n su m er P ri ce -V al u e C o n sc io u s Im p u ls iv en es s C o n fu se d b y O ve r ch o ic e H ab it u al , B ra n d L o ya l Q u al it y C o n sc io u s V ar ie ty -S ee ki n g En jo ym en t- V ar ie ty S ee ki n g R ec re at io n al -H ed o n is ti c P ri ce C o n sc io u s Ti m e- En er gy C o n se rv in g Ti m e c o n sc io u s St o re P ro m is cu o u s St o re L o ya l In fo rm at io n U ti liz at io n Sa ti sf yi n g Fa sh io n -s al e se e ki n g Ti m e r es tr ic te d Ec o n o m y se e ki n g Im p er fe ct io n is m B ar ga in S ee ki n g Lo w p ri ce s ee ki n g C ar el es s C o n su m er 1 CN (Fan & Xio, 1998) x ✓ x x x x x x ✓ ⚫ ⚫ ⚫ ✓ ⚫ ✓ ⚫ ⚫ ✓ ⚫ ⚫ ⚫ ⚫ ⚫ ⚫ ⚫ ⚫ 2 US (Cowart & Goldsmith, 2007) x ✓ ✓ ✓ x ✓ x ✓ ⚫ ⚫ ⚫ ⚫ ⚫ ⚫ ⚫ ⚫ ⚫ ⚫ ⚫ ⚫ ⚫ ⚫ ⚫ ⚫ ⚫ ⚫ 3 KR US (Wickliffe, 2004) x x x x x x x x ⚫ ⚫ ⚫ ⚫ ⚫ ⚫ ⚫ ⚫ ⚫ ⚫ ⚫ ⚫ ⚫ ⚫ ⚫ ⚫ ⚫ ⚫ 4 PL US (Solka et al., 2011) x ✓ x x ✓ x x x ✓ ⚫ ✓ ⚫ ⚫ ⚫ ✓ ⚫ ⚫ ⚫ ⚫ ⚫ ⚫ ⚫ ⚫ ⚫ ⚫ ⚫ Key: (✓) = CSI factor is supported, (x) = CSI factor not supported, (⚫) = Factor not considered , CN-China, US- United States of America, KR-Korea, PL-Poland | 19 Table 2.4 The Wickliffe study summary Kasper et al., (2010); and Sproles (1985) also indicated that perfectionists are less confused by over-choice and information overload because if a product does not meet their criteria it is dropped off the list. As a result, the perfectionist approach to consumer decision-making has been viewed as an effective way to shield the consumer against confusion and information overload, and of downsizing the set of considerations (Kasper et al., 2010; Mitchell, Walsh, & Yamin, 2004; Wiedmann et al., 2001). For perfectionists are regarded as knowledgeable consumers (Kasper et al., 2010; Sproles & Sproles, 1990), with a probability that they can make an informed buying decision. The more knowledgeable the consumer is, the less their confusion, and the higher the probability of uncovering buying risks. Therefore, marketers should give quality a higher priority when dealing with perfectionists. Marketers should also be aware that this type of customer can handle information overload, and confusion, and can challenge marketers because they are knowledgeable and follow a serious systematic approach to learning (Sproles & Sproles, 1990). On the other hand, there is a concern that the information consumers search for, to help them make decisions, can be biased, manipulated, inadequate, concealed (patented information) or miscomprehended by the consumer, and hence imperfect decisions with undesirable outcomes. This can cast a shadow on the concept of a perfectionist consumer decision-making style. High-income earners tend to exhibit perfectionism more than their counterparts (Wesley et al., 2006) because a high income empowers them to demand and afford quality. However, Wang et al., 2004 have shown that people of lower incomes may also demand high quality, not just the rich. In addition, Baoku, Cuixia, and Weimin (2010) concluded that even in poor rural areas there are perfectionists. This could be a sign that perfectionism is not only for the rich and affluent. Also, the original CSI was tested with students, who in an economic sense are dependents on their parents or guardians, yet they were able to show perfectionism. Therefore, it can be inferred that economic status has limited influence on perfectionism. When it comes to social class in relation to perfectionism, research shows that the higher the social class the higher the perfectionism (Shim, 1996). Bakewell & Mitchell, 2006 concluded that perfectionism is one of the ways of showing superiority where perfectionists classed as high-end consumers tend to spend more than average (Wesley et al., 2006). Coun try Reference Study Objective Sample Instrument Analysis Results-Conclusion KR, US (Wickliffe, 2004) To examine the psychometric properties of a popular instrument used to measure CDMS and its findings were compared to earlier studies 126 American factory workers and students 156 Korean factory workers and students Sproles and Kendall (1986) CSI Sproles and Kendall (1986) Factor Analysis CSI not a reliable or valid measure of CDMS for both Korea and the US. The confused impulsive consumer was the new construct and in contrast with previous studies. Key: KR- Korea, US-United States of America | 20 Perfectionist consumers seem to be more responsible shoppers, and more rational than emotional (Kasper et al., 2010) (even though it is not known yet about their green consumption behaviour). This could be due to the fact that consumer education has a direct influence on them (Shim, 1996), which makes them highly knowledgeable and responsible shoppers (Sproles & Sproles, 1990; Sproles , 1985; Sproles & Kendall, 1986; Wang et al., 2004). However, it is not only consumer education that influences perfectionism; some researchers have indicated that collectivist societies are likely to be perfectionists (Doran, 2002) for any of the following three reasons: to show off; being frugal with their finances; or peer influence (Shim, 1996). Bakewell and Mitchell (2006) has shown that most men are perfectionists in order to show their superiority. In addition, Baoku et al. (2010) reported that even uneducated poor peasants can be perfectionists, which means that perfectionism is not influenced by education only, but by a combination of different factors such as the market and macro environments (Kwan et al., 2008). This is to say, the perfectionist consumer may seem more responsible and rational due to their high knowledge, but one should not ignore other factors that influence the perfectionism decision-making style, as discussed above. 