Promoting Sustainable Behaviour: Using a Lifestyle Quickscan With Impact Score and Positive or Negative Feedback

(1)

Promoting Sustainable Behaviour: Using a Lifestyle

Quickscan With Impact Score and Positive or

Negative Feedback

Lizz Splinter (s2053950) Distelvink 80

7827 DD Emmen

e.d.splinter@student.rug.nl

Master Communication and Information sciences University of Groningen

First reviewer: Prof. dr. John Hoeks Second reviewer: Gregory Mills, PhD. 04-10-2020

(2)

Preface

First of all, I would like to thanks professor Hoeks from the bottom of my heart for his guidance, support and feedback in finishing this thesis (even with the challenges of the coronacrisis). Without having u as my supervisor I would not be able to complete this chapter of my educational journey. This whole adventure ignited a desire to – even after finishing my masters – continue to researching means of promoting sustainable behaviour. (This weekend I had my first vegetarian ‘bitterbal’ from the Lidl and it was delicious!) Please continue the quest of finding persuasive interventions in terms of promoting sustainable behaviour with the same drive and passion u had while supervising me. Without exaggeration, our world depends on it.

Secondly, I would like to thank dr. Vuijk for not giving up on me and handing me over when he could not be my supervisor anymore. This goes for all wonderful people who work at RUG and I have encountered during my studies. Thank you for your guidance!

Lastly, I would like to thank my family and partner for supporting me through many breakdowns and academical monologues. Without you guys, I couldn’t have done it. I love you strongly.

To all who are reading this preface voluntarily, I would like to thank you too. This preface was written (and this thesis was finished) on World Animal Day. Without a doubt, I believe it was meant this way. No matter what your take on sustainability is, the negative consequences of our behaviour on animal welfare, climate and child labour are apparent. I do not ask you to change your behaviour, but to educate yourself on how you can contribute to a better world (which will have the same result and is a silly example of the power of persuasive communication - for which I apologize). We can do it, but only together.

“Non scholae, sed vitae.”

[Learning] not for school, but for life.

(4)

Abstract

In this study, the effectiveness of a Lifestyle Quickscan with impact score and corresponding framed positive or negative personalized textual feedback on promoting sustainable behaviour was tested in a pretest/posttest design. 162 Dutch participants were asked (in an English questionnaire) about their behavioural intention and attitude in three domains concerning sustainability (animal welfare, climate and child labour). Following, they were asked to self-report on their current sustainable behaviour in these domains by filling in a ‘Lifestyle Quickscan’. The results of this Lifestyle Quickscan were presented with an ‘impact’ score. Based on that impact score, participants were assigned to either positive or negative feedback. The control group did not receive any feedback. After the intervention participants were again asked about their behavioural intention and attitude towards sustainable behaviour in the three domains (animal welfare, climate and child labour) and emotions they were currently experiencing. In addition, participants were asked about three opinion leader traits for explorative analysis. Participants that scored above the 50 percentile on all three traits were identified as possible opinion leaders. There was no significant difference found in the effect of the Lifestyle Quickscan with impact score and framed feedback between the feedback conditions. Participants in the negative feedback showed a slight increase on attitude on all three domains. The positive feedback condition only showed a positive increase on animal welfare. The control group only showed a positive increase on climate. Participants in the two feedback conditions and the control group showed a decrease on behavioural intention on all three domains. An interesting remark is that participants that received positive feedback seem to experience the most guilt and sadness in comparison with participants that received negative or no feedback. The identified possible opinion leaders had a lower impact score on all three domains in comparison with other participants. Following, the identified possible opinion leaders showed an increase on behavioural intention on all three domains. These results suggest that using a Lifestyle Quickscan with impact score and positive or negative feedback is not effective regarding promoting sustainable behaviour. In addition, the explorative results suggest that opinion leaders and other participants are differentially sensitive to persuasive strategies regarding promoting sustainable behaviour. It seems to be worthwhile to further research underlying mechanisms of forming attitudes and behavioural intention towards sustainable behaviour in a persuasive intervention and the role of opinion leaders in promoting sustainable behaviour.

(5)

1. Introduction

There is overwhelming evidence that global warming is resulting in profound consequences for ecosystems and people (IPCC, 2019). According to IPCC (2019), it is extremely likely that human

influence is the dominant cause of the observed warming since the mid-20th century. Although

multiple global parties, like the UNF CCC, have agreed that deep cuts in emissions are required and global warming should be limited to below 2 °C in the Paris Agreement, the earth’s average surface temperature is still increasing (IPCC, 2018).

1.1. Sustainable behaviour

To combat global warming and reduce emissions it is important that consumers adapt their lifestyle into more sustainable life choices. Concerning, for example, the decrease of meat and dairy consumption, energy use and travelling. Although the awareness and concern about global warming among consumers are vastly increasing, sustainable behaviour is not (I&O research, 2019). Most consumers do have the desire to behave more ethically, but this desire is often not reflected in their actual behaviour (Zaal, Mills, Hagen, Huisman, Hoeks, 2017, p. 3627). To change this consumer behavioural intention, into corresponding with current beliefs, it is important to change people’s cognitions (Zaal et al., 2017). One way to do so, is via persuasive intervention in the form of personalized feedback.

1.2. The effect of framed personalized feedback

The research of Zaal (2018) investigated the effects of positively versus negatively framed personalized feedback on willingness to adapting behaviour towards becoming more sustainable. The participants were asked to self-report their current sustainable behaviour by filling in a ‘Lifestyle Quick Scan’. The results of this Lifestyle Quick Scan were presented in an ‘impact’ score, implying the impact of their current behaviour on three domains: animal welfare, climate change and child labour. In two feedback conditions, the participants received personalized either positively or negatively framed textual feedback based on that impact score. Participants in the third condition, the control group, did not receive any feedback. The results of the research of Zaal (2018) showed no difference between either positively or negatively framed feedback on behavioural intention. In some cases, the effect of positively framed feedback was stronger than the effect of receiving no feedback. In other cases, the absence of feedback seemed to have a more

(6)

persuasive effect than including feedback, either positively or negatively framed. The effect of negatively framed feedback never exceeded the effects of positively framed or the absence of feedback (Zaal, 2018, p.6). These results are quite remarkable. According to Hoeken, Hustinx & Hornikx (2012, p.131) if there is any difference between positive or negative frames, positive framing (feedback) should have more effect. What is quite unexpected is that negative framed feedback seem to work just as well as positive feedback, when based on literature one might expect negative framing not to work. An important observation on the study of Zaal (2018) is that it measured behavioural intention after the intervention with positively or negatively framed textual feedback. Whether participants were willing to change their behaviour towards becoming more sustainable before the experiment remains unanswered. Measuring behavioural intention before the intervention and after the intervention might have led to other results.

