Msc Communication Science Track: Persuasive Communication
University of Amsterdam – Faculty of Social and Behavioural Sciences Master thesis
Influencing pro-environmental behaviour beyond nudging
A quantitative research of behavioural interventions on attitude, immediate pro-environmental choices and long-term environmental choices
By Linde Mouw
12451606
Supervisor: Dr. M.L. Fransen
Statement of originality
This document is written by Linde Mouw, who declares to take full responsibility of 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 Social and Behavioural Sciences is responsible solely for the supervision of completion of the work, not for the contents.
Abstract
The compelling evidence of global warming demands climate change mitigation and adaption on a global and individual level. As the magnitude of climate change is deeply rooted in and under influence of human activity, behavioural sciences can provide insights to influence the underlying decision-making process by means of behavioural interventions. The present study examines the effectiveness of behavioural interventions (nudge vs. rational override) on influencing pro-environmental choices (eating vegetarian) on the short and long term. The results suggest that a Nudge intervention and a Rational Override intervention both lead to more pro-environmental choices on the short term, compared to no behavioural intervention. However, both interventions do not show effects on pro-environmental choices on the long term. A Rational Override does lead to a more positive attitude towards a vegetarian diet compared to a Nudge or no behavioural intervention, but these effects do not affect for pro-environmental choices on the long term. This study was one of the first to experimentally study the effectiveness of behavioural interventions on the long term, in attempts to clarify scientific evidence and to improve the practical application of behavioural interventions. However, future research is needed to refine the findings and to further examine the fruitful possibilities of behavioural interventions in the context of pro-environmental behaviour.
Keywords: rational override, nudge, pro-environmental choices, attitude towards a vegetarian diet.
Introduction
Climate change science shows that human activity is the dominant cause of observed global warming since the mid-20th century (IPCC, 2013). The magnitude of this biophysical effect is deeply rooted in and under influence of human behaviour. Especially, the routines of
everyday life and the social structure of modern societies appear prominent drivers of anthropogenic climate change (Dunlap & Brulle, 2015).
To prevent further environmental damages and mitigate the global warming process, behavioural change on both a global and individual level are urgent. Climate change
mitigation and adaptation go beyond interdisciplinary research across the natural and social sciences due to the centrality of human behaviour. Greater knowledge of human behaviour and societal dynamics are fundamental to face this complex, long-term challenge. Therefore, the role of behavioural sciences has been acknowledged for its usefulness in studying the antecedents and cognitive processes related to human behaviour (Urry, 2015). Previous research contributed to the understanding of underlying decision-making processes that determines specific behaviour and the possible usage to effectively altering human behaviours that contribute to environmental problems. Moreover, the insights of behavioural science caught attention of governmental instruments as it provides empirical-scientific evidence to substantiate policies and increase effectiveness (Oliver, 2015).
One way to explicitly apply behavioural insights into public policy domains is by means of nudging. Nudging draws on facts of human nature and psychology and finds its origin in behavioural economics. The term nudging describes any aspect of a design in which choices can be presented to alter people’s behaviour in a predictable way (Thaler & Sunstein, 2008). By altering the architecture of choice, the decision-making process is influenced and people are nudged towards a desired and unified outcome (Van Lieren, Calabretta, & Schoormans, 2018). For example, a fly sticker placed in a urinal functioning as focal point
when urinating to reduce spillage in the bathroom is considered a nudge. Just like making healthy food options more prominent in a cafeteria or supermarket (Hansen & Jespersen, 2013). People are gently pushed towards performing the desired behaviour while leaving them free to choose otherwise.
However, it is yet unclear how permanent the effects of nudging interventions are. Mixed evidence on the long-term effects of nudging have led to political and scientific
discussions on the sustainability of this behavioural intervention (Marchiori, Adriaanse, & De Ridder, 2017). Therefore, nudging has attracted interest from diverse disciplines within the academic field to shed light on the positive and the controversial effects of nudging. As nudges are influencing the automatic and subconscious part of the human brain, there is no conscious, cognitive engagement with performing the action. Moreover, people do not have any reflection on their own behaviour because an automatic nudge motivates a passive response. People in question do not easily recognize a nudge and are thus often unaware of the nudge and its influence on their own behaviour (Strassheim, 2016). For instance, when people are unconsciously nudged towards taking the stairs instead of the elevator by means of contextual cues, it is uncertain if this behaviour will be repeated in different environments at different times. This implies that a nudge is highly context-specific and will most likely not persist when removed from a particular context. These suggestions add up to the existing critics towards nudging and question its effectiveness on the long term. An alternative design to influence behaviour are rational overrides. Based on the definition of Van Lieren et al., a rational override refers to “a small moment of intentional friction that attempts to influence people’s behaviour or decision-making by intervening automatic thinking and activating
reflective conscious thinking” (Van Lieren et al. p.2171, 2018). Rational overrides are thus,
on the contrary of nudging, used to disrupt mindless automatic interactions and enhance moments of reflection. By stimulating the conscious decision-making process, the behavioural
outcome would be the result of some degree of deliberation (Strassheim, 2016). For example, energy bills allowing for social comparison of electricity consumption in the neighbourhood provokes reflective thinking with behavioural change as a result (Hansen & Jespersen, 2013). Rational override interventions have high potential to change behaviours also on the long term as they are more effective to change routines and habits in different environments (Van Lieren et al., 2018). Especially as Rational Overrides might alter attitudes towards the desired behaviour as well, which will contribute to the persistence of the effects on the long term. Therefore, the application of rational override interventions seems relevant for changes in society concerning the climate as it has been suggested that long-term changes towards pro-sustainable behaviour require deliberate processes and conscious choices (Mont, Lehner, & Heiskanen, 2014).
