1
To what extent are biospheric values expressed
automatically when it comes to pro-environmental
behaviour?
Date:
22th of June, 2018
Student:
Pieter Zuiderveld - 10292691
Education:
Msc. Business Administration
Track:
Digital Business
University:
Amsterdam Business School - University of Amsterdam
2
Abstract
In environmental literature, attempts are being made to turn consumer attention more towards pro-environmental behaviour. One of the drivers of pro-pro-environmental behaviour are biospheric values, but unfortunately values do not always result in the right behaviour. Therefore, it is important to look at how people express their values. Values can be expressed through automatic judgements or more through reasoning. In the interest of better understanding this relationship and increasing pro-environmental behaviour, this study examines how biospheric values are expressed. The direct effect of biospheric values on pro-environmental behaviour and the moderating effect of the type of expression on the relationship between biospheric values and pro-environmental behaviour are analysed. Biospheric values is expected to have a positive relationship with pro-environmental behaviour and the type of expression is expected to moderate the relationship between biospheric values and pro-environmental behaviour in such a way that with automatic judgements, people will show more pro-environmental behaviour. With reasoned expressions, people will have more time to reason which is expected to decrease pro-environmental behaviour. In order to collect data, a
questionnaire was used to investigate biospheric values and pro-environmental behaviour and half of the participants were cognitively distracted, which leads to automatic judgements. The results show that biospheric values have a significant positive direct effect on pro-environmental behaviour and that there is a significant moderation effect of automatic judgements on the relationship between biospheric values and pro-environmental behaviour. However, the moderation effect is only
significant for participants with high biospheric values. The results demonstrate that if the biospheric values of people can be increased and if they make more automatic judgements the pressures on the environment can be reduced.
Key words: pro-environmental behaviour, biospheric values, cognitive distraction,
3
Statement of originality
This document is written by Pieter Zuiderveld 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.
4
Table of content
List of figures ... 5 List of tables ... 5 1. Introduction ... 6 2. Literature review ... 9 2.1 Pro-environmental Behaviour ... 9 2.2 Biospheric Values ... 11 2.3 Cognitive Distraction ... 14 2.4 Conceptual model ... 16 3. Methods ...16 3.1 Procedure ... 16 3.2 Sample ... 17 3.3 Measurements ... 18 3.3.1 Pro-environmental behaviour ... 18 3.3.2 Cognitive distraction ... 19 3.3.3 Biospheric Values ... 19 3.3.4 Control variables ... 20 3.4 Analyses ... 20 4. Results ...224.1 Correlations, means, SE’s ... 22
4.2 Regression analysis biospheric values ... 22
4.3 Moderation analysis without control variables ... 23
4.4 Moderation analysis with control variables ... 26
5. Discussion ...30
5.1 General discussion ... 30
5.2 Theoretical implications ... 31
5.3 Limitations and future research ... 33
5.4 Practical implications ... 34
6. Conclusion ...37
5
List of figures
Figure 1. Conceptual model
Figure 2. Graph of simple slopes without control variables
Figure 3. Graph of simple slopes with control variables
Figure 4. Conceptual model with results
List of tables
Table 1. Descriptive statistics and correlations
Table 2. Results of the hierarchical regression analysis
Table 3. Results of moderation analysis without control variables
Table 4. Results of the conditional effects without control variables
Table 5. Results of moderation analysis with control variables
Table 6. Results of the conditional effects with control variables
6
1. Introduction
Life on earth is currently facing environmental problems such as global warming, air and water pollution, declining amounts of drinking water, declining forests and desertification. Because human behaviour has a large impact on these issues, the pressures on the environment can be reduced if we would change the way we behave and act (Steg, Bolderdijk, Keizer & Perlaviciute, 2014; IPCC, 2007). A growing number of people, organizations and countries now believe that we cannot continue our current rate of consumption and combustion, because the planet can probably not sustain it in the long term, especially with the increasing global population (McDonagh & Prothero, 2014). The consumption behaviour of individuals and households have a large impact on the environment. In most countries, household consumption over the lifecycle of products accounts for more than 60% of all environmental impacts of consumption (Griskevicius, Cantú & Vugt, 2012). The importance of consumers purchasing eco-friendly products (EFP) to reduce the pressures on the environmental is therefore discussed in recent literature and addressed by policy makers, NGO’s and companies. Ethical product labels such as Fairtrade which are sometimes better for the environment are nothing new to retail shops in developed countries. These products try to do things different than their non-ethical competitors. This mostly has to do with better labour conditions, reduced pressures on the environment, how animals are being treated or consumers can help with a particular cause if they purchase the ethical product (Hainmueller, Hiscox & Sequeira, 2015). Still, despite their efforts, the current market shares of EFP’s are still fairly low. Recent studies show that even though people want to be green and lower their carbon footprint, only 9% use any EFP’s (Griskevicius et al., 2012).
One of the reasons that still few people buy environmental and ethical products could be the price of a product because it is an important aspect in the decision-making process of consumers, especially for consumers with limited resources. Some consumers think of green products as merely being more expensive than non-green products and are therefore not willing to pay a premium price for it. Other consumers do buy green products, but just because they are driven by status, especially
7 if they buy daily consumer goods in a public setting (Griskevicius, Tybur & Van den Bergh, 2010). Gatersleben et al. (2010), suggests that people are motivated by environmental concerns to behave in an environmental friendly manner but they also have material concerns which motivates them to buy new products, thereby increasing their environmental impact. It is therefore important to promote environmental awareness and to address material concerns.
Policy makers, NGO’s, companies and consumers around the globe are making attempts to increase pro-environmental behaviour. Investors have also noticed the relevance and importance of sustainability for the profits of companies (McDonagh & Prothero, 2014). Expressing this behaviour comes from a concern for other people, future generations, ecosystems and the planet as a whole. It is also in a person’s self-interest because it minimizes health risks. Because making consumers behave more pro-environmental is such an urgent case for society, understanding the behaviour of consumers is of utmost importance (Morren & Grinstein, 2016). One urgent question we should ask ourselves is; What drives pro-environmental behaviour and how can people be encouraged to behave
more pro-environmental?
In order to answer this question, we have to look at the drivers of pro-environmental
behaviour and what it takes to enhance these drivers. Recent studies have shown that environmental attitudes are related to values and that especially biospheric values are related to pro-environmental behaviour (Schultz & Zelezny, 1999; Steg et al., 2014). Stern et al. (1999) developed the Value-belief-norm theory which suggests that values are the first link in changing worldviews and creating awareness for the negative results of bad behaviours. However, values do not always result in the right behaviour, therefore it is necessary to investigate how people express values because it can play an important role on the relationship between values and behaviour.
