Buying a better future : a longitudinal qualitative study, exploring the relationship between attitude and green buying behaviour for millennials

(1)

Buying a Better Future

A longitudinal qualitative study, exploring the relationship between attitude

and green buying behaviour for millennials

Author: Christine Florence Hogendoorn Student number: 10126929

30 September 2015 (Final)

MSc. in Business Administration – Strategy and Marketing Track Universiteit van Amsterdam (UvA)

First supervisor: Monika Kakovic Second supervisor: Erik Dirksen

(2)

2 Statement of originality

This document is written by Student Christine Hogendoorn who declares to take full responsibility for the contents of this document.

I declare that the text and the work presented in this document is original and that no sources other than those mentioned in the text and its references have been used in creating it.

The Faculty of Economics and Business is responsible solely for the supervision of completion of the work, not for the contents.

(3)

3

Abstract

Rising consumption worldwide leads to increased pressure on the environment. More and more consumers have realized they can make a positive contribution through the consumption of green products. Green products are produced in a way that they contribute to preserving the environment. Millennials (born 1977 - 1994) are thought to be an influential consumer group and are seen as core supporters of green. However, previous research has shown there is a gap between what they think and do: the attitude-behaviour gap. In this paper, I focus on how the factors awareness, trust, habit, and emotional, social and functional value influence the relationship between millennials’ attitude towards- and their purchasing behaviour of green products, and how this changed between 2012 and 2015.

Key findings show that the attitude-behaviour gap has decreased since 2012, caused by the increased influence of others (social value). Another finding is that awareness is still relatively low. This could be improved by convincing millennials they are ‘doing the right thing’, instead of simply providing information. This research contributes to literature because the influence of these factors is not studied in isolation or at one single point in time, but in relation to the same factor over time, i.e. how has the influence of this factor evolved over time and is it more, less or equally influential? Based on longitudinal qualitative research, a dynamic perspective could be given on the millennials green behaviour, behavioural change and the processes associated with this change.

Keywords: attitude behaviour gap; millennials; purchasing behaviour; green products; CSR; qualitative longitudinal research (QLR)

(4)

4

1. Introduction

Rising consumption worldwide, driven by a steep growth in the world population from approximately 6 billion in 2000 to an expected 8.2 billion in 2030, leads to increased pressure on the environment and natural resources, causing problems such as erosion, climate change, the exhaustion of resources and other environmental damage (Chen & Chai, 2010; McLeay et al., 2012; OECD, 2008; Otmann, 1992). More and more consumers are aware of these issues and in line, researchers have reported on consumers having very positive attitudes towards preventing social and environmental damage (Vermeir & Verbeke, 2006; Bhattacharya & Sen., 2004; Sen & Bhattacharya, 2001; Wagner, Bicen & Hall, 2008; Dastous & Legendre, 2009; Ellen, Webb & Mohr, 2006). Consumers even demand companies to be dedicated to these issues and to act in an environmentally friendly way (Schmeltz, 2012; Bockman, Razzouk & Sirotnik, 2009; Kanarattanavong & Ruenrom, 2009). The concept whereby companies integrate social and environmental concerns in their business operations is called Corporate Social Responsibility (CSR) (European Commission, 2011). CSR initiatives encompass a wide range of activities, varying from corporate philanthropy, minority support programs, green manufacturing practices and producing green products (Peloza & Shang, 2011; Bhattacharya & Sen, 2001; Drumright 1994; Smith 1994). More and more consumers realize one of the most significant ways they can make a positive contribution to prevent or decrease the environmental damage is through their private consumption (Smith and Brower, 2012; Laroche et al., 2001). Consequently, they require from companies to meet these concerns by implementing green manufacturing processes and produce green products. Green products can be defined as products that are produced in a way that they do not pollute the earth or deplete natural resources and contribute to preserving the environment or natural habitat (Erve, 2013; Shamdasami, Chon-Lin & Richmont, 1993). These type of products are non-toxic, energy- and water efficient, they have less packaging to reduce its environmental impact and they can be recycled or are biodegradable (Chan & Chai, 2010). Green products are produced under environmental friendly circumstances. Examples of green products are: biodegradable paint, CFC-free hairspray or unbleached coffee filters (Laroche et al., 2001). However, despite the reported positive attitude consumers have towards CSR, green manufacturing and green products, research indicates a gap between what consumers think and what they do when it comes to actually purchasing green products: the attitude-behaviour gap (Young, 2004; Vermeir & Verbeke, 2006; Chen & Chai, 2010; Kolkailah, Aish & Bassiouny, 2012, Eck, 2009; Anon, 2009; Finisterra, et al., 2009; Basu & Hicks, 2008;). When looking at how consumer purchasing is set to shift in the coming years, the generation of millennials (born between 1977 - 1994) make up for an influential consumer group that is expected to reach their buying peak within the next ten years. Moreover, they are seen as core supporters of green products and services (Heaney, 2007). However, millennials also appear to be “low on trust”, which translates to a

(7)

7 generation that is aware but cynical (Lantieri & Chiagouris 2009). Hume (2010) describes that there is a clear pattern of contradiction between what millennials know and think versus how they act with regards to green purchasing, and they show almost no sign of adopting sustainable practices in terms of actual purchasing. In 2010, Smith conducted a survey amongst 602 millennials, asking this group about their actual buying behaviour of green products. Her findings clearly show a gap between attitude and actual behaviour; millennials are taking note of a company’s reputation, read labels, and care if a product is environmentally friendly, but just half of the millennials in her survey indicate that they sometimes make an effort to actually buy green products and only 10 percent indicates they often make the effort. It appears that, even though millennials seem to attach great importance to companies behaving responsibly and produce green products, this does not always relate directly to their actual purchase behaviour (Öberseder et al, 2011). Gaining insight into this relationship is of importance when it comes to understanding, interpreting, foreseeing and influencing such an important and influential consumer group as the millennials (Bagozzi, 1993; Carrigan et al., 2010). Researchers produced many classifications and reports attempting to generalize the complex and often subjective different theories that explain the relationship between attitude and (green) purchase behaviour (Kollmuss and Agyeman, 2002). Even though there is at present not a single conclusive model, there are some key factors. In this research, a large body of existing literature and frameworks was reviewed, and the factors emotional-, social- and functional value, habit, awareness and trust were identified as possible factors that are said to influence the relationship between attitude and actual green purchasing behaviour. As such, the following research question will be answered:

How has the influence of the factors awareness, trust, habit, and emotional, social and functional values on the relationship between millennials’ attitude towards and their purchasing behaviour of green products changed between 2012 and 2015?