The appropriate application of both behavioural interventions is contingent on the intended outcome. It is important to systematically employ the right type of intervention for the right situation, taken existing empirical evidence into account. However, only a few studies have presented impactful nudging designs in order to mitigate climate change (e.g. Henkel, Seidler, Kranz and Fiedler, 2019; Loock, Staake and Thiesse, 2013). Moreover, research by Van Lieren et al. (2018) highlight the need for extended research to validate the effects and identify specific elements of these types of behavioural intervention strategies. In light of limited existing research, this study seeks to broaden the understanding of nudging vs. rational overrides and the effectiveness of both behavioural interventions on influencing pro-environmental choices on the short and long term. We will focus on the reduction of climate emissions from food consumption as food is the most environmentally relevant area of consumption with meat and dairy products having the highest global footprint of carbon, raw materials and water per kg of any food (EEA, 2014). By integrating key principles from behavioural economics, psychology and behavioural sciences, this approach goes beyond the
existing theory and current applicability of behavioural interventions. This study examines the direct and comparative influences of two behavioural interventions - nudging and rational overrides. We will argue that the effects of nudges are not lasting as they are context-specific
and designed to effect immediate behaviour. In order to substantiate existing literature in the promising and more effective usage of rational overrides compared to nudging on the long term (Van Lieren et al., 2018), this research will integrate both behavioural interventions to draw conclusions about the effectiveness and application of both interventions. Hence, this study addresses the following research question:
RQ: What are the effects of nudging vs. rational overrides on influencing pro-environmental
choices (eating vegetarian) and attitudes towards a vegetarian diet on the short and the long
Theoretical Framework
In this section, the literature review discusses the Dual Process Theory proposed by Kahneman (2003) regarding the architecture of cognition in the human brain and specifically, the role of decision-making processes in the determination of behaviour and the application of interventions to influence and change specific behaviour. Moreover, the nudging theory as postulated by Thaler and Sunstein (2008) and its potential for influencing behaviour will be outlined. We will argue that the effects of nudges are not lasting as they are context-specific and designed to effect immediate behaviour. Therefore, we will introduce an alternative behavioural intervention, rational override as formulated by Van Lieren, Calabretta, and Schoormans (2018). This intervention requires conscious deliberation as opposed to nudging. To formulate recommendations on long-term strategies for behavioural interventions, we will consider both interventions and its suitability and sustainability promote pro-environmental behaviour on the long term. The conceptual framework including hypotheses are also presented in this section.
Homo Economicus & The Dual Process Theory
For decennia the realms of science assumed that rationality was the antecedent and determinant of human behaviour. This rational view of humanity was translated in the Rational Choice Theory (Scott, 2000). This theory builds upon the idea that all action is fundamentally rational in character and is the result of deliberate, calculating decision-making processes. Within this framework, outcomes and patterns of choices are predicted based on the premise that individual preferences are always self-interested. By outweighing the potential costs against benefits, choices are translated into action that maximizes personal advantage. This view is often indicated with the term Homo Economicus; the reflective and logic decision maker, the master of rationality (Thaler, 2000). However, through the years
many opposing theories emphasized the irrationality of human behaviour and demonstrated this by means of scientific evidence (Tversky & Kahneman, 1979; Kahneman, 2003).
Extended research into intuition and reasoning stressed the bounded rationality of the human brain (Kahneman, 2003). The Dual Process Theory (Kahneman, 2003) asserts that the human brain operates in two different systems referred to as system 1 and system 2. The operations of System 1 are fast, automatic, effortless and often emotionally charged, while the operations of System 2 are slower, rational, effortful and deliberately controlled (Kahneman, 2003). The systems are also respectively referred to as automatic and reflective thinking (Thaler & Sunstein, 2008). These two modes of thinking and deciding roughly correspond to everyday concepts of reasoning and intuition. Automatic thinking is intuitive, fast and occurs
spontaneously, while reflective thinking is associated with a deliberate processing of information; it is slow, strenuous and requires focus (Kahneman, 2003).
Empirical and systematic research indicate that most thoughts and actions are
primarily intuitive in nature instead of the outcome of deliberate thinking (Kahneman, 2003). Mindless, intuitive behaviour is often associated with powerful, accurate and highly skilled performances. Due to its rapid and effortless nature, experience and high skills are easily acquired by means of prolonged practice. The automation of tasks is energy saving for the human brain, in which the overall capacity for mental effort is limited (Kahneman, 2003). Especially the decision-making process requires a lot of effort and brain capacity to make the right decisions. To reduce this cognitive capacity, people are inclined to take mental shortcuts. Within these shortcuts, people rely on their instinctive subconscious mindset instead of a rational and conscious deliberation (Van Lieren et al., 2018).
The Nudging Theory
Knowledge of cognitive operations being primarily automatic in nature has been implemented for multiple purposes. Especially the application of interventions to influence and change specific behaviour are on the rise. These behavioural and cognitive phenomena are considered as a toolbox for policy makers in attempts to improve the effectiveness of their interventions (Oliver, 2015). Behavioural interventions include a wide variety of approaches to altering social or physical environments to make certain behavioural choices more likely. The Nudging Theory is a pioneer within this field as postulated by Thaler and Sunstein in 2008. Based on their definition, a nudge refers to “any aspects of the choice architecture that alters people’s behaviour in a predictable way without forbidding any options of significantly
changing their economic incentives” (Thaler & Sunstein, 2008). The theory finds its origin in
the Dual Process Theory and goes along with the fact that mental operations are primarily intuitive in nature. People often rely on their intuitive subconscious mind set by means of mental shortcuts due to limited cognitive capacity of the brain. These shortcuts are contingent on basic, core capacities of the human brain. Therefore, actions and thoughts are strongly influenced by context. By slightly altering this choice environment, people can be navigated towards a desired and predictable outcome (Van Lieren et al., 2018). Therefore, nudging appears to be a very promising tool for altering automatic consumption into more sustainable and pro-environmental behaviour (Mont et al., 2014). Nudging successes range from
experiments increasing healthy food choices by placing fruits and vegetables closer to the cashier in a store than sweets (Goldberg & Gunasti 2007), to making the green electricity option easier to access by selecting it as the default option (Mont et al., 2014). More specifically, the implementation of reduced plate size in all-you-can-eat environment
effectively reduced energy intake and food waste (Freedman & Brochado, 2010). Altering the choice environment by changing a plate size from 24 to 21 cm reduced food waste by 15%
(Mont et al., 2014). So, based on previous research we can conclude that nudging is an effective tool for altering behaviour in a specific choice architecture.
Automatic & Reflective Nudges
The basic premise of the Nudging theory lies within the fact that the human decision-making process and accompanied behaviour is influenced by subtle, seemingly insignificant changes in the choice architecture or context. This principle activates a reaction in the subconscious part of the brain, whereby the person concerning is often unaware of the manipulation (Strassheim, 2016). In line with the distinction between automatic processes on the one hand and deliberate, reflective processes on the other hand (Kahneman, 2003; Thaler & Sunstein, 2008), Hansen and Jesperson (2013) suggest that there are also 2 types of nudging to
differentiate. The traditional nudge as postulated by Thaler and Sunstein always aims at influencing behaviour caused by automatic thinking without involving reflective thinking. Hansen and Jesperson introduce a second type of nudge, which involves reflective thinking and aims to influence behaviour as a result of a degree of deliberation (Hansen & Jespersen, 2013). Within this typology, the classification depends on the degree of conscious
engagement and the required response in order for the nudge to succeed. Both nudge
interventions require an active response, however type 1 nudges work automatically and type 2 nudges require reflection. Therefore, the distinction is also referred to as automatic nudges and reflective nudges. (Strassheim, 2016).