According to Cornelissen, Dewitte & Warlop (2011), people tend to struggle to go after long-term goals such as pro-environmental behaviour if there are attractive alternatives. If this behaviour is expressed automatically it could be obvious for people to go after their long-term goals because the alternatives will no longer be considered. However, if long-term goals are expressed through a
8 more reasoned process, it is probably more difficult to refuse the alternatives. According to Dewitte, Bruyneel & Geyskens, (2009) adopted values are first in our reasoning system, but after training the expression goes to the automatic system. This suggests that if people would express
pro-environmental behaviour automatically, the pressures on the environment could be reduced more easily. Since little research has been done on this topic, this study will analyse how the underlying values of pro-environmental behaviour are expressed. For academics, managers and the
environment this is necessary in order to make recommendations about how to increase pro-environmental behaviour in the future.
In order to fulfil the objectives written above, the following main research question was formulated:
- To what extent are biospheric values expressed automatically when it comes to pro-environmental behaviour?
This study will contribute to the literature because of two reasons. First of all, the results will reveal if biospheric values are indeed a driver of pro-environmental behaviour. Secondly, the interaction between the drivers of pro-environmental behaviour and the way people express behaviour is a topic which has not received a lot of attention. The outcome will give a better understanding of how to turn consumer attention to more environmentally friendly products and behaviours.
This study has five sections; the theoretical framework, methods, results, discussion and conclusion. In the theoretical framework, the literature on pro-environmental behaviour, biospheric values and the expression of values is discussed. In addition, the link between these three concepts and the moderating effect of the expression of values is investigated. The methodology and sample are described in the method section and the results are presented in the result section. The
discussion section reviews the results, its implications, the limitations of the study and recommendations for future research. At the end, the conclusions of this study are given.
9
2. Literature review
2.1 Pro-environmental behaviour
Pro-environmental behaviour has received growing attention from scholars in the field of consumer behaviour and can be defined as any action that enhances the quality of the environment (Steg et al., 2014; Griskevicius et al., 2012; Nguyen, Lobo & Greenland, 2016). Azjen (1991),
developed the Theory of Planned Behaviour (TPB) which suggests that people make reasoned choices and base a selection on the highest benefits against the lowest costs either in terms of money, effort or social approval. The behaviour follows a certain intention which displays how much effort a person wants to take to show a certain behaviour. Intentions depend on three factors, namely the attitudes towards the behaviour which reflects the degree to which engagement in behaviour is positively valued, social norms which are social pressures to activate a particular behaviour and perceived behavioural control which are thoughts about whether the person is able to express the behaviour. Factors such as demographics or values are thought to have an indirect effect on
intentions through one of the three factors mentioned above (Steg & Abrahamse, 2010). The TPB has been used in behavioural areas which relate to environmental issues and has shown to be successful in describing different types of environmental behaviour, such as household recycling (Kaiser & Gutscher, 2003), waste composting (Mannetti, Pierro, & Livi, 2004), buying energy-saving light bulbs, water use, meat consumption and pro-environmental behaviour in general (Kaiser, Wölfing, & Fuhrer, (1999). According to the TPB, pro-environmental behaviour will more likely occur when people have a positive attitude towards pro-environmental behaviour, believe others are already doing it, just think it should be done and when people believe that they can do it (Gatersleben, Murtagh & Abrahamse, 2014).
10 People are generally motivated to behave pro-environmental for hedonic reasons (because it is enjoyable), gain reasons (because it saves money) or normative reasons (because they think protecting the environment is the right thing to do) (Steg et al., 2014). If pro-environmental behaviour is profitable, pleasurable and the reasons are clear, people are likely to express it, but if that is no longer the case, people will most likely stop expressing pro-environmental behaviour. Still, people do not always express pro-environmental behaviour because it might not be activated by the context, therefore strategies that want to boost pro-environmental behaviour must make sure they are also activated in the context. With many pro-environmental actions, there is a conflict between hedonic reasons on the one hand and gain and normative reasons on the other hand. Behaving pro-environmental is generally seen as the right thing to do, but it is usually more expensive. According to Steg et al. (2014), in order to reduce this conflict, one could try to strengthen gain and hedonic goals, which can encourage people to behave in a more pro-environmental way. This can be done by reducing prices for pro-environmental products, by increasing prices for environmentally harmful products, by making pro-environmental products fun and by making pro-environmental products more convenient. Even though this strategy seems positive, focusing on hedonic and gain reasons alone can have negative consequences, because people might only behave pro-environmental when it is financially interesting. Hedonic and gain goals could also support instead of conflict with
normative goals. If this is the case, behaving pro-environmentally is not only the right thing to do, but also makes people feel good and enhances their status. Therefore, it is proposed that strengthening normative goals is at least as important as strengthening gain and hedonic goal. The strength of normative goals depends on certain values which some people find important. According to Geller (2002) and Steg & Vlek (2009), promoting behaviour change is more effective when following four steps; (1) selecting the behaviours that need to be changed, (2) investigating what causes those behaviours, (3) applying interventions to change those behaviours and their antecedents and (4) evaluating the effects of the interventions.
11
2.2 Biospheric values
Schwartz (1992) defined values as desirable and trans situational goals that serve as guiding principles in one’s life. Values are generally formed during childhood and is a product of the needs, traits, experiences, socialization and culture of a person. Once values are formed, they mostly remain stable because people normally do not tend to challenge their values. People can however change their values over time, if their original values are challenged frequently they could start to re-evaluate them. Still, if we would like to change values, it is best to start at a young age. According to Schwartz (1992), human values can be grouped into 10 clusters which can be plotted in two
dimensions: self-enhancement versus self-transcendence and conservation versus openness to change. Schultz & Zelezny (1999) show that environmental attitudes are related to values and there are several studies that show that pro-environmental behaviour is negatively correlated with self-enhancement values and positively correlated with self-transcendent values (Karp 1996; Klöckner 2013; Nordlund and Garvill 2002; Thøgersen and Ölander 2002).
Steg et al. (2014) categorized two types of self-enhancement values (hedonic and egoistic values) and two types of self-transcendence values (altruistic and biospheric values) that are important for understanding environmental behaviour. Hedonic values follow an interest in
improving feelings and reducing effort, while egoistic values are about increasing or protecting his or her own resources. People with great altruistic values care for the welfare of others while people with high biospheric values have a strong connection with the environment. The results were that in general, people with strong self-enhancement values are less likely to express pro-environmental behaviour compared to people with strong self-transcendence values. Of these four values, biospheric values are most related to pro-environmental behaviour, so if we would like to enhance pro-environmental behaviour it is important to understand how biospheric values develop and drive pro-environmental behaviour. Altruistic values also play a large role in pro-environmental behaviour and it is crucial to decrease the friction between egoistic values on the one hand and altruistic and biospheric values on the other hand at the same time (De Groot & Steg, 2009).