1.3. Behavioural intention and opinion leaders

Another possible way to change consumers behavioural intention towards sustainable behaviour is via opinion leaders. Keys, Thomsen and Smith (2010) studied the role of informal leadership and its utility in influencing societal attitudes and practice. They based their study on theories of diffusion in which learning about new ideas, practices or technologies occurs through interpersonal communication with informal opinion leaders. According to Boster et al. (2015) opinion leaders play an important role to achieve a so-called ‘tipping point’ in a societal context. Once opinion leaders form a positive attitude towards a type of behaviour, they are likely to use their knowledge (mavenness), connections and persuader skills to influence other people in their social network. When the circumstances are correct, this may cause a ripple effect, resulting in a ‘tipping point’ on a regional level. Keys et al. (2010) and Boster et al. (2015) cite the role of opinion leaders as an important component of social change. They describe turning to opinion leaders to identify meaningful innovations to sustainability problems as a departure from studies of pre-defined innovations.

1.4. This present study

The purpose of this present study is to further research the effect of the Lifestyle Quickscan with impact score and corresponding personalized positively or negatively framed textual feedback on attitude and intentional behaviour towards becoming more sustainable. The present study uses a

(7)

pretest/posttest design to exclude the possibility that the attitude and intention to change behaviour were already present before the experiment. The goal is understanding the effect of the Lifestyle Quickscan with impact score and corresponding framed textual feedback on attitude and behavioural intention by taking already present attitudes and behavioural intentions (before the intervention) into consideration.

This study also aims to identify mechanisms and strategies involved in individual and societal context by looking into the role of opinion leaders to complex global sustainability issues such as animal welfare, climate change and child labour. Researching the effect of the Lifestyle Quikscan with impact score and personalized positively or negatively framed textual feedback on opinion leaders and other participants might lead to interesting results. Bases on the elaborations above, the research questions regarding this present study are made explicit:

RQ1: “To what extent is filling in the Lifestyle Quick Scan effective on the attitude and behavioural intention of individuals towards sustainable behaviour?”

RQ2: “Is there a difference in the effect of either positively or negative framed textual feedback on attitude and behavioural intention towards sustainable behaviour after confronting individuals with their current (un)sustainable behaviour?”

RQ3: “Is there a difference in the effect of either positively or negative framed textual feedback on attitude and behavioural intention towards sustainable behaviour after confronting opinion leaders with their current (un)sustainable behaviour?”

1.5. Relevance

Replicating and adjusting the approach towards researching the effect of the Lifestyle Quickscan on individual attitude and behavioural intention towards becoming more sustainable via personalized framed textual feedback might contribute to the further understanding the effect of the Lifestyle Quickscan with impact score and framing, personalized feedback as a persuasive intervention. In addition, looking into the effect of the Lifestyle Quickscan with impact score and personalized framed feedback on the opinion leader’s behavioural intention towards becoming

(8)

more sustainable might lead to interesting insights. At the practical level, future persuasive interventions planned by governments, institutions and other persuaders with the aim for sustainable behaviour might consider looking into the approach of instigating opinion leaders and informal social networks. Hopefully, this research will be a starting point for further research about achieving this so-called tipping point in people’s willingness to adapt their lifestyle to making sustainable life choices.

(9)

2. Theoretical framework

As described in the previous chapter, the need for more research into persuasive interventions aiming towards positively affecting the behavioural intention of individuals towards behaving more sustainably is essential. Before the underlying mechanism of persuasive intervention in the form of personalized positively or negatively framed textual feedback is described, it is important to create a deeper understanding of the complexity of promoting sustainable behaviour and targeting behaviour change in general. Therefore, in this chapter, the complexity of promoting sustainable behaviour is described. More specifically, the determinants underlying behavioural intention and the linkages between individual behavioural intention and societal context. In addition, the content and form of the intervention design, the personalized positively or negatively framed textual feedback, will be outlined. Lastly, the role of informal opinion leaders and their influence on societal attitudes and practice will be explained elaborately.

2.1. Promoting sustainable behaviour

Climate change, animal welfare and child labour are three domains within a broad range of ethical issues. Because these domains are complex issues with responses and projected impacts affecting the broadest range of social, environmental, economic, and personal issues, it can be argued that every individual is a stakeholder (Keys et al., 2010, p.193). Research related to the social response to ethical issues has focused on the need to reform at the public policy and institutional levels following commitments by governments to develop measures to adapt to more sustainable life choices under the UN Framework Convention of Climate Change (McCarthy, 2001, Smit and Wandel, 2006, Klein et al., 2007, cited in Keys et al., 2010). On the other end, a large body of research has focused on promoting sustainable behaviour on an individual level. Studied persuasive interventions seem to have promising positive effects on behavioural intention towards behaving more sustainably. The persuasive intervention in the form of framed positively and negatively framed feedback is among the most researched and expected promising persuasive effects (Michie et al., 2013, cited in Zaal, 2018). However, the discrepancy between theory and practice lies between the behavioural intention and actual behaviour. As described in the research of Zaal (2018) most individuals have the behavioural intention towards more sustainable behaviour, but this intention is often not reflected in their actual behaviour. In other words, individuals know what is ethical behaviour, but simply not do so because of intervening factors.

(10)

The complexity of promoting sustainable behaviour lies in understanding the underlying mechanisms that create this discrepancy between behavioural intentions and actual behaviour. Therefore, a deeper understanding in forming behaviours in general is necessary.

One of the most fundamental models of behavioural prediction is the model of Fishbein & Ajzen (2010): The Integrative Model of Behavioural Prediction. While other determinants play an important role in predicting behaviour (like the perceived norm and self-efficacy), the determinant

attitude plays a central role in the integrative model of behavioural prediction. The determinants

underlying actual behaviour are described as a summary of beliefs (Hoeken et al., 2012, p.43). For example, the main determinant attitude is seen as an evaluative cumulation of the presumed consequences of behaviour. Underlying determinants are behavioural beliefs (how likely is this behaviour to cause consequences?) and evaluations of these consequences (how desirable are these consequences?). To understand these beliefs and the mismatch between these beliefs and actual behaviour it is important to describe the relationship between attitude and behaviour.

Image 1: The Integrative Model of Behavioural Prediction. Fishbein & Ajzen (2010).