2.4.2 Brand consciousness Consumers who prominently exhibit brand consciousness are oriented toward buying expensive, well- known brands (Shim, 1996). They also believe that higher price reflects higher quality, and have a positive attitude towards pricey high-end stores with bestselling advertised brands (Sproles & Kendall, 1986). In addition, brand conscious–oriented consumers use price and brand as a sign of quality, prestige, and superiority (Bakewell & Mitchell, 2006; Fan & Xio, 1998 ; Wang et al., 2004), and they may be the attributes to higher prices (Forsythe, 1991). Some brand-conscious consumers use brand-conscious orientation to convey fashion, image, and meaning, particularly those from individualistic cultures (Bao, Zhou, & Su, 2003; G.-S. Kim, Lee, & Park, 2010; Park & Rabolt, 2009). Most brand-conscious consumers use brand and price as symbols of status and prestige and are common in areas where there is a high-power distance culture (Hofstede & Hofstede, 2001). It seems that these consumers are also price conscious; however, they see price in a positive, rather than a negative way. It is unknown why they behave in such a manner. Brand-conscious people do not respond well to consumer education and learning and they lean much towards hedonism (Shim, 1996; Sproles & Kendall, 1987; Zhou et al., 2010). This may pose a challenge to consumer educators, counsellors, advisors, and guardians.There seems to be some crossover between brand consciousness and perfectionism (Wesley et al., 2006; Wickliffe, 2004), which can cause some confusion when it comes to precisely differentiating these two dimensions of the CSI. For example, both brand-conscious and perfectionist consumers are highly educated (Wang et al., 2004) and are likely to downsize the set of considerations (Kasper et al., 2010); have high income and plan their expenditures (Wesley et al., 2006); have high sensory innovativeness tendencies, and are not comparison shoppers (Zhou et al., 2010). However, the question remains: how do they choose a brand or service without comparing different alternatives? In this case, it may be true for the price but not for other factors, as brand-conscious consumers are associated with price insensitivity (Warrington & Shim, 2000). | 21 Demographically, urban consumers are more brand conscious than those in rural areas (Zhou et al., 2010). This could be due to the higher presence of media channels used by brands in urban, rather than rural, areas. Also, females seem to be more brand conscious than males (Mitchell & Walsh, 2004; Yasin, 2009). In contrast, when it comes to online buying, male consumers have stronger brand consciousness than women (Sinkovics et al., 2010). However, Bakewell and Mitchell (2006) show that men are as brand conscious as women at equal levels. Yet, Shim (1996) reported mixed results on brand consciousness between boys and girls. Furthermore, (Kasper et al., 2010; Wang et al., 2004 ; Weiss, 2003) reported that youths are more brand conscious than consumers of other ages. On the other hand, (Shim & Gehrt, 1996; Solka et al., 2011) found contradictory results indicating that most youths have a lower level of brand consciousness than others. From the above observations, it can be inferred that demographically things have been changing from mixed results to female brand consciousness, then to males, and later to youth dominance in brand consciousness. Therefore, it can be noted that, demographically there is a trend showing the youth leading on brand consciousness. Hence, marketers can take advantage of this trend to enhance their marketing success within the youth market segment. Other studies have shown that brand consciousness is at a different state of development in different cultures (Lysonski et al., 1996) because culture affects brand consciousness (Leo, Bennett, & Härtel, 2005a). For example, collective culture societies are more brand conscious than individualistic ones (Watson & Wright, 2000). This is why brand consciousness is growing in popularity amongst collective societies, such as with Chinese consumers, as Wang et al. (2004) indicated. The growing popularity of brand consciousness among these collective societies makes it one of the most applicable and stable CSI factors (Durvasula, Lysonski, & Andrews, 1993; Hafstrom et al., 1992; Hiu et al., 2001; Leo et al., 2005a; Lysonski et al., 1996 ; Mitchell & Bates, 1998; Mitchell & Walsh, 2004; Sproles & Kendall, 1986; Walsh, Thurau, et al., 2001; Zhou et al., 2010). However, as an exception, research conducted in Malaysia (Mokhlis & Salleh, 2009a, 2009b) does not confirm the brand consciousness factor. 2.4.3 Novelty-fashion consciousness Consumers who are fashion and novelty conscious are interested and excited in the pleasure of seeking out new things; keeping up-to-date with latest products and styles; and variety seeking (Khare, 2012; Sproles & Kendall, 1986; Wesley et al., 2006). This calls for marketers to stress variety and novelty when marketing to this type of consumer (Michaelidou, 2012). Also, novelty-conscious consumers are more cognitive and visually oriented (Zhou et al., 2010), therefore the marketer may use more visuals when marketing to this kind of customer. However, some research findings give the impression that novelty-fashion-conscious and recreational, hedonistic consumers are indistinguishable (Lysonski et al., 1996). And their argument is that novelty- fashion consumers are more hedonistically inclined, the same as recreational consumers (Babin & Harris, 2009; Zhou et al., 2010). Also, there are features that are found in both groups, such as: being easy going, light-hearted, dreamers, impulsive, and loving pleasure (Zhou et al., 2010). Both groups treat shopping as a | 22 recreational activity (Shim & Gehrt, 1996) and are less concerned with the implications of their purchase of new and novel products (whether it is negative or positive) (Sproles & Sproles, 1990); and they care less about prices (Kasper et al., 2010). Novelty- and fashion-