12 There are different views in the literature regarding whether altruistic values or biospheric values play the largest role in pro-environmental behaviour. A study by Howell (2013) found that the main motivator of action for people who have adopted a pro-environmental lifestyle was not always their biospheric values, her participants scored altruistic values significantly higher than biospheric values. They stated that the main motivator for adopting this new lifestyle was more because of human right and community issues. However, unlike altruistic values, biospheric values also relate to environmental self-identity, which is defined as the extent to which you see yourself as a type of person who acts environmentally-friendly (Van der Werff, Steg & Keizer, 2013). These two constructs are related in a wide range of environmental preferences, intentions, and behaviours, including behaviours reflecting direct and indirect energy use, product choices, intentions to save energy, curtailment and efficiency behaviours, preferences and willingness to pay for green energy. Van der Werff, Steg & Keizer (2013) propose that both constructs can be seen as a general antecedent of environmental preferences, intentions, and behaviour. It Is therefore assumed that people with high biospheric values and environmental self-identity will mostly behave pro-environmental since they see themselves as a person who does so, which is different for other people who could sometimes behave pro-environmental and sometimes not. Biospheric values also relate to problem awareness, while people with high egoistic values have lower problem awareness (Steg et al. 2014). Steg (2016) discussed five strategies that can strengthen biospheric values and thereby enhance
pro-environmental behaviour: changing the costs and benefits of behaviour, reducing cognitive effort, providing information and feedback on costs and benefits, strategies that take advantage of people's need to be consistent, and social influence strategies. De Groot & Steg (2009) reported that people with high biospheric values will mainly build their decision to behave pro-environmentally on the perceived costs and benefits for the ecosystem and the total biosphere. For people with high biospheric values it is therefore important to create products where the benefits for the ecosystem are larger than the perceived costs, otherwise they will not express pro-environmental behaviour.
13 Regarding biospheric values and pro-environmental behaviour, it is expected that people with high biospheric values will express more pro-environmental behaviour than people with low biospheric values because people with high biospheric values have a stronger connection with the environment. In the interest of better understanding this relationship, the following hypothesis was formulated;
H1: There is a positive relationship between biospheric values and pro-environmental
behaviour meaning that participants with high biospheric values behave more
pro-environmental than people with low biospheric values.
In the interest of reducing the pressures on the environment it is important to make consumers express more environmentally friendly behaviours. Because biospheric values are an important driver of pro-environmental behaviour we have to learn more about how they are
expressed and how biospheric values can be promoted in such a way that people will manifest more pro-environmental behaviour.
14
2.3 Cognitive distraction; automatic versus reasoned responses
According to the social intuitionist model developed by Haidt (2001), moral decisions, like the choice between cooperation and defection are in most cases the result of quick automatic intuitions. These automatic responses are formed by social and cultural influences that become incorporated within one’s self during the course of personality development. The role of a thinking process is largely restricted to and based on one’s feelings about what is seen as right or wrong. These quick and automatic judgments in dilemma situations generate a movement either to cooperate
(prosocials) or to defect (proselfs). Cornelissen, Dewitte & Warlop (2011) suggest that decisions are made through either automatic responses or through more reasoned processes. They tested a model of prosocial versus self-interested behaviour. Their idea was that social values that search for the interests of others or self-interest are expressed automatically and they argue that if people make decisions through a more reasoned process, they will think more about themselves compared to the automatic process. In order to test automatic expression versus more reasoned expressions, half of the participants in their study were cognitively distracted by having to memorize a difficult number, while the other half had to memorize an easier number. The cognitively distracted participants with the difficult number represented the group with automatic expressions, because their working memory capacity is influenced by remembering the number. The participants with the easier number represented the group with more reasoned responses since their working memory capacity is put under less pressure. Their results show that social values are indeed expressed automatically and if based on a reasoned process, the generosity of prosocial people tends to decrease to the level of self-interested people. Due to the fact that the way values are expressed by consumers plays an important role in the link between values and actual behaviour, it is critical to investigate which way of expressing values stimulates this link. If the decisions we make regarding biospheric values and pro-environmental behaviour are made more through reasoning, the attractive alternatives, products that are cheaper but pollute our environment will likely become more difficult to resist.
15 The results of Cornelissen, Dewitte & Warlop (2011) show that there is a difference between
prosocial and self-interested people in such a way that if cognitively distracted, prosocial people are more generous than self-interested people. However, if not cognitively distracted, the generosity of prosocial and self-interested people tends to become the same level. Thus, cognitive distraction influences the relationship between social values and the generosity of the participants.
In this study, instead of measuring the generosity of people, I will measure
pro-environmental behaviour and instead of social values I will investigate biospheric values. Because someone with high biospheric values will probably behave more pro-environmental than someone with low biospheric values and due to the fact that if cognitively distracted, prosocial people tend to be more generous than self-interested people, it is expected that the relationship between
biospheric values and pro-environmental behaviour is influenced if people are cognitively distracted. The relationship is influenced in such a way that if cognitively distracted, people with high biospheric values will behave more pro-environmental than people with low biospheric values. Also, if not cognitively distracted, the differences in pro-environmental behaviour between people with high and low biospheric values will become smaller because they both have time to reason and are less dependent on their automatic moral judgements. In order to investigate if this assumption is right, the following hypothesis was formulated;
H2: Cognitive distraction moderates the relationship between biospheric values and
pro-environmental behaviour in such a way that the positive relationship between biospheric
values and pro-environmental behaviour becomes stronger for people who are cognitively
distracted.
16
2.4 Conceptual model
Figure 1: Conceptual model
3. Methods
3.1 Procedure
In order to research to what extent biospheric values are expressed automatically when it comes to pro-environmental behaviour, I used an experimental approach. With experimental approaches the effects of intentional manipulations to the environment are examined. Experiments usually have high internal validity and are therefore favourable for determining causal effects (Saunders & Lewis, 2009). In this study, I consciously manipulated the environment in such a way that I created and compared two groups, one cognitively distracted group and one control group. This was done by a cognitive distraction method and allows me to analyse the effects of these two groups on pro-environmental behaviour. Because this study analyses data collected at a specific point in time, a cross-sectional quantitative study by means of a survey was conducted in order to answer the hypotheses. Participants were gathered by asking fellow students at the University of Amsterdam and the Hogeschool van Amsterdam face to face if they would be willing to help me with an experiment. This was done face to face because I had to be present to ensure that the participants
17 were not able to cheat. People who were willing to help had to fill in a questionnaire which was made in Qualtrics and only available in English. The participants first got a short introduction and were told a false purpose of the study, which was the ability to make decisions after having to memorize a number. They were not told the true purpose because that could affect the results as it could put the participants in a pro-environmental state. Hereafter, I would flip a coin and depending on the outcome the participants would be assigned to either the cognitive distracted group or the control group, which ensured that there was no selection bias. Participants in the cognitively distracted group had to memorize a random seven-digit number while the other group had to memorize a structured seven-digit number (1234567) which is obviously easier. At the end of the questionnaire, participants had to reproduce the memorized number. Memorizing the random number makes it much more difficult to deeply think about a decision, while memorizing the structured number allows participants to make more use of their reasoning capacity. Data of participants who were unable to reproduce the memorized was automatically discarded.