2.2. Attitudes and behaviour

As the observation of Zaal (2018), individuals often believe behaving ethically is the right thing to do, but this belief often does not reflect actual behaviour. This mismatch between beliefs, behavioural intention and actual behaviour is described as ‘cognitive dissonance’: “An aversive

(11)

state which motivates cognitive or behavioural actions to lower itself” (Dijkstra, 2009, p.792, cited in Zaal, 2018). A large body of research has focused on the dissonance that occurs as a result of a discrepancy between attitudes individuals have and their actual behaviour (Hoeken et al., 2012, p.55). In order to solve this cognitive dissonance, two pathways are described. Individuals could change their behaviour in line with their current belief (‘behaving more sustainably helps combat ethical issues’) or individuals could change their attitude in relation to their behaviour (thinking more positively about the relationship between behaving sustainably and ethical issues). The latter emphasizes the importance of changing an individual’s belief about a given behaviour. In order to change an individual's belief about a given behaviour, one has to address the most focal cognitions an individual has about a particular behaviour (Fishbein & Ajzen, 2010).

On the other end, there are individuals that actually behave in line with their attitude and beliefs. Even when attitude and beliefs are in line, current behaviour can still form a foundation for an attitude (Hoeken et al., 2012, p.55). This theory is based on the research of Bem (1967, 1972) which resulted in the self-perception-theory. This theory describes the way other individuals form their attitudes, often forms a base on which individuals form their own attitude. In other words, when an individual sees another individual have a positive attitude towards behaving sustainably, they are more open to that behaviour. Based on that other individual’s behaviour a new attitude is formed. As observed by Keys et al., (2010) this elaboration raises the question if the restricted focus on direct behavioural changes of individuals via persuasive intervention, while excluding the influence of the behaviour of other individuals, is the right focus of study.

Either an individual’s focal cognitions (or beliefs/attitude) needs to be changed via a persuasive intervention or another individual's behaviour forms the leading example into changing behaviour towards more sustainable behaviour in their social network. Before the role of other individuals or so-called ‘informal opinion leaders’ in societal context is described, it is important to describe the possible persuasive intervention to change an individual’s focal cognition, which is described in the paragraph below.

(12)

2.3. Framing textual feedback: a persuasive intervention

In order to change an individual's belief about a given behaviour, one has to address the most focal cognitions an individual has about this behaviour (Fishbein & Ajzen, 2010). But how does one address the most focal cognitions of an individual? One way to do so is via persuasive intervention. In order to describe ‘persuasive intervention’, it is important to outline a definition of ‘persuasion’ in general. Giving a definition of persuasion is challenging, therefore it is difficult to distinguish the difference between persuasion and manipulation (Hoeken et al. 2012, p. 13). O’Keefe (2002a, p.5) attempts to describe persuasion as follows:

“A successful, intentional effort at influencing another's mental state through communication in a circumstance in which the persuadee has some measure of freedom.”

According to Hoeken et al. (2012, p.13) the described ‘mental state’ is often equated with attitude. The most cited definition of an attitude is from Eagly and Chaiken (1993, p.1, cited in Hoeken et al., 2012, p.14):

“(...) a psychological tendency that is expressed by evaluating a particular entity with some degree of favour or disfavour.”

Although the definition of O’Keefe describes a mental state (or attitude), most persuasive interventions aim at behavioural change. As observed by Hoeken et al. (2012), persuasion takes place via communication: the transfer of information. Information transfer influences an individual’s focal cognitions before it affects their actual behaviour. Moreover, communication leads to the formation or change of an attitude, which in return possibly influences behaviour. Therefore, communication cannot influence behaviour directly and is often not used as a term in describing persuasion. Given the definition of persuasion elaborated above, a definition of persuasive intervention can be made:

“Persuasive interventions (or messages) are designed with the goal to change an individual’s attitude via information transfer, in which the persuadee has some measure of freedom (Hoeken et al., 2012, p.14).”

(13)

As described before, persuasive intervention in the form of personalized framed positively and negatively framed feedback is among the most researched and expected promising persuasive effects (Michie et al., 2013., cited in Zaal, 2018). The two most important aspects of this persuasive intervention are ‘personalized feedback’ and ‘positively or negatively framed’. Zaal (2018, p.20) describes two types of mechanisms that are relevant in relation to effect of personalized feedback: (1) facilitating message processing and (2) increasing a communication’s impact by altering the receiver’s behavioural determinants (i.e. belief structures) towards some desired outcome (e.g. Noar et al., 2007, Hawkins et al., 2008, Skinner et al., 1999, in Zaal, 2018). The mechanism to facilitating message processing can be described via one of the most cited models in persuasive communication: The Elaboration Likelihood Model of Persuasion of Petty and Cacioppo (1981). This persuasion model is a dual process theory that describes the forming and changing of attitudes and behaviour. In short, the Elaboration Likelihood Model proposes two pathways to persuasion via message processing: the central route and the peripheral route. The central route is used when the persuadee is motivated to think (elaborate) about the message and has the ability to do so. When the persuadee is involved with the matter of subject and processes the message with the least distraction, it is likely the persuadee will elaborate on the message. When the persuadee thinks or rehearses, favourable thoughts about the message, lasting persuasion is likely. The persuadee will possibly move away from the advocated position if the persuadee rehearses unfavourable thoughts about the message. When the persuadee is unable or not motivated to process the message the persuadee will resort to peripheral cues. Message processing via peripheral cues is associated with a temporary attitude change, with a chance of elaboration in the future. When the peripheral cue is not accepted or present, the persuadee will return to the attitude initially held. The facilitation of message processing in relation to the effect of personalized feedback lies in the tailoring or personalizing of the message. Because the message directly applies to an individual’s current behaviour, attention and elaboration are likely established. The impact of the message (or communication) on an individual's belief structures towards the desired outcome is increased because of the increased elaboration. In other words, the message is relevant to an individual because of the personal focus of the message on the unique individual. An important factor that might play a role in the effect of personalized framed textual feedback is the effect of emotions that are associated with this persuasive intervention. Emotions are often pointed out as the main influence for experiential, or affective attitudes, which in turn can influence an individual’s

(14)

behaviour via behavioural intention and attitude (Zaal, 2018, p.22). After being exposed to a persuasive intervention, positive and negative affective reactions may be aroused (O’Keefe, 2012, Ajzen & Fishbein, 2010, in Zaal, 2018).

In addition, framing the personalized feedback adds another persuasive effect. The framing effect is among the most studied persuasive effects in literature, especially in health communication (Hoeken et al., 2012, p.129). As described in the Model of Integrative Behavioural Prediction of Fishbein & Ajzen (2010), the main determinant ‘attitude’ is seen as an evaluative cumulation of the presumed consequences of behaviour. Underlying determinants are behavioural beliefs (how likely is this behaviour to cause consequences?) and evaluations of these consequences (how desirable are these consequences?). Framing persuasive communication affects these behavioural beliefs and the evaluation of consequences as underlying determinants of an individual’s attitude. Framing is used as a description of the desirability or undesirability of consequences for an individual’s behaviour (Hoeken et al. 2012, p. 128). Although meta-analyses of the effect of positively versus negatively framed feedback remain inconclusive, positive framed messages on propagated behaviour seem to have more effect (Salovey et al., 2002 in Hoeken et al., 2012). On the other end, negatively framed messages seem to have more effect when an individual is not thoroughly aware of the possible consequences. However, when an individual is indeed aware of the possible consequences, the advice against certain behaviour via negative framing is counterproductive (Earl & Albarracin, 2007, in Hoeken et al., 2012).