This reflective nudge intervention, also rational override as formulated by Van Lieren, Calabretta, and Schoormans in 2018, focuses on the implementation of micro moments of deliberate friction in the human brain. These rational override interventions are designed to disrupt mindless automatic interactions and stimulate conscious decision making on the contrary of a system 1 nudge or automatic nudge (Van Lieren et al., 2018). As a rational
override requires an active choice, conscious engagement and reflection, it has high potential for influencing a conscious learning process and make people more aware of their cognitive limitations (Strassheim, 2016).
For instance, the implementation of calorie labels on restaurant menu’s impacted food choices and food consumption. Adding a calorie label and a recommendation for daily caloric intake for an average adult resulted in 14% fewer calories compared to a menu without labels (Roberto, Larsen, Agnew, Baik, & Brownell, 2010). By providing these calorie labels,
automatic mindless behaviour was disrupted and reflective, conscious consideration was activated. Reflective nudging in the food domain in attempts to reduce meat consumption appeared to be effective as well. The introduction of carbon labels by a Swedish fast-food chain on their burgers resulted in a 16% increase in sales of burgers with a lower carbon footprint (Van Gilder Cooke, 2012). These positive effects suggest that rational override interventions are a promising tool for promoting pro-environmental behaviour and specifically, reducing meat consumption.
Based on previous research we can conclude that reflective and automatic nudges are effective in altering specific behaviour. Consequently, we expect that both behavioural intentions, nudging and rational overrides, lead to a greater level of immediate pro-environmental behaviour. Thus, based on previous findings and in line with theoretical evidence we propose the following hypothesis:
H1: Nudging and rational overrides both lead to a greater level of immediate pro-environmental choices (eating vegetarian) compared to providing no behavioural
While the immediate effectiveness of nudging is demonstrated by experimental research, less is known about the persistent effects on behavioural change. Therefore, political and scientific discussions about the suitability and sustainability of nudging are on the rise (Marchiori et al., 2017). Nudges are originally designed to alter immediate behaviour in a specific environment and are thus context-specific. This implies that the effects are not likely to persist when the nudge is removed (Van Lieren et al., 2018; Strassheim, 2016).
An experimental nudge intervention aimed at encouraging household recycling of waste shows an immediate increase in recycling, but the effects reduced three months later at the follow-up after the intervention (John, Cotterill, Moseley, Richardson, Smith, & Stoker, 2019). Moreover, a randomized controlled experiment into the effect of nudging on ordering climate-friendly dishes (e.g. fish, vegetarian meals) revealed unstable results. Besides observing a decrease in the treatment effect over time, post-intervention measurements
revealed that removing the nudge immediately restored pre-treatment meat sales. Specifically, a decrease in fish and vegetarian meals and an increase in meat sales were observed after removing the nudge (Gravert & Kurz, 2019). Therefore, academics highlight the need for more research on the impact of nudges over time to formulate recommendations on long-term strategies.
Moreover, nudging has been criticized for merely focusing on System 1 type of thinking, while leaving interaction with System 2 out of the process. This criticism is especially relevant in the pro-environmental context as research has shown that long-term changes towards sustainable behaviour require deliberative processes and conscious choices to be implemented in automatic, daily routinized behaviours (Mont et al., 2014). As automatic nudges primarily address system 1, effects will be less likely to persist on the long term. On the other hand, the effects of a reflective nudge are more likely to endure on the long-term due to the activation of system 2. Especially since it is sometimes required to help people slow
down, create awareness and let them actively make a decision. To generate a lasting impact on behaviour change, it is important to override the subconscious and stimulate people to be active and aware. Particularly in the promotion of pro-environmental behaviour, it is
necessary to make significant changes in patterns of consumption. Sustainable solutions may not succeed unless people are persuaded on a deeper level so that they implement the changes in their behavioural patterns (Stoker, 2012).
In attempts to combine numerous theoretical perspectives, Fishbein and Cappella (2006) propose the Integrative Model, where attitude is defined as one of the primary determinants of behaviour. In line with this theory, one’s attitude towards performing the behaviour must be addressed to increase its effectiveness on behavioural change. However, influencing attitudes is not part of the nudge criterion as system 1 is primarily addressed (Mont et al., 2014). These characteristics of a nudge are accountable for its immediate effectiveness and its wide suitability as we do not need to change minds in order to change behaviour. However, this simultaneously disputes the effectiveness of nudging on the long term.
Previous research revealed that environmental attitudes are strongly related to behaviours that do not affect people’s daily live enormously (e.g. food consumption).
Environmental attitudes are therefore considered as a strong behavioural determinant for pro-environmental behaviour (Gatersleben, Steg, & Vlek, 2002). In line with this argument, attitude is an important aspect in the construction of an effective behavioural intervention in the of pro-environmental context. By changing how people think, interventions could produce persisting behaviour change, as long as the attitude is retained (Frey & Rogers, 2014).
By means of a rational override, system 2 is stimulated and people are triggered to actively think about the implemented moment of friction. With the activation of the critical mind, people are able to make a deliberate decision and form a judgement autonomously. This
process is important as it may impact the endurance of attitude and behaviour change. It is expected that people become more positive towards the desired behaviour when conscious deliberation is addressed. Especially as a cognitively processed message may result in more persistent attitude change, whereas for messages processed less cognitively, (e.g. automatic nudge) attitude change may be temporal or absent (Rodrigue, 2006). Consequently, the effects will be more likely to be translated in actual behaviour, as attitude is a prominent predictor for behaviour. An automatic nudge circumvents this thinking process by operating unconsciously which makes it less likely that attitudes are created or changed. (Mont et al., 2014), and therefore the desired behaviour will be less likely to be administered in a different context or on the long term. Moreover, cognitive learning theories emphasize the importance of retention and learning in the effectiveness of a persuasive communication. The persisting effects of persuasion communications are based on a cognitive response in which rehearsal and learning are key (Greenwald, 1968). A greater amount of cognitive processing increases knowledge about the intent of the message, and is therefore more likely to be considered acceptable (Rodrigue, 2006). This reasoning indicates that active, cognitive participation of the recipient and the formation of an attitude is fundamental for the effects to be lasting. These
requirements are met by means of a rational override intervention while not part of the nudge criterion. Therefore, we propose the following hypothesis:
H2: Rational overrides result in more positive attitudes towards a vegetarian diet compared to a nudging intervention or no behavioural intervention.
In line with this argument, the effects of a rational override behavioural intervention, compared to a nudge intervention, are more likely to persist on the long term as a Rational Override is not context-specific, requires conscious deliberation and is more likely to change
attitudes. So, based on previous findings and in line with aforementioned inference we propose the following hypothesis:
H3: A Rational override leads to a greater level of pro-environmental choices (eating vegetarian) compared to a nudging intervention or no behavioural intervention on the long
term. No difference in long term pro-environmental choices will be found for the nudging
intervention compared to the control condition.