3.2 Sample
Because I wanted to analyse the differences in pro-environmental behaviour between the group with the random number and the group with the structured number, it was important to have a similar people in the two groups. If there is a large difference between the two groups it would be more difficult to know if the results are caused by that difference or because half of them are being cognitively distracted. Therefore, the sample consists of only higher educated students. All
respondents were between 18-34 years old, 61% of them already had their Bachelor’s degree and almost 59% were males. After discarding respondents who were not able to reproduce the memorized number there is a total sample of 80 respondents.
18
3.3 Measurements
3.3.1 Pro-environmental behaviour
There are different methods to investigate pro-environmental behaviour, one of them is to give participants a task measuring pro-environmental behaviour. De Groot & Steg (2008) developed such a task by looking at the donation intention of participants towards environmental organizations. This was done because they found that people with high biospheric values have a preference for donating to environmental organizations. Donation intention was measured by means of an instrument developed by De Groot & Steg (2008). The donation intention of participants was measured by asking how much money they would donate to a humanitarian or an environmental organisation if they would have €10. The question was: “Suppose you have 10 Euro that you are willing to donate to charity. Below, I list nine pairs of donating choices in which you can choose to donate between two organisations. Please indicate in every situation how you would divide your €10.” In five of the nine cases, respondents were given a choice between a humanitarian and an environmental organisation, the other 4 situations were not relevant to mask the true purpose of the task. (De Groot & Steg, 2008). Allocating €5 to each organisation was not possible. It was
hypothesized that people who are more altruistically oriented donate more to humanitarian
organizations, whereas people who are more biospherically oriented donate more to environmental organizations. Scores on donating intention were calculated by summing up the amount of money a participant gave to an environmental organisation. The scale score could range from 0 ‘no donations to environmental organisations and €50 Euro to humanitarian organisations’ to 50 “€50 Euro to environmental organisations and no donations to humanitarian organisations”. A score of ‘35’ means that a respondent chose to donate €35 to an environmental and €15 to a humanitarian organisation after five donations.
19
3.3.2 Biospheric values
Biospheric values were measured with the use of a 4-item questionnaire which has also been used in previous studies (De Groot & Steg, 2008; De Groot & Steg, 2009; De Groot & Steg, 2010; Ruepert et al., 2016). The participants rated the importance of each item on a 9-point Likert scale from “−1 (opposed to my values)” to “7 (of supreme importance)”. Biospheric values were
represented by 4 items (Respecting the earth: harmony with other species; Unity with nature: fitting into nature; Protecting the environment: preserving nature; Preventing pollution: protecting natural resources). With these four items I will be able to examine if there is a relationship between
biospheric values and pro-environmental behaviour.
3.3.3 Cognitive distraction
In order to test cognitive distraction, I used the same approach as Cornelissen, Dewitte & Warlop (2011). On the second page of the questionnaire just after the introduction, half of the participants were instructed to memorize a seven-digit number (5684524) while the other half had to memorize an easier and structured seven-digit number (1234567). Both numbers remained on the screen for 60 seconds before automatically continuing to the next page. Keeping a difficult number active in working memory taxes cognitive resources and thereby prevents active contemplation of the decision. Memorizing the structured number is not effortful, allowing for remaining reasoning capacity. After the questionnaire, participants were asked to reproduce the number they had memorized. The data of participants who failed to reproduce the number correctly was discarded right away (Cornelissen, Dewitte & Warlop, 2011). By cognitively distracting half of the respondents I can compare automatic responses for the group with the random number versus reasoned responses for the group with the structured number and see if there are differences in pro-environmental behaviour and biospheric values between both groups.
20
3.3.4 Control variables
The age, gender, education and income level of the families of participants were taken into account in the analyses as control variables. This makes it possible to check whether any of these variables influence pro-environmental behaviour. At the end of the questionnaire, respondents were asked four demographic questions, the responses were extracted and used for these control
variables. Age was measured with a 9-point Likert scale ranging from “Under 18” to “85 or older”, education was measured with the use of a 6-point Likert scale ranging from “Less than high school” to “Doctorate” and the income level of the families of participants was measured with a 9-point Likert scale which ranged from “Less than €15.000 a year” to “More than €150.000 a year”.
3.4 Analyses
The first part of my analyses was to create a dummy variable for the nominal variable cognitive distraction, the structured numbers were transformed into zero and the random numbers were transformed into one. Then I had to control the assumptions before I could start with the actual analyses. The next step was to check for errors and missing values and with the use of a frequency check it was possible to look for errors, but these were not found. Participants with missing values were excluded listwise to ensure that only data with no missing values were taken into account in the analyses. Hereafter, it was important to make sure that the variables were normally distributed and had no outliers. I used several methods to test for normality. First of all, I checked the histogram and Q-Q plots for the three variables, the histograms were bell-shaped and the Q-Q plots were normal. The skewness and kurtosis for both variables were between 0.4 and 0.6 which fell between -1 and 1 and therefore indicate a normal distribution.
In order to check for outliers, I did a multiple linear regression with the three variables and selected Cooks, Leverage and Mahalanobis in the save box which resulted in the construction of three new columns, one for Cooks, one for Leverage and one for Mahalanobis. With these columns and a cut-off score which I calculated for each of these three approaches I could determine if any of
21 the data was identified as an outlier. If any data would be identified as an outlier by at least two of the three I would have to delete the data. The outcome showed that none of my data was identified as an outlier by at least two approaches and therefore there were no outliers. In order to check multivariate normality, I looked at the plots of the output of the regression. The plots showed a bell-shaped histogram which indicates a normal distribution, the P-P plot showed good linearity and the residual scatterplot looked good which indicates that there are no homogeneity or homoscedasticity problems. Another issue that had to be considered is that moderation can easily create a
multicollinearity problem because I looked at two variables and their interaction which means that the interaction is not unique and can be calculated by multiplying the other two variables. One way to solve this is to mean centre the variables which puts the mean of the variable at zero and creates a standard deviation of one. The correlation output between the two independent variables showed a correlation of 0.306** which does not seem problematic.