This research question addresses the attitude behaviour gap for millennials, but moves beyond the search for a general correlation between attitudes and behaviour. Instead, the aim is to explain what the influence is of these specific factors and in what direction they influence the relationship between the attitude towards- and the purchasing behaviour of green products. Are all factors positively contributing to the actual purchase, or are some factors negatively influencing the purchase and thereby further increasing the attitude behaviour gap? The contribution to literature that is made in this thesis, is that the influence of these factors is not studied in isolation or at one single point in time but in relation to the same factor over time, i.e. how has the influence of this

(8)

8 factor evolved over time and is it more, less or equally influential as the same construct three years ago?

An appropriate research method to address such questions is qualitative longitudinal research (QLR) (Farrel, 2006). This is because this type of research embodies in-depth interview-based studies conducted at different points in time, which allows to return to the respondents to verify and explore (perceived) changes between the interviews. This way, a dynamic perspective can be created on the millennials behaviour, behavioural change and the processes associated with this change.

A group of self-claimed green millennials were interviewed in 2012 and again in 2015. They were asked about the same topics, both in 2012 and in 2015, and they were also asked themselves to reflect on any changes that occurred. In the final stage, the results from this study were linked back to the literature and compared to prior literature so that conclusions could be drawn.

The following chapter outlines prior literature. First, the general concept CSR, the business case for CSR (why do companies engage in CSR) and different types of CSR will be explored, such as green manufacturing processes and the production of green products. Secondly, the attitude behaviour gap is discussed. Thirdly, millennials and their buying behaviour will be explored, followed by a review of the factors that have been identified as influential (emotional-, social- and functional value, habit, awareness and trust). Chapter three provides an overview of the used research methods, explains how the data was collected and discusses the reliability and validity of this research. This is followed by the results. The discussion, limitations and suggestions for future research are presented in chapter five and the conclusion is presented in chapter six.

(9)

9

2. Theoretical Framework

In this chapter, relevant key literature is reviewed and linked to the main question. The articles reviewed, cover key papers in the areas of CSR, green consumerism and the attitude behaviour relationship. This literature review led to the development of the research question and provides a basis for the topic list that was used during the interviews. The structure of this chapter is set up in a way that is starts with a general description of CRS and gradually links other relevant theories and frameworks to finally come to the identification of factors that are most identified in prior research as having a (potential) influence on the relationship between the millennials attitude towards green consumption and actual green consumption.

2.1 CSR and the business case for CSR

Due to a shift of values in Western parts of society, combined with a revolution in communication technology, traditional roles of businesses, consumers, governments and society changed radically and where traditionally the government had to deal with responsibilities, now companies were also expected to play their part (Schmeltz, 2012; Benabou and Tirole, 2010; Moon and Vogel, 2009; Bockman, Razzouk & Sirotnik, 2009; Kanarattanavong & Ruenrom). In 1953, Howard Bowen was the first one to make a significant scholarly contribution about CSR, which he defined as “the obligations of business to pursue those policies, to make those decisions or to follow those lines of action which are desirable in terms of the objectives and values of our society (Bowen, 1953). Over the subsequent decades, CSR definitions, practices and adoption of CSR expanded immensely. The term CSR is still widely used, although one can find many different names for the same or comparable concepts, such as conscious capitalism, corporate consciousness or corporate citizenship (Peloza and Sheng, 2011; Carrol, 1999; Wood, 1991). Carrol (2007) states that, in the new millennium, ‘companies would be expected to do all the following: demonstrate their commitment to society’s values on social, environmental, and economic goals through their actions; fully insulate society from the negative impacts of company actions; share the benefits of company activities with key stakeholders, as well as shareholders, and demonstrate that the company can be more profitable by doing the right thing’. He concludes that this approach reassures stakeholders that new behaviours will outlast good intentions. In turn, a company that would meet the expectations of the stakeholder, would benefit from many positive effects; the business case for CSR (Hillenbrandt et al., 2011; Weber, 2008).

A stakeholder can be defined as ‘any group or individual who is affected by- or can affect the achievement of an organization’s objectives’ (Freeman and Vae, 2001). Primary stakeholder groups that can be identified are: Employees, customers, suppliers, and shareholders (McElroy, 2012).

(10)

10 According to Smith et al. (2010), consumers are often identified as one of the most important stakeholders driving the CSR business case. As more and more consumers have realized that their private consumption behaviour not only impacts the environment but also is an influential way to affect production, they require from companies to meet these concerns related to products and manufacturing (Smith and Brower, 2012). Consequently, many companies see the production of green products and adjustment of manufacturing processes not only as way to put sustainable practices into action but also a means to gain competitive advantage, because this may lead to positive company evaluations (e.g. Green and Peloza, 2011; Brown and Dacin, 1997), higher purchase intentions (e.g. Green and Peloza, 2011; Mohr and Webb, 2005), resilience to negative information about the organization (e.g. Green and Peloza, 2011; Peloza, 2006), positive word-of-mouth (e.g. Green and Peloza, 2011; Hoeffler and Keller, 2002), and a willingness to pay higher prices by some consumers (e.g. Green and Peloza, 2011; Laroche et al., 2001).

2.2 Green products

About twenty years ago, CSR was limited to corporate charity and, for some businesses, the adherence to environmental legislation (McPherson, 2012). What can be seen today is a far more complex picture, which can mainly be attributed to increased awareness, an ever-widening stakeholder universe, economic development and a growing access to resources and technologies (Rana et al., 2008). In 2005, Ashridge provided a broad overview of what internationally, especially in Europe and North America, constitutes the most common forms of CSR activities that enterprises are undertaking. The most often mentioned elements of CSR in both reports, and the way they are generally measured, are listed in table 1 below.