As empirical research and theory on behaviour explains that positive attitudes are strongly related to behaviour, we argue that the formation of these attitudes are accountable for the duration of the effects. As rational overrides address attitudes towards the desired behaviour, this behaviour will be more likely to persist on the long term. In a similar vein, we argue that attitude is the underlying mechanism that mediates the effect of rational overrides on pro-environmental choices on the long term. In this study, we therefore hypothesize:
H4: Rational overrides (vs nudges) result in more positive attitudes towards a vegetarian diet and will therefore positively affect pro-environmental choices on the long term.
Methods Participants
For the purpose of testing our hypotheses, an online experiment was conducted. This research uses data that was specifically collected for this study by means of a convenience sampling technique. The sample was drawn from friends, family and social media acquaintances of the researcher. The questionnaire was filled in by 128 respondents in total. Twelve subjects had to be excluded from the analysis due to unfinished surveys. The remaining sample consisted of 116 participants of which 87,1% women. The age range varied from 20 years to 67 years (M = 38.08, SD = 14.40). In addition, the majority of the participants in the experiment had either completed a Bachelor’s Degree (48.3%) or a Master’s Degree (31%). Table 1 provides an overview of participant’s demographic characteristics.
Table 1. Sample Characteristics
Variable Characteristics Age (years 18 - 24 17.2% 25 - 34 35.4% 35 - 44 8.6% 45 - 55 24.1% 55 + 14.7% Gender Female 87.1% Male 12.9%
Education Primary Education 0.9%
High School Diploma 19%
Bachelor’s Degree 48.3%
Master’s Degree 31%
Doctoral Degree (PhD) 0.9%
Research Design
The aim of this study was to research the effects of nudging vs. rational overrides on
influencing pro-environmental behaviour (reducing meat consumption) and attitudes towards a vegetarian diet on the short and long term. Accordingly, the study employed a three-group between-subjects design (nudge vs. rational override vs. control), and participants were randomly assigned to one of the three conditions in which the behavioural intervention was experimentally manipulated.
Procedure
Prior to the experiment, participants were shown an informed consent statement (see
Appendix 1). The participants were informed that this study was about their eating behaviour and food preferences and that the data from this experiment was collected for the purpose of a master thesis. The exact goal of the study was not revealed to prevent participants of
becoming aware of this purpose. Moreover, their anonymity was guaranteed and the survey was entirely on voluntary basis. To participate in the experiment, participants had to agree with these terms and conditions. The participants were instructed to answer all question truthfully. First, participants were asked some basic demographic questions including gender, age and education. Gender identification was measured by a multiple-choice question
including the answers ‘Male’, ‘Female’, ‘Identify as neither’ and ‘Prefer not to say’. Age was measured by means of an open-ended question where participants where askes to fill in their current age. To identify participant’s highest level of completed education, a multiple-choice question was implemented consisting of the answers ‘None’, ‘Primary Education’, ‘High School Diploma’, ‘Bachelor’s Degree’, ‘Master’s Degree’ and ‘Doctoral Degree (PhD)’. After basic information about the participants was assembled, participants were randomly assigned to one of three conditions, using the randomizer option in Qualtrics. For each
condition, a menu was created that reflected a true menu to increase the external validity of the research. The lay-out was kept consistent in all conditions. After the manipulation, participant indicated what they wanted to eat from the menu to measure immediate choices. After this, several intelligence riddles were presented for the participant to solve. These questions served as a break in the experiment to stretch time and to distract participants from the true purpose of the experiment. Next, long term food choices and attitudes towards a vegetarian diet were measured.
Independent variables
Control Condition. In the control condition, a menu consisting of meat and vegetarian dishes was presented (see Appendix 2). The dishes were subdivided under the heading ‘Appetizer’, ‘Main course’ and ‘Dessert’. The situation sketch contained the
following sentence; ‘Imagine you are in a restaurant and the menu below is presented to you’. Meat and vegetarian appetizer options contained ‘Chicken Soup’, ‘Beef Carpaccio’, ‘Burrata’ and ‘Mediterranean Salad’. Main course options featured ‘Mushroom Risotto’, ‘Tournedos’, ‘Spaghetti Aglio Olio’, ‘Rib Eye’, ‘Pizza Margherita’ and ‘Cheese Burger’. Lastly, dessert options were ‘Tiramisu’, ‘Cheesecake’ and ‘Lemon Meringue’.
Nudging. In the nudge condition, a menu consisting of solely vegetarian dishes was presented to the participant (see Appendix 2). The situation sketch contained the following sentence; ‘Imagine you are in a restaurant and the menu below is presented to you’. The dishes on the menu were subdivided under the heading ‘Appetizer’, ‘Main course’ and ‘Dessert’. Vegetarian appetizer options contained ‘Burrata’ and ‘Mediterranean Salad’. Main course options featured ‘Mushroom Risotto’, ‘Spaghetti Aglio Olio’ and ‘Pizza Margherita’. Lastly, dessert options were ‘Tiramisu’, ‘Cheesecake’ and ‘Lemon Meringue’. Besides, participants were given the option to see the menu including meat options by the following
text ‘If you would like to see the menu with meat option, please check this box. If not, please continue and indicate your preferences below.’
Rational Override. In the rational override condition, a menu consisting of meat and vegetarian options was presented to the participant (see Appendix 2). The situation sketch contained the following sentence; ‘Imagine you are in a restaurant and the menu below is presented to you’. To disrupt mindless thinking and stimulate a deliberate decision, the following information was implemented; ‘Did you know? Eating vegetarian saves per day:
- 1 month of shower water usage
- Approximately 1740-gram Co2 emissions - 1 month of work for a tree
This information was acquainted from development organization Oxfam Novib that mobilizes the power of people and strives to defeat poverty and promote sustainable development. The dishes on the menu were subdivided under the heading ‘Appetizer’, ‘Main course’ and ‘Dessert’ and include both vegetarian and meat options.
Dependent variables
Immediate environmental choices. To measure participant’s immediate pro-environmental choices, a multiple-choice question with regard to meal choice was
implemented. Participants were asked to choose an appetizer, main course and dessert. The multiple-choice options corresponded exactly with the options presented in the menu.
Depending on the condition, participants were able to choose from either vegetarian and meat options or vegetarian as the default option as previously described. The multiple-choice answers were coded dichotomously with 0 for meat and 1 for vegetarian meals. This resulted in a binary variable for the immediate pro-environmental choices for appetizer and main course. To measure these choices as a whole, the variables immediate pro-environmental
choices for appetizer and main course were combined. Again, the answers were coded dichotomously with 0 for meat and 1 for both appetizer and main course vegetarian.