Before running any analysis, it is important to check if there are counter-indicative items in the dataset. For the variables biospheric values and pro-environmental behaviour it was therefore necessary to conduct a reliability analysis in order to examine if the data collection techniques yield consistent findings (Saunders et al., 2009). The biospheric values scale has high reliability, with Cronbach’s Alpha = .864. The pro-environmental behaviour scale also has high reliability, with Cronbach’s Alpha = .781. Because both scales are above .70 and are therefore considered good. The corrected item-total correlations indicate that all the items for both variables have a good correlation with the total score of the scale (all above .30). For both variables none of the items would affect reliability more than 0.10 if they were to be deleted, so I could continue with the actual analyses.
22
4. Results
4.1 Correlations, means, SE’s
With the use of descriptive statistics and bivariate correlations in SPSS it was possible to generate the means, standard deviations and correlations of the variables. The output shows that none of the variables are highly correlated with each other, some variables are however moderately correlated.
Table 1: Descriptive statistics and correlations
Variables Mean SD 1 2 3 4 5 6 7 1. Gender 1.41 0.50 - 2. Age 2.53 0.50 -.22* - 3. Education 4.11 0.62 .03 .42** - 4. Income 4.16 2.21 .007 .05 .19 - 5. Cognitive Distraction 0.50 0.50 -.13 .10 .18 .13 - 6. Biospheric Values 6.81 1.40 -.11 .29** .49** .04 .31** (.86) 7. Donation Intention 23.13 8.23 -.05 0.15 .27* .05 .25* .44** (.78)
**. Correlation is significant at the 0.01 level (2-tailed). *. Correlation is significant at the 0.05 level (2-tailed).
4.2 Regression analysis biospheric values
In order to answer the first hypothesis, I analysed the relationship between biospheric values and pro-environmental behaviour to examine the direct effects. A hierarchical regression analysis was performed because I wanted to control for the variables Gender, Age, Education and Income. With this regression, it is possible to test whether the effect of biospheric values on
pro-environmental behaviour is independent of the effect of the control variables on pro-pro-environmental behaviour. The output of the regression can be found on the next page.
23
Table 2: Hierarchical regression model of pro-environmental behaviour
R R2 R2 Change B SE β T Step 1 .274 .075 .075 Gender -.607 1.896 -.037 -.320 Age .490 2.060 .030 .238 Education 3.419 1.668 .256* 2.049 Income -.008 .421 -.002 -.019 Step 2 .442** .195** .120** Gender -.114 1.786 -.007 -.064 Age -.058 1.941 -.004 -.030 Education .945 1.734 .071 .545 Income .072 .396 .019 .183 Biospheric values 3.308 .994 .402** 3.328 Statistical significance: *p <.05; **p <.01; ***p <.001
In step 1 the model is not statistically significant (F (4,75) = 1.520; p>.05), even though the effect of the control variable education on pro-environmental behaviour is (β = .256, p <0.05). However, by including biospheric values as a predictor in step 2 the overall model becomes significant (F (1,74) = 3.594; p<.01) and accounts for 19.5% of the variance of pro-environmental behaviour. The effect of biospheric values on pro-environmental behaviour is statistically significant (β = .402, p <0.01).
Hypothesis 1 suggests that there is a positive relationship between biospheric values and pro-environmental behaviour, meaning that participants with high biospheric values express more pro-environmental behaviour. The results of the regression analysis show that biospheric values indeed has a significant positive direct effect on pro-environmental behaviour. Therefore, hypothesis 1 can be supported.
4.3 Moderation analysis without control variables
In order to answer the second and last hypothesis, I performed a moderation analysis with the use of Process in SPSS. The first step was to mean centre the numerical variables, this was done in Process while testing for moderation. With Process I could run the moderation analysis with and without covariates, first I will show the analysis without control variables and later on with control variables. Table 3 and 4 show the results of the moderation analysis without control variables.
24
Table 3: Moderation output without control variables
Coefficient SE t p Intercept i₁ 22.48 0.85 26.60 <0.001*** Cognitive Distraction (X) c₁ 1.92 1.69 1.14 0.258 Biospheric Values (M) c₂ 2.60 0.62 4.22 <0.001*** Biospheric Values * c₃ 3.08 1.23 2.50 0.015* Cognitive Distraction (XM) R²=0.267,p<0.05** F(3,76)=9.219 Statistical significance: *p <.05; **p <.01; ***p <.001
Table 4: Conditional effects output without control variables
Conditional effect of Cognitive Distraction (X) on Donation Intention (Y) at levels of Biospheric Values (M)
Effect SE t p
Low Biospheric Values -2.385 2.479 -0.962 0.339
Average Biospheric
Values 1.924 1.689 1.139 0.258
High Biospheric Values 6.234 2.348 2.655 0.009**
Statistical significance: *p <.05; **p <.01; ***p <.001
The moderation output shows that cognitive distraction alone does not have a significant effect on pro-environmental behaviour (c1=1.92, p>0.05). Biospheric values however does have a significant effect on pro-environmental behaviour (c2=2.60, p<0.001), meaning that 2.60 is the estimated difference in pro-environmental behaviour between 2 participants who differ by one unit in biospheric values, among those scoring zero on cognitive distraction (i.e., among participants who were not cognitively distracted and had to memorize the easier structured number). The moderation output also shows a significant interaction between cognitive distraction and biospheric values on environmental behaviour (c3=3.08, p<0.05). Thus, the effect of biospheric values on pro-environmental behaviour is moderated by cognitive distraction. More specifically, as participants become cognitively distracted, the difference in pro-environmental behaviour between participants with high and low biospheric values increases with 3.08 units. The model accounts for 26.7% of variance in pro-environmental behaviour, which is 7.2% more than the first model which only included biospheric values and the control variables.
25 The output of the conditional effects shows the slopes for cognitive distraction predicting pro-environmental behaviour at each level of biospheric values. The conditional effects indicate that the moderating effect of cognitive distraction on the relationship between biospheric values and pro-environmental behaviour is only significant for the participants with high biospheric values (effect= 6.23, SE = 2.35, CI: 1.56 to 10.91). For participants with low (“low biospheric values” effect= -2.39, SE = 2.48, CI: -7.32 to 2.55) and average biospheric values (“average biospheric values” effect= 1.92, SE = 1.69, CI: -1.44 to 5.29), the moderating effect is not significant. Together with the output above came data for visualizing the conditional effect of the focal predictor, which was used to build a graph of biospheric values with pro-environmental behaviour for both the cognitive distracted group and structured group. As can be seen in the graph below, the slope linking cognitive distraction and pro-environmental behaviour is only positive for participants with average or high biospheric values. In other words, being cognitively distracted only significantly affects pro-environmental behaviour for respondent with high biospheric values, although such a trend has the same direction for
respondents with average biospheric values.