Given the elaboration above, the results of the research of Zaal (2018) into the effect of either positively or negatively framed personalized feedback on behavioural intention towards becoming more sustainable are remarkable. The results of the research of Zaal (2018) showed no difference between the effect of either positively or negatively framed feedback. In general, including personalized framed feedback did not seem to affect attitude and behavioural intention much.The results suggest that only in some cases positive feedback was stronger than no feedback. However, the absence of feedback also seemed to work more persuasive than including feedback in some cases (Zaal, 2018, p.6). The fact that the results of the research of Zaal (2018) showed no difference between either positively or negatively framed personalized feedback is interesting. Especially

(15)

when taken into consideration that the usage of negatively framed personalized feedback often seemed to be viewed as counterproductive (Earl & Albarracin, 2007, in Hoeken et al., 2012).

2.4. The role of opinion leaders

The framing effect is often paired with the use of an exemplar in persuasive messages (Zillman, 2006). An exemplar is a vivid, narrative form, in which one or more individuals are staged to illustrate the desirable or undesirable consequences of certain behaviour. The frequent use of exemplars in persuasive messages once more hints towards the impact of other individual's behaviour on an individual’s (attitude towards) a certain behaviour. As described before, this raises the question whether societal context should be taken into consideration when researching the effect of persuasive intervention (like personalized positively or negatively framed textual feedback) on behavioural intention in promoting sustainable behaviour. Keys et al. (2009) substantiate this theory by concluding that relationships with social groups and networks are significantly influencing sustainable decision-making on an individual level.

Like the study of Zaal (2018), research related to behavioural intention towards becoming more sustainable has predominantly focused on individual behaviour change. At the other extreme, research has focused on the need to reform at the public policy and institutional level (Keys et al., 2010, p.187). Keys et al. (2010, p.188) conclude that all levels of response are at least indirectly connected and attempts to understand the capacity for change in different contexts is crucial for responding to complex problems such as climate change. This raises the question whether behavioural intention should be studied on an individual level, leaving out societal context. As observed by Zaal (2018), “Many persuasive interventions employ so-called exemplars, which often include stories about other individuals who already successfully carry out some targeted behaviour” (Zaal, 2018, p.11). Keys et al. (2010) argue that “a methodological weakness in many attitude-behaviour as reviewed by Kollmuss and Agyeman (cited in Keys et al., 2010, p.191) is the restricted focus on direct behaviour changes and exclusion of educational or political actions, thereby focusing the burden of environmental responsibility on the individual and the household”. Therefore, Keys et al. (2010) conclude that relationships with social groups and networks are significantly influencing sustainable decision-making on an individual level. In addition, it can be

(16)

argued that ethical issues like climate change, animal welfare and child labour are societal issues, outreach individual actions.

In a large body of research, opinion leaders are often identified as focal points for the communication of innovative ideas and practices in social networks relevant to ethical issues. Opinion leaders aid in developing knowledge and motivation for change in their social networks. Moreover, individual decisions-making in relation to responding to ethical issues is influenced significantly by relationships with social groups or networks. (Keys et al., 2010, p.188-192). In addition, the self-perception-theory of Bem (1967, 1972) substantiates this elaboration by describing that the way other individuals (in an individual’s societal context) form their attitudes, often forms a base on which individuals form their own attitude. In other words, when an individual sees another individual have a positive attitude towards behaving sustainably, they become more open to changing attitudes and behavioural intention towards that behaviour themselves. The importance of the role of other individuals attitudes and behavioural intention towards certain behaviour is substantiated by the underlying determinants ‘normative beliefs’ or so-called motivation to comply and ‘perceived (social) norm’ of behavioural intention in the Integrated Model for Behavioural Prediction of Fishbein & Ajzen (2010). According to Fishbein & Ajzen (2010) these determinants play an important role next to attitude and self-efficacy (how capable am I in performing this propagated behaviour) in forming a behavioural intention. Mackie, et al., (2015) describe a social norm as a social expectation which is constructed by an individual’s beliefs about what others do, and by an individual’s beliefs about what others think one should do. In other words, the expectation and beliefs of an individual of what is appropriate behaviour in group context plays an important role in forming individual behavioural intention towards certain behaviour.

According to Boster et al. (2015), opinion leaders play an important role in interpersonal influence persuasive campaigns. An opinion leader is described as someone with a strong social status and influence on (the beliefs of) people around them. Therefore, opinion leaders may play an important role in the (intention of) behaviour of people who follow or look up to them. Take for example an opinion leader who has a very positive attitude towards certain behaviour, like eating vegan. As a

(17)

rule of thumb, individuals in the social group of the opinion leader might consider eating vegan as well.

In conclusion, identifying and igniting opinion leaders towards behaving more sustainably, might make a pathway to achieving the ‘tipping point’ in relation to sustainable behaviour in the societal context. But how does one change the attitude and behavioural intention of an opinion leader towards behaving more sustainably? Assuming that opinion leaders are individuals as well, studying the effect of the Lifestyle Quickscan with impact score and positively or negatively framed personalized textual feedback on attitude and behavioural intention of opinion leaders might lead to interesting results. Therefore, this present study focuses on replication and adjustment towards researching the effect of the Lifestyle Quickscan with impact score and personalized framed textual feedback on individuals and (in addition) opinion leaders, of which the methodology is described in the following chapter.

(18)

3. Method

3.1. Participants

The participants for this research were recruited through snowball sampling (Babbie & Baxter, 2003: 135). This means that the first selection of participants was done by the researcher and from there the participants invite other people to participate in the study by sharing the link to the questionnaire (online). The website Qualtrics automatically ensures an ad random distribution of the stimulus material among the participants. Ethical approval was given by the Research Ethics Commission Committee (CETO) of the faculty of Arts, University of Groningen.

Due to limited time and resources 162 participants filled out the questionnaire. Of the 162 responses, 96 participants did not complete the questionnaire. This means that 66 participants were used for this study. Table 1 below shows the demographics and distribution of the participants in this study. One participant specified ‘highest earned degree’ as other by filling in ‘conservatory’.

Table 1: Relative distribution of gender and highest earned degree. Means and standard deviation of age throughout the feedback conditions (positive vs negative) and the control group.