Table 1. Types, measurement and most common forms of CSR activities (Ashridge, 2005)

In October 2011, the European Commission issued a paper aimed at enhancing the visibility of CSR and disseminating good practices. They provide a list of practices that companies can engage in but

CSR type Example activities

Compliance ethical conduct of staff transparency of

operations no conflicts of interest

compliance with listing rules

compliance with corporate law

Risk Management organisational health &

safety

responsible business

practices regulatory compliance

reputation management

stakeholder engagement

Value Chain responsible procurement supplier management externalisation of costs fair trade supplier screening & feedback

Social human rights community

engagement philanthropic activities staff volunteering local capacity building

Employees turnover satisfaction & work-life balance

internal culture (transparency)

Diversity & equal

opportunity training

Environment environmental footprint waste & pollution

avoidance eco-efficiency product recycling sustainability

Measurement Global Reporting

(11)

11 they should at least cover one of these areas: human rights, labour and employment practices (such as training, diversity, gender equality and employee health and well-being), environmental issues (such as biodiversity, climate change, resource efficiency, life-cycle assessment and pollution prevention), combating bribery and corruption and Sustainable Consumption and Production. With regards to the latter, they mention that companies are responsible to improve the overall environmental performance of products throughout their life-cycle, to boost the demand for better products and production technologies and to help consumers in making informed choices (the European Commission, 2008). Chen & Chai, 2010 add to this that in producing green products less packaging and less toxic materials are used so they can be recycled and environmental impact is minimal. Green products can be classified in four categories (Ethical Consumerism Report, 2010): (1) Green food and drinks: organic, Fair Trade, Rainforest Alliance, free range eggs, free range

poultry and sustainable fish;

(2) Green home: energy efficient light-bulbs, ethical cleaning products, sustainable timber and paper, rechargeable batteries and buying for re-use household products;

(3) Green personal products: clothing and make-up which are produced under environmental friendly circumstances and are non-polluting and free of synthetic dyes or perfumes, buying for re-use clothing and real nappies;

(4) Community: local shopping for green products. Purchasing green products is called: green purchasing behaviour.

The spread of green products has now broadened to the mainstream public (Carrington et al., 2010), and many consumers express their concerns about the environment and our society through their consumption behaviour. Media has also picked up on this concern, and the emergence of organised activist groups (such as Greenpeace; Wakker Dier), political attention (Partij voor de Dieren) and the increased availability of green products, have all led to a growing awareness of the impact of our modern consumption behaviour (Connolly and Shaw, 2006; Crane and Matten, 2004). Why is it then, that despite this apparent attention, rising awareness and demand, there still appears to be a gap between attitudes and actual green purchasing behaviour? This question will be explore in the next chapters.

2.3 The Attitude behaviour gap

Over the past decade, many researchers from various disciplines (psychology, sociology, economics, marketing, health) have tried to close the attitude behaviour gap, but the outcomes remain fragmented (Aguinis and Glavas, 2012). Although hundreds of studies have been undertaken, no definitive explanation has yet been found (Bhattacharya & Sen., 2004; Kollmuss and Agyeman, 2002).

(12)

12 There are several reasons found for this fragmentation in outcomes. Aguinis and Glavas (2012) reviewed 588 articles and 102 books on CSR and suggested several research approaches that may close the knowledge gaps. The first one of them concerns the level of analysis and the authors suggest there is a need for research at the level of individuals instead of organizations. This is backed up by Hillenbrand et al. (2011) who argue that prior research ‘investigated CSR from the perspective of organizations, often focusing on how organizations define, manage and implement CSR to gain benefits or competitive advantage’. Aguinis and Glavas (2012) found that different methodological approaches are needed to develop an understanding of underlying mechanisms and micro foundations (i.e., foundations based on individual action and interactions). Clifton and Handy (2001) emphasize the need to use qualitative methods because: ‘the more we understand about consumer buying behaviour, the more we recognize how much there is that we don't understand’. Another reason for these different outcomes could be that many of these studies do not directly address behaviour, but are only aimed at providing a broad background to the general concept (Davis, 2012). As Davis (2012) tries to explain, a study which focuses on purchasing intention will very likely be misleading, since in the context of green consumption there is a big difference between intention and actual behaviour. In line, Erve (213) finds that the factors influencing the attitude-intention relationship are not necessarily the same as factors influencing the attitude-behaviour relationship. There is evidence to suggest that having a certain attitude does relate to behaviour or may lead to behaviour change, but this relationship is not always predictable (Erve, 2013). When researching the relationship between attitude and behaviour, it is important to note the difference between behaviour and intention.

Öberseder et al. (2011) find shortcomings in the experimental setting of many studies. They claim that some studies are based on artificial awareness of CSR and, while such studies offer some useful insights, the effect of CSR on actual consumer behaviour may not be fully captured. Many of these cognitive methods assume perfect and constant conditions without attention for environmental, situational or social settings. These types of study seem to oversimplify the complex translation of having a positive attitude into actual buying behaviour (Fukukawa, 2003; Carrington et al., 2010). Secondly, at present, consumer research is primarily based on a "snapshot" approach, focussed on only one particular point in time. Erve (2013) and Smith and Brouwer (2012) emphasize the need to advance theory through longitudinal research.

When looking at the milleninials’ attitude towards green products in general, research shows a positive attitude among consumers (Erve, 2013; Paladino & Serena, 2012; Chen & Chai, 2010; Mohr et al., 2001). However, Erve (2013), Smith (2010) and Vermeir & Verbeke (2006) also confirm the

(13)

13 attitude behaviour gap for this generation. The authors found that having a ‘green attitude’ may lead to green purchasing behaviour when looked at in isolation, but in the broader purchasing decision, a positive attitude does not always lead to actual green purchasing. Erve (2013) and Vermeir & Verbeke (2006) conclude that, although millennials express their willingness to make ethical purchases, social responsibility is apparently not the most dominant criteria in their purchase decision. Despite these findings, Erve (2013) indicates that the relationship between attitudes and behaviour cannot be rejected completely and the underlying factors and how they evolve over time needs more clarification.

Before we can move on to identifying what factors might influence the attitude behaviour relationship, it is important to gain a better understanding of this new type of consumer, the millennial.

2.4 Millennials

This thesis is focussed on the generation of young consumers called the millennials. Born between 1977 and 1994, millennials (also known as Generation Y or Echo Boomers) constitute a radically different generation compared to their parents. They are known for certain specific characteristics that may influence their buying behaviour (Nowak et al., 2006). A very important characteristic is that millennials are very technology savvy. This generation has more liberal views on social issues and they are shaped by their access to technology and the internet. Millennials have grown-up with the Internet and use it to buy products or do research (Nowak et al., 2006) According to Moriarty (2004), they also trust the Internet as a valuable source of information.

Next to this, this generation is known to be the most consumption orientated generation ever, caused by the abundance and constant availability of services and products (Sullivan & Heitmeyer, 2008). They are also known for contradicting themselves when it comes to their behaviour and their thoughts about certain topics, making this generation not easily manageable or easy to target by marketeers, due to their many faces (Erve, 2013).

They show concern towards the environment and social responsibility issues and related to this, they possess a strong feeling of responsibility and are looking to express their concerns and ideals through consumption and purchasing (De Pelsmacker et al., 2005; Carrington et al., 2010). Other characteristics that distinguish this generation from others is that they are very optimistic and believe they can make a difference in the world (Nowak et al., 2006). They also find that that life is fun and enjoyable and are always looking for new experiences (Nowak et al., 2006).