26 With the use of the Johnson-Neyman significance region(s) output it is possible to take a closer look at which levels of biospheric values cognitive distraction and pro-environmental
behaviour are significantly related. The output shows that when biospheric values are at least 7.33, cognitive distraction and pro-environmental behaviour are significantly related t(76)=1.99, p=0.05 and b=3.54. As the score for biospheric values increases, the relationship between cognitive distraction and pro-environmental behaviour becomes more positive with the highest biospheric values score of 9, b=8.67, t(76) and p=0.0066.
4.4 Moderation analysis with control variables
The final step in the analyses was to examine the moderating effect with control variables, table 5 and 6 show the results of the analysis. The overall model is statistically significant and explains 26.9% of the variance in pro-environmental behaviour (F (7,72) = 3.7836, p=.0015), which is only 0.2% more than the model without control variables. Therefore, including the control variables into the model does only increase the predictive power slightly. None of the control variables are statistically significant and their coefficients are all very low.
After including the control variables, cognitive distraction is still insignificant (c1=1.98, p>0.05), while biospheric values are still statistically significant (c2=2.50, p<0.001), which means that 2.5 is the estimated difference in pro-environmental behaviour between 2 participants who differ by one unit in biospheric values, among those scoring zero on cognitive distraction (i.e., among
participants who were not cognitively distracted and had to memorize the easier structured number). The output also shows that the interaction is statistically significant (c3=3.09, p<0.05), so cognitive distraction still moderates the relationship between biospheric values and
environmental behaviour. As participants become cognitively distracted, the difference in pro-environmental behaviour between participants with high and low biospheric values increases with 3.09 units.
27
Table 5: Moderation output with control variables
Statistical significance: *p <.05; **p <.01; ***p <.001
Table 6: Conditional effects output with control variables
Conditional effect of Cognitive Distraction (X) on Donation Intention (Y) at levels of Biospheric Values (M)
Effect SE t p
Low Biospheric Values -2.338 2.459 -0.917 0.362
Average Biospheric
Values 1.983 1.756 1.130 0.262
High Biospheric Values 6.304 2.472 2.550 0.013*
Statistical significance: *p <.05; **p <.01; ***p <.001
The output of the conditional effects shows the slopes for cognitive distraction predicting pro-environmental behaviour at each level of biospheric values. The results indicate that the moderating effect of cognitive distraction on the relationship between biospheric values and pro-environmental behaviour is again only significant for the group of participants with high biospheric values (effect= 6.30, SE = 2.47, CI: 1.38 to 11.23). For the participants with low (“low biospheric values” effect= -2.39, SE = 2.46, CI: -7.42 to 2.74) and average biospheric values (“average”
biospheric values” effect= 1.98, SE = 1.76, CI: -1.52 to 5.48), the moderating effect is not significant.
Coefficient SE t p Intercept i₁ 20.31 7.40 2.75 <0.05* Cognitive Distraction (X) c₁ 1.98 1.76 1.13 0.262 Biospheric Values (M) c₂ 2.50 0.72 3.45 <0.001*** Biospheric Values * c₃ 3.09 1.28 2.41 0.019* Cognitive Distraction (XM) Gender c₄ 0.31 1.74 0.18 0.859 Age c₅ -0.16 1.88 -0.08 0.934 Education c₆ 0.62 1.68 0.36 0.717 Income c₇ -0.09 0.39 -0.24 0.810 R²=0.269,p<0.05 F(7,72)=3,784
28 Together with the output above came data for visualizing the conditional effect of the focal
predictor, which was used to build a graph of biospheric values with pro-environmental behaviour for both the cognitive distracted group and the structured group. Again, the graph below shows that the slope linking cognitive distraction and pro-environmental behaviour is only positive for
participants with average or high biospheric values. Just as without control variables, being cognitively distracted only significantly affects pro-environmental behaviour for respondent with high biospheric values, although the line for respondents with average biospheric values also has a positive direction.
Figure 3: Graph of simple slopes with control variables
With the use of the Johnson-Neyman significance region(s) output I could take a closer look at which biospheric values cognitive distraction and pro-environmental behaviour are related. The output showed that when biospheric values levels are at least 7.33, cognitive distraction and pro-environmental behaviour are significantly related b= 3.75, t (72) =1.99, p=0.05. As the score for biospheric values increases, the relationship between cognitive distraction and pro-environmental behaviour becomes more positive with the highest biospheric values score of 9, b=8.74, t (72) = 2.68 and p<0.01.
29 Hypothesis 2 suggests that cognitive distraction moderates the relationship between biospheric values and pro-environmental behaviour in such a way that the positive relationship between
biospheric values and pro-environmental behaviour becomes stronger for people who are cognitively distracted. Both models (without and with control variables) show that there is a significant
interaction effect (c3=3.08, p=0.015 and c3=3.09, p=0.019). Thus, the effect of biospheric values and cognitive distraction on pro-environmental behaviour are co-dependant. However, the output also shows that the relationship between cognitive distraction and pro-environmental behaviour is only significant for certain levels of biospheric values. For participants with low biospheric values for instance, the relationship is not significant. Therefore, H3 is partially supported because cognitive distraction only moderates the relationship between biospheric values and pro-environmental behaviour for participants with high biospheric values.
The conceptual model with results is illustrated below, the value behind H1 represents the β-value of biospheric β-values of the regression output and the β-value behind H2 represents the
coefficient value of the interaction of the moderation output.
Figure 4: Conceptual model with results
30
5. Discussion
5.1 General discussion
In this study, the relationship between cognitive distraction, biospheric values and environmental behaviour has been examined. The expectation was to find differences in
pro-environmental behaviour between participants with dissimilar biospheric values. How this related to participants who were and were not cognitively distracted was also analysed. This study was
designed to answer the following research question; To what extent are biospheric values expressed
automatically when it comes to pro-environmental behaviour? In order to answer the research
question, two models were tested. The first model tested if biospheric values directly affects pro-environmental behaviour, the results showed that biospheric values has a significant positive direct effect on pro-environmental behaviour and explains 19.5% of the variation in pro-environmental behaviour. In the second model, the results demonstrate that cognitive distraction moderates the relationship between biospheric values and pro-environmental behaviour. The results of the
moderation analysis revealed that the interaction effect is statistically significant and explains 26.9% of the variance in pro-environmental behaviour, which is 7.4% more than the direct effect of
biospheric values on pro-environmental behaviour. Furthermore, cognitive distraction only moderates the relationship between biospheric values and pro-environmental behaviour for
participants with high biospheric values. For participants with low and average biospheric values this relationship was not significant. Including the control variables into the model does only increase the predictive power slightly and none of the control variables were statistically significant. Regarding the main research question, the results of the analyses demonstrate that biospheric values are expressed automatically when it comes to pro-environmental behaviour although only for people with high biospheric values.