Positive Feedback

Negative Feedback Control

Gender

Male 13 (56%) 6 (22%) 11 (39%)

Female 10 (44%) 21 (78%) 17 (61%)

Total 23 (100%) 27 (100%) 28 (100%)

M(SD)Age 36.0 (16.53) 40.19 (17.02) 46.36 (19.81)

Highest earned degree

Secondary school 3 (13%) 3 (11%) 2 (7%)

Intermediate vocational

training (MBO) 7 (31%) 6 (22%) 7 (25%)

College education (HBO) 9 (40%) 12 (45%) 10 (36%)

(19)

Master (University) 2 (8%) 2 (7%) 6 (21%)

Other 0 1 (4%) 0

Total 25 (100%) 21 (100%) 19 (100%)

3.2. Instrument 3.2.1. The system

Zaal’s (2018) system served as a base for the computer-based feedback intervention used in this study. This feedback system provides individually tailored feedback and was built using Qualtrics, a software platform for making online questionnaires. The system of Zaal (2018) served as a measurement instrument recording participants’ behavioural impact and willingness to change behaviour. The questionnaire of Zaal (2018) was adapted by adding (1) the measurement of willingness to change behaviour before the intervention and (2) the measurement of the three personality traits that can be used to identify opinion leaders. In addition, Although the participants were Dutch, the questionnaire’s main language was English. The specific components of the questionnaire are shown in table 2 below.

Table 2: Components of the Lifestyle Quick Scan.

MC = Multiple Choice, T/F = True/False questions, LS = Likert Scale (1 till 7, ranging from fully disagree to fully agree).

Component Type

1. Welcome/Introduction Text

2. Demographics Questions (MC)

3. Behaviour change Questions (MC)

4. Connector items Statements (LS) 5. Persuader items Statements (LS)

(20)

6. Environmental maven items Statements (LS)

7. Meat and dairy consumption Questions (MC + T/F)

8. Travelling Questions (MC)

9. Clothes/accessories Questions (T/F)

10. Electronic devices Questions (MC + T/F)

11. Feedback Personalized score + textual framed feedback

12. Emotions Statements (LS)

13. Social norm Statements (LS)

14. Behaviour change Questions (MC)

15. End of research Text

3.2.3. Measurement components

Most of the measurement components below were derived from Zaal’s (2018) questionnaire. The components “connector items, persuader items and environmental maven items” were added for this study.

2. Demographics

The questionnaire starts with questions about the participants' demographics, including gender (“What gender do you identify with?”), age (“What is your age?”) and highest earned degree (“What is your highest earned degree?”) following by multiple-choice or text entry when “other” is specified.

(21)

3. Behaviour change

The component behaviour change consists of two elements: the variable attitude and the variable

behavioural intention. The variable attitude consists of a pre-test/post-test measurement of the

dependent variable ‘attitude towards behaviour into becoming more sustainable’ on the three sustainable domains: animal welfare, climate and child labour. This component is being questioned on all participants before and after the intervention. The following items below were questioned rating to (1) I fully disagree to (5) I fully agree.

Table 3: Components on the attitude towards sustainable behaviour. LS = Likert Scale (1 till 5, ranging from fully disagree to fully agree).

Component

1. I think animal welfare in the livestock sector is a large and urgent problem

2. I think global warming is a large and urgent problem

3. I think child labour and inhumane working conditions are large and urgent problems

4. Animal suffering in livestock is something that should be terminated as soon as possible

5. Global warming is something that should be stopped as soon as possible

6. Child labour and inhumane working conditions should be terminated as soon as possible

The variable behavioural intention consists of a pre-test/post-test measurement of the dependent variable ‘willingness to change behaviour into becoming more sustainable’ on three sustainable domains: animal welfare, climate and child labour. This component is being questioned on all participants before and after the intervention. Instead of using a Likert scale the number of behaviours that participants plan to adapt for each condition was measured. The following items below are an example of the statements that were questioned via multiple choice (multiple answers possible).

(22)

Table 4: Components of behavioural intention. MC = Multiple choice

Component: Which behaviour(s) are you willing to change? (multiple answers possible)

1. Humans and Environment: My buying behaviour concerning clothes and/or electronic devices

2. Animal welfare and Environment: My eating behaviour concerning meat and fish 3. Animal welfare and Environment: My eating behaviour concerning dairy and eggs

4. Environment: My travel behaviour

5. None

Personal traits of opinion leaders

The personal traits of opinion leaders were measured via three scales: connector items, persuader items and maven items. Per item, 5 statements on a 7-point Likert scale were questioned. Participants that score above the 75 percentile on all three scales were identified as opinion leaders. The following items below were used to measure the personal traits of opinion leaders.

Table 5: Items of personal traits of opinion leaders

LS = Likert Scale (1 till 7, ranging from not at all to very much). Connector items

The following questions are about you and your behaviour in social context.

1. I am often the link between friends in different groups 2. I often find myself introducing people to each other

(23)

3. I try to bring people I know together when I think they would find each other interesting

4. I frequently find that I am the connection between people who would not otherwise know one another

5. The people I know often know each other because of me

Persuader items

6. I am good at thinking of multiple ways to explain my position on an issue

7. When in discussion, I’m able to make others see my side of the issue 8. I am able to adapt my method of argument in order to persuade someone

9. I can effortlessly offer multiple perspectives on an issue that all support my position

10. More often than not, I am able to convince others of my position during an argument

Environmental Maven items

11. When I know something about environmental issues, I feel it is important to share that information with others

12. I like to be aware of the most up-to-date environmental information so I can help others by sharing when it is relevant

13. If someone asked me about an environmental issue that I was unsure of, I would know how to help them find the answer

14. Being knowledgeable enough about environmental issues so that I could teach someone else is important to me

(24)

15. People often seek me out for answers when they have questions about an environmental issue

4. Connector items

The connector items consisted of 5 statements on a Likert scale ranging from (1) fully disagree to (7) fully agree. Measuring the personal traits of ‘connectors’, concerning behaviour in social contexts. Per example: “I am often the link between friends in different groups”.

5. Persuader items

The persuader items consisted of 5 statements on a Likert scale ranging from (1) fully disagree to (7) fully agree. Measuring the personal traits of ‘persuaders’, concerning persuasive behaviour in social contexts. Per example: “ I am good at thinking of multiple ways to explain my position on an issue”.

6. Environmental (sustainability) maven items

The maven items consists of 5 statements on a Likert scale ranging from (1) fully disagree to (7) fully agree. Measuring the personal traits of ‘mavens’ in the environmental domain, concerning the use of certain knowledge in social contexts. These items were modified to statements concerning the environmental domain, whereas the original scale used statements concerning health issues. Per example, the original statement “When I know something about a healthy lifestyle, I feel it is important to share that information with others” was modified to “When I know something about environmental issues, I feel it is important to share that information with others”.

Lifestyle Quickscan

The Lifestyle Quickscan is a set of questions about meat and dairy consumption, travelling, clothes and accessories, devices and energy use. Answers given to these questions are coded and counted into the impact score later on. The items are described below.