(14)

14

2.5 Influencing factors

As literature is inconclusive about what factors trigger actual green purchasing behaviour for millennials, it is difficult to design a valid research design that includes all relevant factors. The next section sets out the most important findings from influential academic literature about what factors explain environmental behaviour and the attitude behaviour gap.

Erve (2013) found that, in the relationship between having a positive attitude towards buying green and actual green purchasing behaviour, attitudes only explain 16% of the variance in behaviour. Results from prior studies that try to explain how different forms of CSR impact consumer behaviour assume a “win-win” outcome, resulting from common support and positive consumer reactions (Du et al., 2010; Green and Peloza, 2011). However, as chapter 2.3 shows, consumers must receive value from the exchange (i.e., the “win” for the consumer) to support companies engaging in CSR. Over time, the term ‘value’ was interpreted in different ways by various researchers. According to Zeithaml et al., (2009) value can be defined as the consumer’s overall assessment of the utility of a product or service based on perceptions of what is received and what is given. Lovelock and Wirtz (2011) find that value is a trade-off between perceived benefits and perceived costs. Jankingthong and Gonejanart (2012) find that the most widely accepted definition of value is ‘the ‘consumers global evaluation of the utility of a product according to their perception of what they receive and what they give’ (provided by Kotler and Keller, 2011).

The oldest theory aimed at understanding value and the value action gap is based on linear thinking (Kollmuss and Agyeman, 2002). This type of model represents the link between knowledge and action and assumes the existence of a consumer that behaves as a consistently moral decision maker. A linear progression is assumed in which environmental knowledge leads to environmental awareness and concern (environmental attitudes). However, having a positive attitude towards CSR and green products is not always a strong predictor of actual purchasing behaviour (Kolmuss and Agyeman, 2002). The purely rational approach used in these models from the 1970’s does not sufficiently explain behaviour (Kollmuss and Agyeman, 2002; Kals et al., 1999).

Both the theory of reasoned action and theory of planned behaviour by Fishbein and Ajzen (1975; 1980) address these methodological issues. The authors support that people are essentially rational in a way. They make use of information available to them but this only influences intentions (rather than direct behaviour). Especially the theory of reasoned action (figure 1) is seen as the most influential attitude behaviour model and one of the most predictive persuasion theories in social psychology. Fishbein and Ajzen find that attitudes do not determine behaviour directly, they rather

(15)

15 influence behavioural intentions which in turn shape our actions. Intentions are not only influenced by attitudes but also by social (‘normative’) pressures. Thus ‘the ultimate determinants of any behaviour are the behavioural beliefs concerning its consequences and normative beliefs concerning the prescriptions of others’ (Ajzen & Fishbein, 1980) The ultimate factors regulating behaviour are beliefs concerning behavioural consequences and normative beliefs concerning the opinion of others (social or peer pressures).

Figure 1. Theory of reasoned action (Ajzen & Fishbein, 1980) There are a few important concepts mentioned in his theory:

• Attitude towards behaviour is defined as 'the result of a consumer's assessment of particular behaviours' (Ajzen, 1991). This directly affects intention.

• Behaviour intention is defined as an indication of an individual's readiness to perform a given behaviour and is based on attitude towards the behaviour, subjective norm, and perceived behavioural control.

• Subjective norm directly influences intention and is defined as an individual's perception about the particular behaviour directly influenced by others. Given the traits of millennials, this influence is expected to have a significant impact.

• Behaviour is defined as an individual's observable response in a given situation with respect to a given target. Behaviour is a function of likeminded intentions and perceptions of control. Perceived behavioural control is expected to moderate the effect of intention on behaviour, such that a favourable intention produces the behaviour only when perceived behavioural control is strong.

Although both frameworks by Fishbein and Ajzen are more sophisticated than the linear theory, the identified factors still do not sufficiently explain pro-environmental behaviour. The link between

(16)

16 knowledge and attitudes, attitudes and intentions, and intentions and actual responsible behaviour is still weak.

Later, Sheth et al. (1991) stated that values may predict consumption behaviour and developed a theory of consumption values consisting of a five-fold typology of values influencing consumer choice behaviour. According to the study, consumer choice is influenced by: functional, social, emotional, epistemic and conditional and they argued that these “relate additively and contribute incrementally to choice” and are therefore independent. However, Sweeney and Soutar (2001) questioned this independency and they found that values are indeed interrelated and value dimensions combined are a better predictor of consumer choice. They discovered three types of values are most apparent: functional, social and emotional value. This was later on confirmed by Green and Peloza (2011) who found that consumer categorization of CSR and their subsequent perception of value corresponds to three of the value drivers identified by Sheth et al. (1991).

Over time, some researchers have noticed a difference between what they call primary motives (the larger motives that let us engage in a whole set of behaviours, e.g. social or functional value, and a second set of moves that influence one specific action, e.g. ‘I would like to buy free range eggs but the shop is very far away’ (Kollmuss & Agyman, 2006; Moisander, 1998). Usually, such internal barriers to pro-environmental behaviour are non-environmental motivations that are more intense and directed differently. In the example, primary motives (environmental values) are overruled by selective motives (personal comfort). It can therefore be hypothesized that ‘primary motives, such as altruistic and social values, are often covered up by the more immediate, selective motives, which evolve around one’s own needs’ (Kollmuss & Agyman, 2002). Habit, environmental awareness and trust are seen as the most influential factors influencing the millennials behaviour. In the next chapters will be explained why.

More recent consumer behaviour theories suggests that consumers possess a range of strategies to deal with both simple and complex decisions (Szmigin et al., 2009). Consumers do typically measure how something relates to their own values, morals, standards and priorities (Green and peloza, 2011). Many studies suggest that the value consumers receive is variable, and that their behaviour is dependent on the value they perceive (Green and Peloza, 2011).

In sum, in the past, researchers produced many classifications and reports attempting to generalize the complex and often subjective different theories that explain the relationship between attitude and (green) purchase behaviour. Even though there is at present not a single conclusive model that integrates all factors behind pro-environmental behaviour, there are some key factors. The table

(17)

17 below (table 2) provides an overview of the factors that are most identified in existing literature. These will be further explained in the next chapters and form the basis of the interview guide that was used during the data collection.