31
5.2 Theoretical implications
This study contributes to the literature and gives empirical support for two important findings. The first finding supports the positive direct effect of biospheric values on
environmental behaviour, meaning that participants with high biospheric values express more pro-environmental behaviour than participants with low biospheric values. This is in line with the existing literature and shows that biospheric values are indeed related to pro-environmental behaviour. (Karp 1996; Klöckner 2013; Nordlund and Garvill 2002; Thøgersen and Ölander 2002; Schultz & Zelezny, 1999). Although the first finding does not progress the level of knowledge, it supports the existence of this relationship and thereby contributes to the literature. The first finding is also important because replicating a study and finding the same results is not always easy in behavioural science. Aarts et al. (2015) conducted replications of 100 studies published in three psychology journals and found that while ninety seven percent of the original studies had significant results, only thirty six percent of the replications had significant results. Because this study demonstrates that biospheric values drive pro-environmental behaviour, it is relevant to explore what can increase the biospheric values of the public. Although biospheric values might predict pro-environmental behaviour, it is important to say that it does not necessarily lead to actual pro-environmental action (Verplanken & Holland, 2002). Just because someone has high biospheric values does not mean that that same person recycles waste, eats less meat and/or only uses renewable energy sources. There are other factors which also play a role such as, the locus of control (Engqvist Jonsson & Nilsson, 2014), situational variables (Corraliza & Berenguer, 2000), norms, the influence of habit and conflicts with other values (Verplanken & Wood, 2006). For academics and future research, it is important to investigate how biospheric values relate to not only environmental behaviour, but also pro-environmental action.
32 The second and most important finding is the moderating effect of cognitive distraction on the relationship between biospheric values and pro-environmental behaviour. This study is the first to find support for this relationship and thereby extends the literature. So, besides that cognitive distraction influences the relationship between social values and generosity (Cornelissen, Dewitte & Warlop, 2011), now we can also say that cognitive distraction influences the relationship between biospheric values and pro-environmental behaviour, although only for people with high biospheric values. There is a difference between people with high and low biospheric values in such a way that if cognitively distracted, people with high biospheric values express more pro-environmental behaviour than people with low biospheric values. Yet, if not cognitively distracted, the differences in pro-environmental behaviour between both groups decreases. This suggests that if people would make more decisions based on their automatic system instead of their reasoning system, there will probably be an increase in the expressions of pro-environmental behaviour. According to Dewitte, Bruyneel & Geyskens (2009), changing values is much like learning to drive a car. At first, it is in our reasoning system and after training it goes to the automatic system. How to change and train our values in such a way that it goes to our automatic system is an important topic which has to be addressed in future research.
To sum up, this study adds to the existing literature because it presents empirical evidence for the relationship between biospheric values and pro-environmental behaviour and it is the first to show a significant moderating effect of cognitive distraction on the relationship between biospheric values and pro-environmental behaviour.
33
5.3 Limitations and future research
The outcomes of this study raise two new important questions, namely; how can the
biospheric values of people be enhanced, especially for people with low biospheric values? and how can we get people to make more decisions based on their automatic system? Both of these questions
need to be investigated and answered in future research in order to increase pro-environmental behaviour. In regard to this study, it would be good to increase the sample size and include other groups of respondents as well instead of only higher educated students. Although the sample was not a good reflection of the Dutch society, it was a homogeneous group which was good for analysing the effects of the model. It is imaginable that by including diverse groups of respondents there will be bigger differences between individuals. It would be interesting to see what that does to the results. The methodology used in this study had some limitations. I wanted to have at least 80 participants, 40 in the random group and 40 in the structured group. In order to randomly assign participants into the two groups I flipped a coin, heads meant the participant had to memorize the random number and tails meant the participant had to memorize the structured number. At first this seemed like a good way to assign participants, but later on I discovered that a large portion of the participants in the random group were not able to memorize the number. Eventually I reached 40 participants for the structured group but I still needed around 15 people for the random group. Because of this, the last participants were all assigned to the random group. To reduce difficulties with randomly assigning participants to a group in the future, I would change the research design slightly. It might be better to make the design more experimental and invite a group of participants at once. The participants will then be assigned to a group when they enter the classroom. If some participants in the random group are not able to remember the number, I invite another group of participants until I have enough people in both groups. Because some participants might have positive or negative associations with the organizations used in the task measuring
pro-environmental behaviour, people could be donating to an organization not by reading the mission statement but by these associations. Therefore, it could be better to leave out the names of the
34 organizations in future research. According to Hinds & Sparks (2008), people who grew up in a rural area report more positive orientations towards the environment than people who grew up in an urban area. This could be taken into account as a control variable in future research. Even though the results show a relationship between biospheric values and pro-environmental behaviour, it does not imply that biospheric values are the specific motivator of pro-environmental behaviour, for future research it would be interesting to investigate specific motivations for climate action (Howell & Allen, 2017). Someone with high biospheric values does not necessarily cycles to work, recycles waste, eats less meat or uses rainwater to water their plants. In order to analyse if that person does more for the environment than someone with lower biospheric values it would be interesting to add questions about how much that person is currently doing to reduce the pressures on the environment.
5.4 Practical implications
Pickett-Baker & Ozaki (2008) reviewed consumer behaviour and advertising to investigate how consumers can be persuaded to buy more pro-environmental products. They suggest that the market for environmental products could be exploited more for consumers with high
pro-environmental values. Laroche, Bergeron & Barbaro-Forleo (2001) suggest that for people who find it important to behave environmentally friendly, marketers should communicate that buying EFP’s can have a significant impact on the environment. With the use of targeted advertising campaigns, marketers can encourage positive attitudes and behaviours and make people aware that their consumptions patterns matter. The results of this research demonstrate that people with high biospheric values express more pro-environmental behaviour than people with low biospheric values. Marketers busy with pro-environmental advertising therefore have to identify and target consumers with high biospheric values. This could increase the market share of EFP’s and reduce the pressures on the environment.