(25)

7. Meat and dairy consumption

This item questioned the frequency in which participants consume meat, dairy products and eggs. For example: “How many days a week do you eat meat (products)?”. When participants answered to eat meat products, they were asked if they only consumed these products if they were certified as organic, via a true or false statement. For example, “I only eat dairy (milk, yoghurt, cheese) or eggs, if it is certified as organic, otherwise I don’t eat it”.

8. Travelling

This item consists of two questions concerning travelling by plane. Participants were asked about the frequency (never, once, twice, three times or more often) and destination (within or outside Europe).

9. Clothes and accessories

This item consists of two true of false statements about the buying behaviour of participants concerning clothes and accessories. For example: “I only buy clothes and accessories if I am sure they are produced without exploitation, forced labour and labour, otherwise I don’t buy them.”

10. Devices

In this item participants were asked whether they purchased electronic devices, such as a mobile phone, laptop or tablet, within the past five years. If indicated buying one or more of such electronic devices, more specific follow-up questions were asked. For example: “In general, how often do you buy a new laptop?”. With answer options ranging from once a year, once every two years, once every three years and I use my laptop longer than three years. Following, participants were confronted with a true or false statement about the production of such device. For example: “I only buy a new laptop if I’m sure the materials (like gold and coltan) are produced without exploitation, forced labour and child labour, otherwise I don’t buy it.”

12. Experienced emotions

The variable emotions consist of 9 items on a Likert scale ranging from (1) not at all to (7) very much, measuring the current emotions of the participants after the intervention. The emotions asked were happy, mad, indignant, guilty, sad, worried, scared, threatened and confused. The

(26)

variable ‘happy’ was later on recoded in the analysis, therefore it was the only ‘positive’ emotion among more ‘negative’ emotions.

3.2.4. Intervention components

The intervention components are a replication of the intervention in Zaal’s research (2018, p. 38-39). No adjustments were made at this point.

Information

Small paragraphs that contain information about the consequences of (un)ethical behaviour, related to behavioural domains. These paragraphs contain textual information only, partially based on Carvalho et al., (2015, in Zaal, 2018) research into drivers of sustainable behaviour.

Impact score of the Lifestyle Quick Scan

After filling in the Lifestyle Quick Scan, participants were confronted with an impact score on each domain (animal welfare, climate and child labour). The higher the score, the higher one’s impact on animal welfare, climate change and child labour. For example, if a participant within the negative feedback condition scored below 50 points on the animal domain, they received corresponding feedback (“As you behave now, your negative impact on animal welfare is minimal”). When another participant within the negative feedback condition scored above 200, they received the following feedback: “As you behave now, your negative impact on animal welfare is very large.”

11. Personalized feedback

Table 6 below shows more examples of how positive and negative feedback based on the Lifestyle Quick Scan score was constructed in the research of Zaal (2018, p.38-39) for each domain.

Table 6: Examples of Individually Tailored Messages (Zaal, 2018, p.39).

(27)

Animal Welfare

You don't consume a lot of meat. Minimizing your meat-intake means minimizing your impact on animal cruelty, you are doing much better than most people!

---

To make things even better, you eat meat (products) that are organic.*

You don't consume a lot of meat, but with every piece of meat, you eat you are still contributing to animal cruelty.

---

To make things worse, you eat meat (products) that aren't organic.*

Human Rights When you buy clothes and accessories you’re sure that they are produced under proper working conditions. In this way, it is most probably the case that you contribute to working conditions adhere to human rights laws. That’s great!*

When you buy clothes and accessories you're not sure whether they are produced under proper working conditions. In this way, it is most probably the case that you contribute to working conditions that don't adhere to human rights laws.*

Environment As you behave now, you have a moderate negative impact on the environment. Your score is lower compared to most participants. This means that you are already minimizing your impact on the environment and that is very impressive!

As you behave now, you have a moderate negative impact on the environment.

---

You consume more non-renewable energy than is necessary*; You quite often travel by plane; you don't care if the material of your clothes or accessories are of ecological or biological origin.*

*Only presented by the system if actually being true or false as reported by the participants.

(28)

3.3. Procedure

The participants in this study entered the survey via Qualtrics, an online survey tool. The complete questionnaire consisted of all components as shown in table 1. The questionnaire was completed in an average time of 25 minutes online. All participants went through the items of the Lifestyle Quick Scan in the same chronological order as outlined in table 1. Participants were welcomed, information about the questionnaire with a brief description of its content and its estimated duration, and thanked in advance for participation. Following, participants were asked about demographic characteristics such as gender, age and highest earned degree.

After the introduction, participants were asked about their attitude and behavioural intention towards climate change, animal welfare and child labour. In addition, participants were asked about their personality traits (connector, persuader and environmental maven traits). After questioning attitude, behavioural intention and personality traits, participants were asked to indicate their current behaviour concerning meat and dairy consumption, travelling, clothes and accessories, (electronic) devices and energy use. After filling out questions concerning their current behaviour, participants in the positive and negative feedback condition were informed that the system calculated an impact score on three sustainable domains (animals, humans and environment):

“You just specified which choices you make concerning food, travel, clothing, electronics, energy use and giving to/working for non-profits. Often you don't realize that each time you make such a decision, you negatively influence the world. This survey calculates your (un)sustainability score on three dimensions - Animals, Humans and the Environment - and it shows you how large your negative impact on our world really is. As your score gets higher, your impact is more negative.”

After this notification participants were confronted with their impact score and corresponding personalized feedback. For example: “You scored 320 of the maximum score of 345 on the Animal Welfare dimension.” Followed by how this score reflects their current impact: “As you behave now, your negative impact on animal welfare is very large.” Lastly, the personalized framed textual feedback, for example in the negative feedback condition: “You consume meat every day. (...)

(29)

With every piece of meat you eat, you contribute to animal cruelty.” Participants in the control group did not receive the notification, the corresponding impact score and personalized framed textual feedback. After being presented with the notification, the impact score and personalized feedback, the participants were asked: “To what extent you feel one or more of the following emotions?”. Followed by 9 emotions and a 7-item Likert Scale, ranging from 1 (not at all) to 7 (very much). For explorative research, participants thereafter were asked 4 questions about their experience with a social norm. Also presented on a 7-item Likert Scale, ranging from 1 (not at all) to 7 (very much). Second-last, participants were once again asked about their attitude and behavioural intention towards sustainable behaviour. Lastly, participants were thanked for participating in this study. If they had any questions or remarks they could fill in a text field.

3.4. Design

In the research of Zaal (2018) a post-test only control group experimental design was employed. This study uses a pre-test/post-test control group experimental design, meaning there is a measurement of attitude and behavioural intention before and after the invention. Table 7 below shows a visual presentation of the design for this experiment. Please note, that for the explorative study (opinion leaders vs. participants) the same experimental design was used.