Long term pro-environmental choices. To determine the effect of the behavioural intervention on the longer term, participants were requested to describe their dinner plans for the upcoming 3 days as specific as possible, including the main ingredients. The question was open-ended with a separate form field which represented the upcoming days. The form fields were indicated with ‘Day 1’, ‘Day 2’ and ‘Day 3’. Due to an increase in blanc fields in this question, seven subjects had to be excluded from the analysis with this variable. The
remaining sample for the long term pro-environmental variable consisted of 109 participants. The open-ended answers were coded dichotomously with 0 for meat and 1 for vegetarian meals. Based on the definition of the Dutch Vegetarian Union, meals consisting of either meat or fish were coded 0. This resulted in three binary variables for Day 1, Day 2 and Day 3. To measure long term pro-environmental choices as a whole, these variables were combined. Resulting in a 0 for meat or fish and 1 for three vegetarian meals.
Attitude towards a vegetarian diet. Attitude towards a vegetarian diet was measured using the Ambivalence Towards Meat scale by Berndsen and Van Der Pligt (2004) by means of two items. The first item focussed on vegetarian dieting with the following statement; ‘I think a vegetarian diet is…’. Four semantic scales measured the respondent’s attitude. The
four items were ‘bad-good’, ‘unpleasant-pleasant’, ‘unfavourable-favourable’ and ‘negative-positive’. The second item addressed participant’s attitude towards the consumption of meat with the following statement; ‘I think eating meat is…’. The multi-item scale represented the respondent’s attitude towards the consumption of meat. The four items were ‘bad-good’, ‘unpleasant-pleasant’, ‘unfavourable-favourable’ and ‘negative-positive’. As this item was negatively phrased in comparison with the first item, this scale needed to be recoded. After reverse coding, a high score on both dimensions indicated a positive attitude towards a
vegetarian diet and a negative attitude towards the consumption of meat (M = 4.90, SD = 1.22). A factor analysis was performed with all 8 items building the scales for attitude towards a vegetarian diet. All the items had higher commonalities than then threshold of 0.4 and were incorporated in the solutions. The items were allocated to two factors with
Eigenvalues larger than 1 and a total variance explained of 77.83%. The rotated component matrix showed that the items exclusively loaded on one component (59.05% of total variance explained), so for the purpose of this research both components were further used in the analysis as one construct. Cronbach’s alpha was calculated for the items to test the reliability and internal consistency of the scales. The Cronbach’s alpha statistic was .899.
Results
Randomization check. To check if participant’s age was comparable over the different behavioural intervention conditions, a One-way Analysis of Variances was conducted. The analysis showed that participant’s mean age in the control condition (M = 39.29, SD = 13.93) was not significantly different from participant’s mean age in the nudge condition (M = 37.70, SD = 14.87) and the rational override condition (M = 37.13, SD = 14.70), F (2, 113) = .0237, p = .789). These results indicate that age did not differ
significantly across the conditions. To check if participant’s gender was evenly distributed over the conditions, a Chi Square Analysis was conducted. The Crosstab Chi-Square showed that gender did not significantly differ across the conditions, X2(2) = .03, p = .984. This means that the random allocation of gender between the conditions was equally divided. A Chi Square Analysis revealed that education differed significantly across the conditions, X2 (8) = 18.75, p = .016. This indicates that the highest level of education obtained by
participants varied across the different conditions and may influence results in further analysis. However, as 40% of the cells have an expected count less than 5, the test is not reliable for this ordinal scale. Therefore, a Kruskal-Wallis Test was completed to generate statistically accurate results. Again, a statistically significant difference between conditions was observed, H (2) = 13.505, p = .001. To ensure that the effects of the behavioural intervention were not caused by educational differences, this variable was included as a control variable in further analysis.
Immediate environmental choices. To test whether participant’s immediate pro-environmental choices differed across conditions, a Chi-Square analysis was conducted. This crosstab showed that in the control condition 56,1% of the participants chose two vegetarian dishes (appetizer and main course). In the Rational Override condition, this percentage was higher, namely 73,7%. The highest amount of immediate pro-environmental choices was
observed in the Nudge condition with 75,7%. However, the difference between the conditions was not statistically significant, X2(2) = 4.24, p = .120. A follow-up analysis to control for educational differences in the conditions was conducted by means of a logistic regression. A binary logistic regression was run to analyse the effects of the conditions on immediate pro-environmental choices. The binary dependent variable was coded dichotomously (meat 0, both vegetarian 1). The analysis indicated that the difference between the Rational Override condition and the Nudge condition did not differ significantly, b = -.006, p = .991. However, the difference between the Nudge condition and the control condition on immediate
environmental choices was marginally significant, b = .892, p = .073. Moreover, the
difference between the Rational Override condition and the control condition on immediate environmental choices was marginally significant as well, b = .898, p = .078. These results indicate that in the Nudge Condition, as well as in the Rational Override condition there was a higher chance of immediate pro-environmental choices compared to the control condition.
In depth analysis on participant’s appetizer choice revealed that 89.5% of the
participants in the Rational Override condition chose a vegetarian appetizer. This is slightly higher than participants in the Nudge condition (75.7%) and the control condition (78.0%). A binary logistic regression revealed that the difference between the Rational Override and the Nudge condition was not statistically significant b = -.948, p = .161. Moreover, the difference between the Rational Override condition and the control condition wasn’t significant either, b = -.815, p = .227. Lastly, the difference between the Nudge condition and the control
condition was not significant, b = -.134, p = .803. These results indicate that participants in the Rational Override condition were most likely to choose a vegetarian appetizer compared to the Nudge and control condition. However, this difference was not statistically significant.
To explore whether these results differed for the main course, a Chi-Square analysis was conducted. This crosstab revealed that 94.6% of the participants in the Nudge condition
chose a vegetarian main course. This percentage was lower in the Rational Override condition (76.3%) and the control condition (73.2%). To test whether this difference was statistically significant, a binary logistic regression was completed. The results revealed that the
difference between the Nudge condition and the control condition was statistically significant, b = 1.860, p = .022. Moreover, the difference between the Nudge condition and the Rational Override condition was marginally significant as well, b = -1.554, p = .064. The difference between the Rational Override condition and the control condition was however not statistically significant, b = .306, p = .576. These results indicate that participants in the Nudge condition chose significantly more vegetarian main courses compared to the Rational Override condition and the control condition.
Hence, the results partly confirm the first hypothesis suggesting that nudging and rational overrides both lead to a greater level of immediate pro-environmental choices compared to providing no behavioural intervention. However, taken into consideration that the results approached the threshold significance value of .05.