Factor Definition Source

Emotional value

Value that is received when a consumer makes a purchase with a social or environmental attribute (mostly related to traditional forms such as donations

Green and Peloza (2011) Sweeney and Soutar (2001) Sheth et al. (1991)

Social value Comparable to the subjective norm, this type of value deals with others’ perception about the particular behaviour and is used to define oneself

Paladino & Serena (2009 ) Green and Peloza (2011) Sweeney and Soutar (2001) Sheth et al. (1991) Ajzen & Fishbein, 1980 Functional

value

The actual benefit the consumer receives from the product or service (egoistic). E.g. consumers purchase organic foods because they are healthier

Erve (2013) Grimmer & Woolley (2012) Green and Peloza (2011) Sweeney and Soutar (2001) Sheth et al. (1991)

Habit A fixed way to think or feel and is determined by prior repetition of an experience or the so called frequently performed past behaviour

Erve (2013) Ajzen , Czasch & Flood, (2009) Kollmuss & Agyman (2002) Aarts et al., (1998) Ramayah et al. (2010)

Awareness The level of knowledge the consumer holds relevant to environmental issues. This kind of knowledge can be gained through experience or study

Schmeltz (2012) Kolkailah et al. (2012) Öberseder et al. (2011) Laivate (2011) Bhattacharya & Sen (2009) Kollmuss and Agyeman (2002)

Trust Consumers ‘perception (trust or scepticism) towards the motivation of the organization to engage in green products perceived concern for its consumers

Erve (2013) Schmeltz (2012) Bhattacharya and Sen (2009) Drumwright, (1996) Fein, Hilton & Miller (1990)

Table 2. Key factors explaining the value behaviour relationship

In the following chapter the factors that have been established as having the most influence will be discussed in more detail and related to what is known about millennials.

2.5.1 Emotional value

Green and Peloza (2011) defined emotional value as value that is perceived when a consumer makes a purchase with a social or environmental characteristic. This is often related to more “traditional” forms such as donations to charity. The perceived value goes beyond self-interest. De Young (1996) found that having such intrinsic values are believed to lead to engagement in a pro-environmental behaviour. According to Kanovich (2007), instead of looking for functional product benefits, the contemporary ethical consumer is asking for demonstrations of good and ethical behaviour. It is important to keep in mind that there also is a “breaking point”; CSR is generally seen as a bonus and not the key factor to favour certain products. When consumers experience more than one form of value, functional value was considered to be more important when making purchase decisions (Green and Peloza, 2011). Consumers will buy a green product when, besides the emotional value stemming from the environmental benefit, they also see direct individual benefit in purchasing the product (Nottage, 2008; Stern, 2000). But, even though consumers give lower priority to products

(18)

18 that generate emotional value, they do not give companies carte blanche when it comes to abandoning CSR (Green and Peloza, 2011). In cases of negative emotional value (such as a bad company reputation), consumers are generally more impacted when compared to positive CSR (Creyer and Ross, 1997). This shows that there are certain baseline expectations towards companies.

2.5.2 Social value

Baker and Ozaki (2008) suggest that the moment when a purchase is made provides to be a crucial point of influence by companies, as consumers would be open and reactive due to others. Indeed, Han et al. (2010) confirm this and found that within (green) consumerism, individuals can be ‘highly influenced by their social surrounding, such as relatives, friends, colleagues and business partners’. When purchasing goods from companies active in CSR, consumers perceive this may lead to being seen as a ‘better person’ in eyes of others (Green and Peloza, 2011). This type of value is also called social value and in this case, buying green products is done to define oneself to their community (Green and Peloza, 2011). Green and Peloza (2011) furthermore explain that, ‘by being seen as someone who cares for the environment, a consumer can ensure they meet social norms and community standards that reinforce such behaviour’. Social value can be obtained through the consumers association with (positively or negatively) stereotyped demographic, socioeconomic and cultural-ethnic groups (Kosonen, 2014). Therefore the consumption of highly visible products, such as clothing, is often driven by social values (Kosonen, 2014).

The concept of social influence has also been included in the theory of reasoned action and theory of planned behaviour by Ajzen & Fishbein (1980). Their theory explains the factor of subjective norms which are associated with social pressures to perform or not to perform a specific behaviour (Han et al., 2010). While most models are conceptualized within individual cognitive space, their theory finds that consumer behaviour is based on perceptions of whether they are expected (by their friends, family and the society) to carry out the recommended behaviour.

Many of the same forms that provide emotional value for consumers also provide social value (Green and Peloza, 2011). This is because buying green products or in other word ‘doing good’ can be perceived as ways to show who you are by what you support. However, the same as with emotional value, findings by Green and Peloza (2011) indicate social value is given a lower priority from consumers than functional value. Conversely, Vermeir and Verbeke (2006) studied young consumers’ green purchasing behaviour and discovered that this generation is very vulnerable to social pressure from peers and this is expected to have a significant influence on their (green) purchasing behaviour. Bush, Martin and Clark (2001) built and extended consumer behaviour theories on the social learning theory. This theory demonstrated that direct role models such as fathers, mothers and teachers had the greatest impact on the attitudes of young adults (Bush et al., 2001). In turn, celebrity role models

(19)

19 may also bring social pressure to buy certain products and are also a good source of product information (Bush et al., 2001). Hume (2010) found that millennials are mainly loyal to their closest circle (Hume, 2010). It is clear that many researchers support the hypothesis that social value (will) play a big role in millennials’ decision making.

2.5.3 Functional value

Self-interest, functional value, economic value or personal benefits all refer to the focus a person has to fulfil his own needs and desires. Sheth et al. (1991) define functional value as ‘the capacity of an alternative to provide utilitarian or physical performance, which are the result of salient functional, utilitarian or physical attributes’. Prior literature on consumer value overall has one thing in common, that is that consumers are largely motivated by economic rationality and therefore generally act in ways that maximize outcomes for the self. Furthermore, consumers appear to be unwilling to trade off economic value for other values (Bazerman, 1998; Cropanzano, Rupp, Mohler, & Schminke, 2001; Peloza and Shang, 2010; Hartmann, 2011). Green and Peloza (2011) confirmed this, and moreover found that functional value is the leading (and, in many cases, the sole) driver behind adding environmental responsibility in their decision-making process. This has also been focus of research for Arora and Henderson (2007) who explicitly created tensions between concern for ‘self’ and ’other’ in three experimental studies. Respondents were asked to trade off price discounts (i.e. “self” component) against donations of equal monetary value for varying social causes (i.e. “other” component). Their findings suggest that promotions with a ‘self’ component seem to be more effective than promotions with a social cause component. This is backed up by Davis (2012), who suggests consumers evaluate whether buying a certain product corresponds with their own personal concerns.

In sustainable purchasing, functional value is expected to significantly impact millennials. This generation is very narcissistic, self-centred and has been criticized due to its hedonistic behaviour (Fajersson and Cerrudo Sampol, 2013). When young consumers engage in (green) purchasing, they are often driven by egoistic motives related pleasure, enjoying life and self-indulgence (Fajersson and Cerrudo Sampol, 2013). Taking this into account, millennials might have a different perception of functional value than their parents, as something that is closer related to pleasure or excitement rather than simply buying products.