Table 7: Visual presentation pre-test/post-test control group experimental design Pre-test/Post-test

experimental design

Time ---→

Pre-test Intervention Post-test

Experimental group 1 (positive) R O1 X O2 O2-O1=de Experimental group 2 (negative) R O3 X O4 O4-O3=de Control group R O5 O6 O6-O5=de

(30)

4. Results

4.1. Effects of personalized positive and negative feedback

Related to the primary research questions, a repeated measures analysis of variance (ANOVA) was conducted with time of measurement (pretest vs posttest) as within-participants factor. Type of framed personalized textual feedback (positive vs negative) were included as between-participants factors. The control group did not receive any framed textual feedback. There were no significant interactions between the pretest and posttest and type of feedback on any of the dependent variables.

4.2. Attitude towards sustainable behaviour

To analyse the attitude towards sustainable behaviour a repeated measures (ANOVA) was conducted with pre/post as within-participants factor and the type of feedback (positive vs. negative) and domain (animal welfare, climate and child labour) as between-participants factors. No significant results were found in attitude towards sustainable behaviour with pre/post as within-participants factor and the type of feedback (positive vs. negative) and domain (animal welfare, climate and child labour) as between-participants factors. Table 8 below shows the results on the attitude towards sustainable behaviour. The means apply to a participant’s attitude towards statements made about sustainable behaviour per domain (1= I fully disagree, 5= I fully agree).

Table 8: Attitude towards sustainable behaviour pretest/posttest versus group and domain Pretest Posttest Mean Standard deviation Mean Standard Deviation Control Animal welfare 4.089 .199 4.000 .201 Climate 3.389 .202 4.054 .203

(31)

Child labour 4.286 .185 4.250 .165 Negative Animal welfare 4.045 .225 4.250 .227 Climate 4.068 .228 4.273 .229 Child labour 4.477 .208 4.773 .186 Positive Animal welfare 3.925 .236 4.100 .238 Climate 4.100 .239 4.050 .240 Child labour 4.350 .219 4.200 .195

4.3. Behavioural intention towards sustainable behaviour

In the pretest measurement, a total of 57 participants indicated to be willing to change their behaviour in one or more categories (animal welfare, climate change and child labour). 9 participants indicated they were not willing to change any behaviour. In the posttest measurement, a total of 55 participants indicated to be willing to change their behaviour in one or more categories. 11 participants indicated they were not willing to change any behaviour after the intervention. Table 9 below shows the behavioural intention of participants pretest/posttest per condition.

Table 9: Behavioural intention pretest/posttest per condition.

Pretest Posttest

Behavioural intention

Positive Negative Control Positive Negative Control

Clothes and devices

11 14 14 10 13 13

(32)

Diary and eggs 6 8 5 5 9 6

Travelling 7 7 9 4 4 8

Total 38 45 34 28 42 42

None 4 2 5 6 1 4

Table 10 shows the average number of behaviours that participants plan to adapt for each condition. The means for behavioural intention with time of measurement (pretest vs posttest) as within-participants factor are described with type of textual personalized framed feedback (positive vs negative) as between-participants factors.

Table 10: Means (and standard deviations) of behavioural intention pretest/posttest per condition

Pretest Posttest N Mean Standard Deviation Mean Standard Deviation Control 25 1.920 .222 1.840 .213 Negative 21 2.238 .243 2.095 .232 Positive 20 2.053 .255 1.789 .244 4.4. Experienced emotions

Table 11 below show the results of the emotions of participants. An explorative result worth mentioning is the effect of the personalized feedback on the emotions ‘guilt’ (p=.045; N=66) and ‘sadness’ (p=.108; N=66). Participants that received positive feedback seem to experience the most guilt in comparison with participants that received negative or no feedback. Table 11 below shows the means and standard deviation of the emotions of participants versus the feedback condition.

(33)

Table 11: Means (and standard deviation) of emotions of participants post-intervention Control (N=25) Negative (N=21) Positive (N=20) Mean Standard Deviation Mean Standard Deviation Mean Standard Deviation Happy 5.25 1.110 4.64 2.158 4.51 1.123 Mad 2.61 1.595 2.36 1.551 2.90 1.446 Indignant 3.04 1.401 2.68 1.701 3.29 1.736 Guilty 2.46 1.574 3.00 1.683 3.67 1.653 Sad 3.11 1.524 2.88 1.856 3.90 1.670 Worried 3.82 1.887 3.24 1.739 3.95 1.936 Scared 2.89 1.750 2.28 1.6969 2.95 1.717 Threatened 2.29 1.536 1.92 1.412 2.62 1.465 Confused 3.32 1.964 2.64 1.912 3.52 1.834

(34)

4.5. Explorative analysis

The explorative analysis firstly applied to the difference between the impact score of the Lifestyle Quickscan of opinion leaders versus other participants. Secondly, the attitude towards sustainable behaviour for opinion leaders with the time of measurement (pretest vs posttest) as within-subject factor and the domain (animal welfare, climate and child labour) as between-participants factor is described.

In order to identify the opinion leaders, three scales with personal traits were measured among participants (connector traits, persuader traits and traits of environmental mavens). Participants that scored above the 75 percentile on all three scales were identified as opinion leaders. The first sample size of identified opinion leaders was very small (N=5). Therefore, no ANOVA was conducted. Lowering the percentile to 50 resulted in 17 participants that could be identified as possible opinion leaders.

4.5.1. Impact score opinion leaders versus other participants

Table 12 below shows the difference between the scores of opinion leaders versus others per domain by presenting the calculated means and standard deviation.

Table 12: Means (and standard deviation) of the impact score of opinion leaders versus others per domain

Opinion leaders others

N Mean Standard Deviation N Mean Standard Deviation Animal welfare 17 194.117 90.298 61 239.590 88.021 Climate 17 142.941 79.580 61 175.491 65.508

(35)

Child labour

17 217.647 130.732 61 303.278 85.935

4.5.2. Attitude towards sustainable behaviour of opinion leaders versus others

Table 13 below shows the results on attitude towards sustainable behaviour for opinion leaders with the test (pre or post) as within-subject factor and the domain (animal welfare, climate and child labour) as between-participants factor.