Attitude towards a vegetarian diet. A Univariate Analysis of Variance was run to check whether the attitude towards a vegetarian diet differed between conditions, with
education as covariate. The Levene’s test was not significant (p = .56) suggesting that we can assume equal variances for the three groups. The analysis shows that the effect of condition on attitude is marginally significant, F (2,112) = 2.97, p = .055. The effect size is very small, eta2 = .05. The conditions thus explain 5% of the variance in attitude towards
pro-environmental choices. Follow-up comparisons showed that among participants in the
Rational Override condition, the average attitude towards a vegetarian diet was notably higher (M = 5.24, SD = 1.10) compared to participants in the Nudge condition (M = 4.66, SD = 1.15). This mean difference (Mdif = .67) is statistically significant, p = .023. The average attitude towards a vegetarian diet was lower among participants in the control condition, (M =
4.79, SD = 1.33) than among participants in the Rational Override condition. Again, this mean difference was marginally significant, p = .058. However, there is no substantial difference between the Nudge condition and the control condition with respect to their effect on attitude towards a vegetarian diet (Mdif = .129, p = .637). Hence, the findings lend support to the prediction that a Rational Overrides results in more positive attitudes towards a vegetarian diet than a Nudging intervention or no behavioural intervention.
Long term pro-environmental choices. The third hypothesis assumed that a Rational override leads to a greater level of pro-environmental choices on the long term, compared to a nudging intervention or no behavioural intervention. A Chi-Square analysis shows that
participants in the Rational Override condition made more pro-environmental dinner choices on the long term (3 days of vegetarian meals). Specifically, 19.4% of the participants in the rational override planned all upcoming days with vegetarian meals. In the Nudge condition this was 11,1% of the participants and 16.2% in the control condition. However, this
difference was not statistically significant, X2(2) = .966, p = .617. A binary logistic regression was run to analyse the effects of the conditions on long term pro-environmental choices, while controlling for educational differences. The analysis revealed that the difference between the Rational Override condition and the control condition on long term
environmental choices did not reach a level of significance, b = -.159, p = .806. Moreover, the difference between the Nudge condition and the control condition was not significant either, b = -.443, p = .532. Lastly, the difference between the Rational Override condition and the Nudge condition was not significant, b = -.284, p = .689.
In depth analysis on participants meal planning revealed that for the first day, 51.4% of the participants in the Rational Override condition were planning on having a vegetarian meal. In the Nudge condition this was 30.6% and in the control condition 42.1%. A binary logistic regression revealed that the difference between the Nudge condition and the control
condition was not statistically significant, b = -.542, p = .277. Moreover, the difference between the Rational override condition and the control condition wasn’t significant either, b = .054, p = .915. The difference between the Rational Override condition and the Nudge condition did not reach significance either, b = -.596, p = .246.
For the next day, 47.4% of the participants in the control condition were planning on having a vegetarian meal. This was lower in the Nudge condition (30.6%) and the Rational Override condition (45.7%). A binary logistic regression revealed that the difference between the Rational Override condition and the control condition was not statistically significant, b = .241 p = .630. The difference between the Rational Override condition and the Nudge
condition wasn’t significant either, b = -.505, p = .326. Lastly, the difference between the Nudge condition and the control condition was not statistically significant, b = -.746, p = .128.
On the third day, 50% of the participants in the control condition planned a vegetarian meal. This was lower in the Rational Override condition (42.4%) and the Nudge condition (36.4%). To check whether these differences were statistically significant, a binary logistic regression was run. The analysis revealed that the difference between the Nudge condition and the control condition was not significant, b = -.563, p = .254, nor the difference between the Rational Override condition and the control condition, b = -.318, p = .534. The difference between the Nudge condition and Rational Override condition wasn’t significant either, b = -.244, p = .638. Therefore, there is not enough evidence in support of the assumption that a rational override intervention leads to a greater level of long term pro-environmental choices compared to a nudging intervention or no behavioural intervention. However, as predicted, no substantial differences in long term pro-environmental choices were found for the nudging intervention compared to the control condition.
Attitude mediating long term pro-environmental choices. To test whether attitude towards a vegetarian diet mediated the relationship between condition and long term pro-environmental choices, several analyses were conducted. To establish the relationship between condition and long-term environmental choices, while controlling for education, a binary logistic regression was run. The analysis revealed that the difference between the Rational Override condition and the control condition on long term environmental choices did not reach a level of significance, b = -.159, p = .806. Moreover, the difference between the Nudge condition and the control condition was not significant either, b = -.443, p = .532. Lastly, the difference between the Rational Override condition and the Nudge condition was not significant, b = -.284, p = .689.
A linear regression was completed to examine the relationship between condition and attitude towards a vegetarian diet. The linear regression with condition and education as independent variables and attitude towards a vegetarian diet as dependent variable was not significant, F (2, 113) = 1.626, p = .201. The prediction was very weak, (R = .028) The conditions thus explain 2.8% of the variance in attitude a vegetarian diet. Condition positively predicts attitude towards a vegetarian diet and this effect was marginally significant, b = .254, p = .076.
To establish the relationship of condition on long term pro-environmental choices, while controlling for education and adding attitude towards a vegetarian diet, a logistic regression was run. Results revealed no statistically difference after implementing attitude towards a vegetarian diet between the Nudge condition and the control condition, b = -.203, p = .798, nor between the Rational Override condition and the control condition, b = -.609, p = .399. Moreover, the difference between the Nudge condition and the Rational Override condition wasn’t statistically significant either, b = -.405, p = .615. However, the effect of attitude towards a vegetarian diet on long term-pro environmental choices was statistically
significant, b = 1.229, p < .001. This indicates that a positive attitude towards a vegetarian diet was related to long term pro-environmental choices. However, the mediation analysis revealed that attitude towards a vegetarian diet was no significant mediator in the relation between condition and long term pro-environmental behaviour. Hence, the findings lend no support to the assumption that Rational Overrides (vs nudges) result in more positive attitudes towards a vegetarian diet and will therefore positively affect pro-environmental choices on the long term.
Table 2. Overview of the Hypotheses
Hypothesis 1 Nudging and rational overrides both lead to a greater level of immediate pro-environmental choices (reducing meat consumption) compared to providing no behavioural intervention.
Accepted
*Note: p < .10
Hypothesis 2 Rational overrides result in more positive attitudes towards a vegetarian diet compared to a nudging intervention or no behavioural intervention.
Accepted
Hypothesis 3 A Rational override leads to a greater level of pro-environmental choices (reducing meat consumption) compared to a nudging intervention or no behavioural intervention on the long term.