(20)

20

2.5.4 Habit

A habit can be seen as a behavioural routine, repeated on a regular basis and often developed subconsciously (Erve, 2013). Habits can also be seen as thinking in a fixed way, building on the prior repetition of behaviour (Ajzen, Czasch & Flood, 2009). When looking at behaviour, this is often determined by frequent past actions rather than by perceptions of new situations. When performing a certain behaviour based on habit, this person will be more likely to use simple decision rules (Sutton, 2006). Aarts et al. (1998) found that people who show certain habits towards green purchasing, searched for less information and only focussed on the habitual choice instead of alternatives. This also shows the frequency of the behaviour determines strength of the habit which directly effects future performance (Oulette and Wood, 1998). Kollmuss and Agyeman (2002) found that habit mainly has a negative influence on green purchasing behaviour. They claim consumers are not persistent enough in practicing new behaviour until it becomes a habit. And when a consumer subsequently never buys green products, it is very hard to change this habit into purchasing green products (Aarts et al., 1998). The majority of consumer buys products they have always bought, instead of making extra effort to change behaviour and be socially responsible (Ramayah, Lee & Mohammed, 2010).

It is expected that habit will also impact the green purchasing behaviour of millennials. Opposed to their parents, this generation shows an accepting attitude towards decisions their parents make and are influenced by their direct role models, as found in chapter 2.6.2 about social values (Rainer & Rainer, 2011). Millennials are therefore expected to be influenced by- or make the same choices as their parents out of habit and therefore the attitude behaviour gap will continue to exist for this generation.

2.5.5 Awareness

Numerous surveys and studies have confirmed an increasing awareness for CSR and environmental or ecological issues (Öberseder et al. 2011). Consumer awareness can be defined as ‘the amount of time that has been spent on processing information about green products’ (Baker, Hutchinson, Moore & Nedungai, 1986). It is confirmed by many researchers over time that awareness holds a key role in green purchasing and some found it is even a perquisite for making a green purchase (Kolkailah et al., 2012; Schmeltz, 2012; Öberseder et al., 2011; Bhattacharya & Sen, 2004). When consumers have access to information, it is said to affect company and product evaluations and eventually their purchase intentions (Kosonen, 2014). In order make a green purchase, it is important that a consumer is aware and notices how the company communicates about the green product (Schmeltz, 2012). In return, when a consumer is not aware this will not influence their attitudes and

(21)

21 intentions towards purchasing green products because they will not be able to form an opinion (Schmeltz, 2012; Paladino & Serena, 2012). Similarly, Preuss distinguishes between an ‘abstract willingness to act’, based on values and knowledge and a ‘concrete willingness to act’ based on habits (Preuss, 1991). And as found in the previous chapter, consumers who show certain habits towards green purchasing, search for less information and only focussed on the habitual choice. When looking at the younger generation, studies find that this generation is marked by being very technologically savvy, with the internet a primary source of information which in turn leads to a high awareness of the environmental- and societal issues (Nowak et al., 2011). It is therefore assumed that this generation will also act on this, given that a higher awareness leads to green purchasing behaviour.

2.5.6 Trust

Trust or scepticism towards the motivation of the organization to engage in green products, is frequently mentioned in literature as a negative influencing factor in the attitude-behaviour relation (Erve, 2013). Blake (1999) points out that a lack of trust often stops consumers from acting pro-environmentally. Heidinger (2015) emphasizes the importance of being trusted as a company, as trust leads to increased loyalty which in turn fosters (re)purchase intention. In turn, a lack of trust makes consumers suspicious which may lead to less willingness to buy. Accordingly, CSR and green branding have become strategic tools that are assumed to generate trusting relationships between brands and consumers, if carried out correctly (Popescu, 2012). Brands with a strong CSR reputation are continuously making efforts to suitably respond to their target groups while strengthening brand trust and reputation (Popescu, 2012).

According to Morgan and Hunt (1994), relevant mediators or antecedents of brand trust are shared values and communication, which stresses the importance of the trust construct for this research. Shared values also form a basis for identification and trust. Another factor that may influences consumers’ trust and in turn, increased consumption, is the conformism pressure they are likely to feel. Whether their peers trust a brand or not, millennials are expected to make choices similar to those made by others (Gupta and Ogden, 2009).

Millennials are seen as a generation that is more sceptical towards corporate messages than previous generations and also more open to other experiences (Schmeltz, 2012). This, and having easy access to sources of information to challenge messages and look for alternative purchases makes it very hard for a company to base their business model on repeat purchases; for every purchase you have to compete for the loyalty of this generation (Nowak, 2006).

(22)

22

2.6 Summary

Consumers, and more specifically millennials, hold positive attitudes towards environmental issues, green products in general and purchasing green products. One of the most significant ways they can make a positive contribution to prevent or decrease the environmental damage, is by purchasing green products. However, having a positive attitude towards buying green products does not always seem to have a direct impact on actual green purchasing: the attitude behaviour gap also exists for this generation. Emotional value, social value, functional value, habit, awareness and trust are expected to significantly influence the attitude-behaviour relation. Each in their own way, as discussed in this chapter. How they impact this relationship, and how this has changed from 2012 to 2015 is described in the results section and conclusion of this thesis. The next chapter presents the set-up of the research method and data collection.

(23)

23

3. Method and data

This chapter will outline and discuss the chosen research method that is used to answer the research question. The interview structure and methods for data collection and analysis will be explained and the last part describes the reliability and validity of the research.

3.1 Research design

The way the topics of sustainable behaviour and green purchasing evolved over the last decades signifies that a change in research focus is needed (Erve, 2013; Sciarelli and Tani, 2012; Aguinis and Glavas, 2012; Kolkailah et al., 2012; Öberseder et al. 2011; Hillenbrand et al. 2011; Carroll and Shabana, 2010; Bhattacharya & Sen., 2004; Kollmuss and Agyeman, 2002). More recent research by Farrel and Twining (2005) shows that sustainable behaviour can be seen as an ever evolving concept or a ‘learning process’ and when aiming to understand this complexity, quantitative approaches only provide a limited explanation. This is in line with research by Clifton and Handy (2001) who demonstrate that quantitative survey methods are not suited to explore areas of research where behavioural issues remain unidentified. Farrel (2006) suggest that qualitative longitudinal research (QLR) is very suitable to address these issues (Farrel, 2006). QLR is a relatively recent development in research, and is based on qualitative interview methods which are performed at more than one point in time. This way, a dynamic perspective and deeper understanding of millennials’ buying behaviour and the influencing factors can be gained to answer the research question, and the gaps left by quantitative methods can be addressed. Rather than focussing at point point in time, this method allows to make a comparative analysis between two or more points in time. In this thesis, the same respondent group of millennials was interviewed first in 2012 and re-interviewed in 2015. This allowed the respondents to reflect on the changes they experienced since the previous interview which provides a dynamic rather than static perspective. Also, by giving the respondents more time to reflect on their answers, interesting outcomes are found which would have otherwise remained concealed. Qualitative methods can also be used to improve both the design and interpretation of traditional surveys because they bring the stories to life (Deakin, 2001). Given the changing need in research focus, it is evident that a qualitative longitudinal research design is appropriate to answer the research question.