Table 13: Attitude towards sustainable behaviour pre/post versus group and domain

Pre Post Mean Standard deviation Mean Standard deviation Control Animal welfare 4.600 .330 4.100 .343 Climate 3.950 .344 3.950 .347 Child labour 4.800 .306 4.400 .280 Negative Animal welfare 4.500 .603 4.500 .626 Climate 4.500 .627 4.500 .633 Child labour 4.500 .558 4.833 .512 Positive Animal welfare 3.750 .738 4.000 .767

(36)

Climate 4.250 .768 4.250 .776

Child labour 4.500 .684 4.500 .627

4.5.3. Behavioural intention of opinion leaders versus others

Table 14 below was based on the average number of behaviours that opinion leaders plan to adapt for each condition. The means of the behavioural intention towards sustainable behaviour with the time of measurement (pretest vs posttest) as within-subject factor and the group (opinion leaders versus others) as between-participants factor are described.

Table 14: Descriptives of the behavioural intention pretest/posttest towards sustainable behaviour of opinion leaders versus other participants

Pre Post N Mean Standard deviation N Mean Standard deviation Opinion leaders 17 2.056 .1088 17 2.400 .1055 Others 49 2.066 1.108 49 1.849 1.063

(37)

5. Discussion

In this study, the persuasive effects of the Lifestyle Quickscan with impact score and either positively or negatively framed personalized textual feedback on attitude and behavioural intention towards more sustainable behaviour was investigated. 162 Dutch participants filled in an English questionnaire, of which 66 participants completed the questionnaire and therefore were used for this study. A between-participants experiment was conducted in which attitude and behavioural intention on three sustainable domains (animal welfare, climate change and child labour) were measured before and after the intervention of receiving an impact score and corresponding (either positively or negatively) framed personalized textual feedback. After the intervention the participants were also asked about emotions they were currently experiencing. In addition, the participants were asked about three personal traits that can be used to identify opinion leaders. Participants that scored above the 50 percentile were marked as possible opinion leaders. Moreover, the effect of the Lifestyle Quickscan with impact score and corresponding feedback on the attitude and behavioural intention towards sustainable behaviour on the three domains of identified possible opinion leaders was researched.

First, results showed filling out the Lifestyle Quickscan and receiving an impact score has no significant effect on attitude or behavioural intention towards sustainable behaviour, in contrast to what Zaal (2018) hypothesized. Evaluating (and elaborating on) one’s behaviour via a Lifestyle Quickscan, receiving an impact score with corresponding framed personalized feedback seem to have no effect. The behavioural intention of participants to change towards sustainable behaviour was relatively high to begin with, and did not change between pretest and posttest. In addition, the results showed that either positively or negatively framed personalized textual feedback has no significant effect on behavioural intention or attitude towards sustainable behaviour, also like to the results of the study of Zaal (2018). According to the results of this present study, there is no difference between the effect of using either positively or negatively framed personalized feedback on an individual’s attitude or behavioural intention towards becoming more sustainable. This is providing support for the claim that it does not matter for persuasive intervention to either confront participants with positive or negative feedback (Hoeken et al., 2012, p.130). However, this is in contradiction with the expectation that when persuasive intervention in the form or framing is used, positive framing has more effect (Hoeken et al., 2012, p.131). This is also in contradiction with

(38)

the claim that when providing persuasive intervention in the form of framed textual feedback, negative framing might harm the intention to change towards sustainable behaviour. For further research, this raises the question if framing personalized textual feedback in the context of promoting sustainable behaviour is effective at all.

Given the measured emotions post experiment, it seems that participants are worried, sad and confused (more than average, M > 2.5) after filling in the questionnaire. Participants that received positive feedback seem to experience the most guilt and sadness in comparison with participants that received negative or no feedback. This cautiously supports the claim that personalized feedback might have an effect on emotions of participants. As described before, emotions are often pointed out as the main influence for experiential, or affective attitudes, which in turn can influence an individual’s behaviour via behavioural intention and attitude (Zaal, 2018, p.22). After being exposed to a persuasive intervention, positive and negative affective reactions may be aroused (O’Keefe, 2012, Fishbein & Ajzen, 2010, in Zaal, 2018). Looking at this present study, it might be a possibility that the subject of this study (sustainable behaviour and ethical issues) already aroused certain emotions among participants before filling in the questionnaire. Sustainability is discussed often on a political, educational and regional level. Measurements already taken have affected almost every household. Take for example, limiting the maximum speed on highways from 130 to 100 km/ph (between 6AM and 7PM) and not allowing gas stoves and boilers for new-built houses in order to reduce emissions. Therefore it is very likely participants have already elaborated and formed attitudes and behavioural intentions on the subject. Therefore, we cannot out rule the possibility that the measured emotions in the feedback conditions were aroused by other factors than receiving the impact score and corresponding feedback. For example, the participants in the positive feedback condition might have felt more sad and guilty due to the fact that their given answers are not in line with their actual behaviour. These emotions of guilt and sadness might have been caused by the surfacing and (experiencing) the discrepancy between certain beliefs and actual behaviour, so-called ‘cognitive dissonance’ (Dijkstra, 2009, p.792, cited in Zaal, 2018). On the other end, the participants might have felt more guilty and sad because they feel like they are not capable of solving ethical issues on their own. Another possibility for the participants in the positive feedback condition is that they do not agree with the aim of the study (like reported by a participant in the feedback field of the questionnaire). In addition, participants

(39)

who do not already make sustainable life choices, probably have certain beliefs why they should not do so. For example, there is a possibility that participants in the negative feedback condition might not see sustainable behaviour as necessary as a consequence of viewing the perceived threat (like melting ice caps as result of global warming) not as likely. This is substantiated by the fact that participants in the negative feedback condition felt less threatened than the participants in the positive feedback condition or control group. According to Hoeken et al. (2012) participants do not act at a perceived threat when the outcome of the threat is not perceived as likely, which is confirmed via the absence of effect on attitude and behavioural intention.

However, the results of this present study do partly support the claim of Zaal (2018) that personalized feedback leads to facilitating message processing and further elaboration of the matter. Participants were able to elaborate and review their current behaviour via the impact score, given the feedback of the participants on the questionnaire:

“When you look at my answers it seems like I agree to everything but do not want to change anything about my behaviour. I hardly ever buy meat, I only eat it at work because there at not many vegetarian options. I rent a small room in a house with roommates, so it is not up to me to decide which (vegetarian) options we use.”

“It is not that black or white as you suggest. Income by example influence also my choices..”

“I think this study is done with specific goals in mind, and I don’t agree with that. I don’t agree with the direction this study has taken.”

Although the results of this present study showed that receiving an impact score and either positively or negatively framed personalized textual feedback is not effective in promoting sustainable behaviour, there is a possibility that the framing effect of the personalized feedback had a methodological weakness. Looking at the notification of Zaal (2018) given to participants before receiving the impact score and corresponding feedback, an important notion has to be made. Zaal (2018) describes the impact of a current behaviour of a participant as “negative influence on the world”. According to Zaal (2018) the impact score: “(...) shows you how large your negative

Promoting Sustainable Behaviour: Using a Lifestyle Quickscan With Impact Score and Positive or Negative Feedback