Not accepted
Hypothesis 4 Rational overrides (vs nudges) result in more positive attitudes towards a vegetarian diet and will therefore
positively affect pro-environmental choices on the long term.
Discussion
The aim of this research was to improve scientific understanding of applied behavioural interventions in a pro-environmental context. More specifically, this study focuses on the effectiveness of a Nudge compared to a Rational Override on influencing pro-environmental choices on the short and long term. Several studies have investigated the impact of behavioural interventions on pro-environmental behaviour, (Henkel et al., 2019; Loock et al., 2013) or the persistence of the effects (Frey & Rogers, 2014; Marchiori et al., 2017). However, this study provides new insights into the domain of behavioural
interventions by experimentally manipulating two interventions and examine its effectiveness and its consistency in the context of climate change adaptation. The findings suggest that both behavioural interventions, a Nudge and Rational Override, lead to the choice of more
vegetarian dishes on the short term. This simultaneously implicates that a simple and unobtrusive adjustment of a menu can contribute to an immediate reduction in meat consumption. Especially altering the convenience of ordering meat in the Nudge condition resulted in significantly more vegetarian main courses. These results add up to existing evidence of the effectiveness of automatic nudges (Freedman & Brochado, 2010; Goldberg & Gunasti 2007) and reflective nudges (Roberto et al., 2010; Van Gilder Cooke, 2012).
Furthermore, results of the experiment showed that a Rational Override leads to a more positive attitude towards a vegetarian diet, compared to a Nudge or no behavioural
intervention. This indicates that a cognitively processed message through a Rational Override is more likely to result in attitude change, while by means of a Nudge it is less likely that attitudes are created or changed. Participants exposed to information about the positive effects of pro-environmental behaviour in the Rational Override condition, were triggered to
cognitively respond, learn and retain (Rodrigue, 2006). The stimulation of system 2 resulted in the activation of the critical mind, in which a judgement was formed or altered.
The results of the experiment lend however no support to the theorizing that a Rational Override leads to a greater level of pro-environmental choices on the long term, compared to a Nudging intervention or no behavioural intervention. These findings contradict the
argumentation that long term changes towards sustainable behaviour require deliberative processes as intended by a Rational Override intervention (Mont et al., 2014). However, there is a possibility that participants already planned their (non)vegetarian dinners before
participating in the experiment. Moreover, as the open-ended questions were coded manually, the interpretation of several meals (e.g. spaghetti, salad) might be subjective and ambiguous. Therefore, the variable is not extremely reliable. Future research should take this into
consideration by obtaining more data and filter out indefinite or questionable answers. Besides, to measure long term pro-environmental choices, several riddles were implemented to stretch time. However, in practice the long term pro-environmental choices took place several minutes after immediate pro-environmental choices. Future research could delay the effects by means of a longitudinal research and measure long-term pro-environmental choices after a longer period of time. Nevertheless, in this manner we did examine the
context-specificity of the nudges, as the long-term behaviour was measured in a distinctive context. The mediation analyses showed that long term pro-environmental behaviour induced by type of behavioural intervention was not mediated by attitude towards a vegetarian diet. In other words; the Rational Override succeeded in positive attitude changes, however this in turn did not result in long term pro-environmental behaviour. These findings lend support for the theorizing that providing information does not necessarily lead to changes in behaviour. The so-called attitude-behaviour gap is confirmed as a change in attitude towards a vegetarian diet did not simultaneously lead to more vegetarian choices on the long term (Mont et al., 2014). To overcome this gap and to formulate recommendations on long-term strategies, other nudges should be considered and examined.
The substantial results in our experiment contribute to existing research in the
promising application of Nudging towards sustainable behaviour (e.g. Mont et al., 2014) and additionally introduces an alternative Rational Override intervention that might be even more effective. Therefore, future research should examine this innovative behavioural intervention and explore its effectiveness in various fields of interest. Especially as nudging is often associated with a violation of individual autonomy (Felsen, Castelo, & Reiner, 2013). The claim that nudging is merely a manipulation of people’s choice and forms a threat to our autonomy is a central tenet in the political and normative criticism (Hansen & Jespersen, 2013). A Rational Override, on the contrary, circumvents these points of criticism by its transparent and reflective character and therefore increases acceptance and applicability. Future research should take the opportunities of a Rational Override for influencing behaviour on the long term seriously, despite the nonsignificant results of this particular research. For instance, by means of a longitudinal research with multiple measurements. Moreover, to emphasize the fact that a Rational Override intervention is not context-specific, future research could detach a Rational Override from specific stimulus material by providing information about the desired behaviour separately. In this way, attention and focus of the participants is determined, which increases the learning aspect of the cognitive process. This could increase the chance of retention and consequently, the re-application of the acquired knowledge on the long term. Based on the effectiveness of a Rational Override on influencing attitude towards the desired behaviour as proven in this experiment, it is still expected, despite nonsignificant results, that this influences actual behaviour on the long term. Future research should take this into consideration, especially in the context of pro-environmental behaviour, as climate change is one of the most critical challenges we are facing. To mitigate the process of global warming, we need substantial change in behaviour which is to a large extent
problem by means of disrupting mindless, habitual behaviour. Besides the need for
awareness, people need to be informed about their individual contribution to mitigate climate change. Future research should therefore focus on a combination of several behavioural interventions instead of the current, independent scientific approach. Research into the effectiveness of an intervention that triggers both mindsets might provide new insights in the academic field.
To maximize effectiveness in the practical application of behavioural interventions, the intended outcome is the most important factor to consider. It appears that nudging might be more effective in situations in which there is one unified outcome, targeted at a relatively large group of people. Rational Overrides on the other hand are more effective to change behavioural routines that take place in different environments, at different times and or require a change in attitude. Bringing these findings into practice, policy makers could consider the implementation of vegetarian meals as the default option as proven to be effective in this research. For instance, in a restaurant, supermarket, online or in an airplane. Making the option for vegetarian dishes more prominent and easier to access than meat, will result in an increase in immediate pro-environmental choices. However, for the new routine to become a habit, a nudge needs to be repeated frequently. Therefore, policy makers should consider a nudge in combination with an additional Rational Override. This combination addresses both system 1 and system 2 and will therefore not only result in immediate effectiveness, but will also create awareness. For instance, informing people about the Co2 emissions of their car, water usage, energy usage and eating patterns in combination with an automatic nudge allows people to autonomously decide to alter this behaviour into more pro-environmental choices. This will increase the chance of a re-application of the desired behaviour in a different context and will also circumvent the autonomy issue as associated with the Nudging Theory in general.