There is currently no definition of how long studies should last, nor is there any guidance in the literature as to how long the time intervals between interviews ought to be (Farrel, 2006). In essence, QLR studies often involve the following sorts of research design, although as Holland et al (2004) note, it is hard to draw precise boundaries around differing styles of QLR studies:

(24)

24 The strengths and weaknesses of this type of research can be classified as follows (Farrel, 2006). Weaknesses:

• This type of research is very time consuming. For example for this research, interviews lasted about one and a half hour, and it took about five hours to write out and interpret the interviews.

• Secondly, as change takes time to emerge, some studies will need to be undertaken for years (Farrel, 2006; Holland et al, 2004). Therefore, the results of any such study may only arise several years after the study was done.

• A commitment to a QLR study may be as much (or more) of a commitment on the part of the researcher as it is the respondents in terms of time, energies and opportunities to undertake follow-ups several years later.

• Like any follow-up or research design in which respondents are returned to, careful thought needs to be given to the design of the research instruments and questions. In this case, the same topic list was used in order to assess the development of green consumption behaviour

Strengths:

• Qualitative longitudinal research allows to gain a prospective rather than retrospective study. Retrospective studies can be influenced by respondent’s failure to recall events or the correct ordering of events and Farrel (2006) found that respondents often also ‘leave themselves open to deliberate distortions as they attempt to imbue their actions with a rationality which they did not have at the time, or non-deliberate distortions due to subconscious suppressions of painful memories’. Prospective studies are less likely to be influenced by biases, since one keeps a record of what was said earlier. This can also bring its own biases but one can triangulate data with either other records or other respondents and it should be easier to detect.

• QLR allows for respondents to reflection on the changes (or lack of them) which they have experienced since the previous interview. You can gather specific data related to the respondent being interviewed and therefore ask about specific events, periods or feelings in order to measure changes.

• In theory, the researcher is able to tailor-make follow-up interviews for each respondent and to plan to ask specific questions of them based on their previous answers and experiences. Of course, this represents a further time-cost.

• Qualitative longitudinal research allows linking the micro perspective to the macro (Farrel, 2006). For example, a study of the reorganisation of a large firm may take several months or more than a year or so to be completed. During this time employees’ feelings about the

(25)

25 processes of change will follow any number of trajectories, depending upon how the changes are progressing and what these mean for them as individuals. A QLR study of these processes of change would need to understand the employees’ feelings in relation to the macro level changes experienced by the organisation where the research was based.

3.2 Interview structure

The interviews are semi structured, which means that not all questions were designed and phrased in advance, but a topic list and open ended questions were used (Habraken et al., 2008). Semi structured interviewing was used because it shows a high validity when asking respondents about complex questions and subjects. Due to the flexible research design, issues can be clarified (such as emotions or values) and respondents may talk freely and in detail and depth (Mathers et al, 1998). The topic list was gathered based on literature in the previous chapter and evolved over time so that

new factors and associations could be discovered. The general objective of semi structured interviewing is to understand the respondent rather than making generalizations about certain behaviour (Barriball, 1994). During the interviews, the respondents were continuously asked to explain their answers and the questions using the laddering technique. The laddering technique means that the interviewer starts with a question, and then another question is asked about their response, for example: Interviewer: "Why x?” Respondent: "Because z" Interviewer: "Why z?” etc. This technique is used to discover the underlying, subconscious reasons for the particular answers (Reynolds and Olson, 2006).

On average, the discussions took one and a half hour. Prior to the interviews, respondents were informed about the confidentiality and discretion of the interviews. The respondents were interviewed in an informal setting (café/bar). This interview setting was chosen in order to provide a more intimate situation for discussion. All interviews began with collecting background information, such as age, education, occupation and household size. In the first round of interviews this was followed by an introduction to the topic and the second time with a refresher. In the next section, general attitudes how respondents generally consume and utilize products were measured. During this part, the attitude and motives towards CSR and green products in general was measured and general patterns for actual green consumption were identified. This way, the existence of the attitude behaviour gap for millennials could also be identified. The following part was used to distinguish specific values and factors and reasons for preferring one type of product over another. To identify how their behaviour changes overtime, Saldaña (2003) developed a set of questions to guide the analysis of longitudinal qualitative data and initially frame data analysis. A subset of these questions were used in this thesis in the second round of interviews and slightly adapted to the topic, such as “How has this changed between t1 and t2?”. Saldaña explains that asking this question

(26)

26 provides a simple but useful way of analysing what is different between two time periods. This way, over time the interviewer can build up a notion of the general direction of an individual’s progression. Another question that was asked is “Did something occur over period x that caused this change?”. With this question the occurrence of sudden events can be explored, which trigger a change or which in some way initiate a process of change. Alternatively, something might have accelerated or developed which brings about a change (Farrel, 2006).

3.3 Respondents

In the current study, age was the common aspect for the respondents. Table 3 shows an overview of the respondents’ profile. Even though a wide range of demographical factors exist, like age, gender and income, the majority of research in the field of psychology underlines that age, gender and education are most influential (Wittgenstein Centre, 2011). In this thesis, personal interviews were conducted with ten higher educated, young (age 18-30 at the time of the first interview) Dutch respondents. Consumers with a green attitude were interviewed and education, age and gender profiles are included as characteristics. Respondents with the same background were chosen, since it is assumed they share the same values and beliefs. This is important in order to compare the outcomes. To overcome inappropriate generalization, qualitative data was collected in the Netherlands. Other respondent groups have not been considered in this thesis, because it have would have made this research far too complicated.