35 The results of this study also suggest that people express biospheric values automatically. This is practically important because making more people express their biospheric values automatically will probably reduce the pressures on the environment. Managers should try to make their customers express their values more automatically instead of through reasoning. The results indicate that one way of doing this is by giving consumers a difficult task. Consumers will then base more decisions on automatic judgements and the results of this research indicate that if based on automatic
judgements, people with high biospheric values tend to behave more pro-environmental. For people with low biospheric values however, the results show that it does not really matter if they are cognitively distracted or not. For managers, marketers and policy makers it is therefore important to reduce the amount of people with low biospheric values. Although marketing alone can probably not change people’s lives radically, it does present ways of changing consumer behaviour and influencing attitudes and beliefs (Jones et al., 2008). For instance, by frequently challenging the values of
consumers with low biospheric values. However, because values are generally formed during childhood, it is best to start at a young age with developing a strong connection to the environment (Schwartz, 1992). Companies could therefore make more attempts in involving children. If these children are brought up in a world where pro-environmental behaviour is the only good thing to do, chances are high that they will express this behaviour automatically. In addition, managers could also focus more on environmental self-identity, which is defined as the extent to which you see yourself as a type of person who acts environmentally-friendly and relates to biospheric values (Van der Werff, Steg & Keizer, 2013). If more people would see themselves as such a person, it is likely that they will behave more pro-environmental. If for instance well known athletes, musicians and other role models would be used more often in pro-environmental advertising, environmental self-identity will most likely go up since people identify themselves with these role models. Because people who grew up in a rural area report more positive orientations towards the environment than people who grew up in an urban area, practitioners can investigate how we can change urban areas to reduce this difference (Hinds & Sparks, 2008). This is needed since more and more people are moving to
36 urban areas. Marketers can also try to identify and target people who grew up in rural areas, since they have more positive orientations with the environment.
However, managers cannot forget other value orientations, especially altruistic values (De Groot & Steg, 2009). Goldstein, Cialdini & Griskevicius (2008) tested the influence of different signs that asked hotel guests to participate in an environmental conservation program. The signs either focused on environmental protection or on descriptive norms. The descriptive norms informed guests that a large group of other guests were also participating in the program. The results showed that guests participated more often to signs with descriptive norms compared to signs that focused only on environmental protection. The results of a study by Howell (2013) show that some of the people they interviewed are more concerned about the poorer people who will suffer directly from climate change than about climate change alone. Because the underlying values to behave pro-environmental could vary from person to person, it is important to link climate action not only to environmental issues but to humanitarian issues as well. Thus, there are different ways to motivate people for climate change action and it would not be smart for climate changers to focus only on one single value orientation.
37
6. Conclusion
Consumers can help with reducing the pressures on the environment by expressing more pro-environmental behaviour. This study adds insights to the literature by suggesting that biospheric values are one of the main drivers of pro-environmental behaviour and that these values can be expressed automatically or more through reasoning. By cognitively distracting half of the participants in the study it was possible to analyse the effect of automatic and reasoned expressions and how they relate to biospheric values and pro-environmental behaviour. The results showed that biospheric values indeed relate to pro-environmental behaviour, which demonstrates that it is relevant to examine how we can increase the biospheric values of the public. This is also practically relevant for organizations in pro-environmental marketing because consumers with high biospheric values are more likely to express pro-environmental behaviour. Therefore, they have to identify and target these consumers. Furthermore, the results of the moderation analysis revealed that when expressed automatically, participants with high biospheric values show more pro-environmental behaviour than when expressed through reasoning. For participants with low and average biospheric values there was no significant difference in pro-environmental behaviour between automatic and reasoned expressions. This indicates that if the biospheric values of people can be increased and if people would make more decisions based on their automatic system instead of their reasoning system, there will probably be an increase in the expressions of pro-environmental behaviour which makes it an important topic for future research.
38
7. References
Aarts, A. A., Anderson, J. E., Anderson, C. J., Attridge, P. R., Attwood, A., Axt, J., ... & Bartmess, E. (2015). Estimating the reproducibility of psychological science. Science, 349(6251), 253-267.
Ajzen, I. (1991). The theory of planned behavior. Organizational behavior and human decision
processes, 50(2), 179-211.
Cornelissen, G., Dewitte, S., & Warlop, L. (2011). Are social value orientations expressed automatically? decision making in the dictator game. Personality and Social Psychology
Bulletin, 37(8), 1080-1090.
Corraliza, J. A., & Berenguer, J. (2000). Environmental values, beliefs, and actions: A situational approach. Environment and behavior, 32(6), 832-848.
De Groot, J. I., & Steg, L. (2008). Value orientations to explain beliefs related to environmental significant behavior: How to measure egoistic, altruistic, and biospheric value
orientations. Environment and Behavior, 40(3), 330-354.
De Groot, J. I., & Steg, L. (2009). Mean or green: which values can promote stable pro‐environmental behavior? Conservation Letters, 2(2), 61-66.
De Groot, J. I., & Steg, L. (2010). Relationships between value orientations, self-determined motivational types and pro-environmental behavioural intentions. Journal of Environmental
Psychology, 30(4), 368-378.
Dewitte, S., Bruyneel, S., & Geyskens, K. (2009). Self-regulating enhances self-regulation in subsequent consumer decisions involving similar response conflicts. Journal of Consumer
39 Engqvist Jonsson, A. K., & Nilsson, A. (2014). Exploring the relationship between values and pro-environmental behaviour: the influence of locus of control. Environmental Values, 23(3), 297-314.
Gatersleben, B., Murtagh, N., & Abrahamse, W. (2014). Values, identity and pro-environmental behaviour. Contemporary Social Science, 9(4), 374-392.
Gatersleben, B., White, E., Abrahamse, W., Jackson, T., & Uzzell, D. (2010). Values and sustainable lifestyles. Architectural Science Review, 53(1), 37-50.
Geller, E. S. (2002). The challenge of increasing proenvironment behavior. Handbook of
environmental psychology, 2, 525-540.
Goldstein, N. J., Cialdini, R. B., & Griskevicius, V. (2008). A room with a viewpoint: Using social norms to motivate environmental conservation in hotels. Journal of consumer Research, 35(3), 472-482.
Griskevicius, V., Tybur, J. M., & Van den Bergh, B. (2010). Going green to be seen: status, reputation, and conspicuous conservation. Journal of personality and social psychology, 98(3), 392.
Griskevicius, V., Cantú, S. M., & Vugt, M. V. (2012). The evolutionary bases for sustainable behavior: Implications for marketing, policy, and social entrepreneurship. Journal of Public Policy &
Marketing, 31(1), 115-128.
Haidt, J. (2001). The emotional dog and its rational tail: a social intuitionist approach to moral judgment. Psychological review, 108(4), 814.
Hainmueller, J., Hiscox, M. J., & Sequeira, S. (2015). Consumer demand for fair trade: Evidence from a multistore field experiment. Review of Economics and Statistics, 97(2), 242-256.
Hinds, J., & Sparks, P. (2008). Engaging with the natural environment: The role of affective connection and identity. Journal of environmental psychology, 28(2), 109-120.