Table 3 shows the characteristics of the respondents in 2012. Both men and women that finished a high education degree (HBO, University) were interviewed. In selecting people to interview, I began with my existing network and then used the ‘snowball sampling’ technique to find participants particular backgrounds (gender, age, education). This sampling technique uses recommendations to find people within the specific range of characteristics that has been determined as being useful (Goodman, 1961). When using this technique the following steps have to be taken:

a. Identify one or more respondents in the desired population

b. Ask the identified respondent to find further respondents and so on until the sample size is met

This way, new relationships are built through mutual association. Relevant connections were made via respondents referring to other respondents. In this thesis, an important characteristic is that respondents are higher educated. These people are, according to literature, most aware of how their behaviour influences the environment and perform pro-environmental behaviour at home. Next to the criteria gender, age and education, selection of the respondents is random.

(27)

27 In 2015, at the time the second round of interviews were held, unfortunately one of the respondents could not be reached (respondent G). It appeared this respondent moved abroad and she was therefore not included in the follow up. Table 4 provides an overview of respondents that were interviewed in the second round of interviews. These are the same respondents and the table shows how their personal situation has changed over time.

Male respondents Age Education Marital status Occupation Family size

Respondent A 26 WO Single Employed 1

Respondent B 30 HBO Married Employed 3

Respondent C 20 HBO Cohabiting Student/Part-time

Employed 2

Respondent D 18 WO Single Student/ Self-employed 1

Respondent E 21 HBO Single Student/ Employed 1

Female respondents Age Education Marital status Occupation Family size

Respondent F 26 HBO Married Employed 2

Respondent G 30 WO Cohabiting Self-employed 3

Respondent H 25 WO Single Employed 1

Respondent I 19 WO Single Student/ Employed 1

Respondent J 27 WO Cohabiting Employed 2

Table 3. Interview respondents (2012)

Male respondents Age Education Marital status Occupation Family size

Respondent A 29 WO Single Employed 1

Respondent B 33 HBO Married Employed 3

Respondent C 23 HBO Cohabiting Student/Part-time

Employed 2

Respondent D 21 WO Single Student/Part-time

Employed 1

Respondent E 23 HBO Relationship Employed 1

Female respondents Age Education Marital status Occupation Family size

Respondent F 29 HBO Married Employed 3

Respondent G 30 WO Cohabiting Self-employed 3

Respondent H 27 WO Cohabiting Employed 2

Respondent I 22 WO Relationship Student/Part-time

Employed 1

Respondent J 29 WO Cohabiting Employed 2

(28)

28

3.4 Data analysis

The analysis in this research is done based on approaches defined by Pope et al (2000) and Saunders et al. (2009). Qualitative Longitudinal Research produces vast amounts of raw data of the research. In this phase, they provide a descriptive record of the research and cannot yet provide explanations. By sifting, classifying and interpreting, the interviewer has to make sense of the data. Babbie (2004) explains inductive reasoning, an interview technique which is used in this thesis. This approach is useful in this thesis since new variables and relationships between variables occur, changing the findings from general to specific (Eisenhardt 1989). Inductive reasoning means that the interviewer uses content analysis to generate categories and explanations from specific observations. This way, patterns and consistencies can be detected and conclusions can be developed.

The analyses were done using software program MAXQDA and consisted of multiple stages:

(1) Sorting of text fragments and the identification of recurring categories and themes. This was done deductively using the primary topic list. For example, a category that was identified is ‘attitude behaviour gap’.

(2) After this, the analyses continued in a more inductive way by providing each part with different labels and sub-labels, with the purpose of finding related connections. Labels were given to specific parts of the text. For example ‘frequency of buying green products’ was given a label. A sub-label indicates the frequency of this behaviour, in this case frequency was labelled with ‘always, very often, sometimes, almost never, never’.

(3) Finally, the data was analysed with the qualitative software program MAXQDA.

Software packages such as MAXQDA are designed for qualitative data analysis and enable complex organisation and retrieval of data (Habraken et al., 2008). A growing number of researchers use computer programs to assist in the analysis of qualitative data. MAXQDA provides more sophisticated support in the analysis stage to identify and create codes. However, qualitative studies are not designed to be statistically generalized, which can be done with questionnaires.

In the final phase of this thesis conclusions are drawn in line with the main question. The outcomes are compared to existing literature. In the last part, recommendations for future research and limitations are described, referring to the implications of the research (Saunders et al., 2009).

3.5 Reliability and validity

Semi structured interviews were conducted in this research. This technique allows collecting qualitative data by setting up a personal interview. Individual interviewing brings several benefits such as the absence of normative pressures or confidentiality issues (Barriball, 1994). In a semi

(29)

29 structured interview, not all questions are designed and phrased in advance but the interviewer makes use of a topic list and open ended questions. The topic list used in this thesis was gathered based on topics gathered during the literature review (see table 2). Furthermore, since questions are not ‘pre-set’, the interviewer is unbiased about what he feels is relevant information. This also holds that the questions used may vary from interview to interview (Saunders et al., 2009). The general objective of semi structured interviewing is to understand the respondent rather than make generalizations about certain behaviour (Barriball, 1994).

When using qualitative techniques, questions about generalizability and validity can be raised. In this thesis, a sample of ten respondents seems relatively small. However the main objective of this thesis is not to produce a generalizable theory but to address the gaps that have been left by quantitative research. In relation to sample size, (external) validity and meaningfulness derived from qualitative research methods are considered to have more to do with the information collected from the selected cases, than with sample size. This is because there are no rules for sample size in qualitative research. This depends on the intention and objective of the research of the research (Jamali et al., 2009).

The informal setting that was created also contributed to a better data collection and obtaining more detailed information. This setting caused respondents to talk freely and in detail and depth. During the interviews, it was important that the meaning and intent of the respondents is understood so that reliable interpretation can be ensured. Due to the flexible research design, complex questions and subjects could be discussed and clarified (such as emotions or values), which proved to be a great advantage during the interviewing. Respondents experienced the interviewing more as a conversation rather than data collection, which provided detailed reports. It was found that the respondents reacted positively to the topics presented to them and they tended to be generous with their answers and time. The respondents could all relate to the topics presented to them.

To ensure validity, some researchers see triangulation as a method for validating findings (Pope, 2000). Qualitative researchers generally use this technique to ensure that findings are rich, robust, inclusive and well-developed. Richards (2006) explains Triangulation as ‘gaining multiple perspectives through studies that have been conducted on the same topic and that directly address each other’s findings’. This way, outcomes can be challenged (for clarification), illuminate it (link outcomes to theory) and verified (provide the same conclusions). Richards (2006) indicates this is a suitable design when the research consists of two studies that reflect on the same phenomenon, such as being done in this research. In this thesis, the output of the two studies that were conducted in 2012 and 2015

Buying a better future : a longitudinal qualitative study, exploring the relationship between attitude and green buying behaviour for millennials