To conclude, this study was one of first to experimentally manipulate two different behavioural interventions separately in the context of pro-environmental behaviour while simultaneously examine its effectiveness on the short and long term. The results are a promising step towards climate change mitigation and adaptation and can be considered as a nudge towards more experimental research on the application of behavioural interventions
References
Berndsen, M., & Van der Pligt, J. (2004). Ambivalence towards meat. Appetite, 42(1), 71-78. Dunlap, R. E., & Brulle, R. J. (Eds.). (2015). Climate change and society: Sociological
perspectives. Oxford University Press.
EEA (2014), Environmental Impacts of Production-Consumption Systems in Europe. Copenhagen, European Environment Agency, pp. 64.
Felsen, G., Castelo, N., & Reiner, P. B. (2013). Decisional enhancement and autonomy: public attitudes towards overt and covert nudges. Judgment & Decision Making, 8(3). Fishbein, M., & Cappella, J. N. (2006). The role of theory in developing effective health
communications. Journal of Communication, 56, S1-S17.
Freedman, M.R. & Brochado, C. (2010), ‘Reducing Portion Size Reduces Food Intake and Plate Waste’ Obesity, Vol. 18 No. 9, pp. 1864– 1866.
Frey, E., & Rogers, T. (2014). Persistence: How treatment effects persist after interventions stop. Policy Insights from the Behavioral and Brain Sciences, 1(1), 172-179.
Gatersleben, B., Steg, L., & Vlek, C. (2002). Measurement and determinants of
environmentally significant consumer behavior. Environment and behavior, 34(3), 335-362.
Goldberg, M. E. & Gunasti, K. (2007), ‘Creating an Environment in Which Youths Are Encouraged to Eat a Healthier Diet’ Journal of Public Policy & Marketing, Vol. 26 No. 2, pp. 162–181.
Gravert, C., & Kurz, V. (2019). Nudging à la carte: a field experiment on climate-friendly food choice. Behavioural Public Policy, 1-18.
Greenwald, A. G. (1968). Cognitive learning, cognitive response to persuasion, and attitude change. Psychological foundations of attitudes, 1968, 147-170.
Hansen, P. G., & Jespersen, A. M. (2013). Nudge and the manipulation of choice: A
framework for the responsible use of the nudge approach to behaviour change in public policy. European Journal of Risk Regulation, 4(1), 3-28.
Henkel, C., Seidler, A. R., Kranz, J., & Fiedler, M. (2019). How to nudge pro-environmental behaviour: An experimental study.
IPCC, 2013: Climate Change 2013: The Physical Science Basis. Contribution of Working Group I to the Fifth Assessment Report of the Intergovern- mental Panel on Climate
Change [Stocker, T.F., D. Qin, G.-K. Plattner, M. Tignor, S.K. Allen, J. Boschung, A.
Nauels, Y. Xia, V. Bex and P.M. Midgley (eds.)]. Cambridge University Press, Cambridge, United Kingdom and New York, NY, USA, 1535 pp.
John, P., Cotterill, S., Moseley, A., Richardson, L., Smith, G., & Stoker, G. (2019). Nudge, nudge, think, think: Experimenting with ways to change citizen behaviour. Manchester
University Press.
Kahneman, D. (2003). Maps of bounded rationality: Psychology for behavioral economics. American economic review, 93(5), 1449-1475.
Loock, C. M., Staake, T., & Thiesse, F. (2013). Motivating energy-efficient behavior with green IS: an investigation of goal setting and the role of defaults. MIS quarterly, 1313-1332.
Marchiori, D. R., Adriaanse, M. A., & De Ridder, D. T. (2017). Unresolved questions in nudging research: Putting the psychology back in nudging. Social and Personality Psychology Compass, 11(1), e12297.
Mont, O., Lehner, M., & Heiskanen, E. (2014). Nudging a tool for sustainable behaviour? The Swedish Environmental Protection Agency.
Oliver, A. (2015). Nudging, shoving, and budging: Behavioural economic‐informed policy. Public Administration, 93(3), 700-714.
Oxfam Novib (n.d.). Dit effect heeft één keer per week vegetarisch eten. Retrieved from
https://www.oxfamnovib.nl/blogs/dagelijks-leven/dit-effect-heeft-een-keer-per-week-vegetarisch-eten
Rodrigue, C. S. (2006). The impact of masking of persuasive intent on persuasive message effectiveness. Louisiana State University and Agricultural and Mechanical College. Roberto, C. A., Larsen, P. D., Agnew, H., Baik, J., & Brownell, K. D. (2010). Evaluating the
impact of menu labeling on food choices and intake. American journal of public health, 100(2), 312-318.
Scott, J. (2000). Rational choice theory. Understanding contemporary society: Theories of the present, 129, 671-85.
Stoker, Gerry (2012), The Politics of Nudge: dilemmas in implementing policies for sustainable consumption. University of Southampton, pp. 17.
Strassheim, H. (2016). Not all nudges are automatic: freedom of choice and informative nudges.
Thaler, R. H. (2000). From homo economicus to homo sapiens. Journal of economic perspectives, 14(1), 133-141.
Thaler, R., & Sunstein, C., (2008), Nudge: improving decisions about health, wealth, and happiness, New Haven.
Tversky, A., & Kahneman, D. (1979). Prospect theory: An analysis of decision under risk. Econometrica, 47(2), 263-291.
Urry J. (2015) Climate Change and Society. In Why the Social Sciences Matter. Palgrave Macmillan, London https://doi.org/10.1057/9781137269928_4
Van Gilder Cooke, Sonia (2012). Why Going Green Can Mean Big Money for Fast-Food Chains. Times, April 9.
Van Lieren, A., Calabretta, G., & Schoormans, J. (2018). Rational Overrides: Influence Behaviour Beyond Nudging. doi: 10.21606/dma.2018.699
Appendix Appendix 1 – Informed Consent
Dear participant,
With this letter, I would like to invite you to participate in a research study to be conducted under the auspices of the Graduate School of Communication, a part of the University of Amsterdam.
In the survey, several questions will be asked about your eating behaviour. The study will take about 5 minutes.
As this research is being carried out under the responsibility of the Graduate School of Communication University of Amsterdam, we can guarantee that:
1) Your anonymity will be safeguarded, and that your personal information will not be passed on to third parties under any conditions, unless you first give your express permission for this.
2) You can refuse to participate in the research or cut short your participation without having to give a reason for doing so.
3) Participating in the research will not entail your being subjected to any appreciable risk or discomfort, the researchers will not deliberately mislead you, and you will not be exposed to any explicitly offensive material.
4) No later than five months after the conclusion of the research, we will be able to provide you with a research report that explains the general results of the research.
For more information about the research and the invitation to participate, you are welcome to contact the project leader Linde Mouw at any time (linde.mouw@student.uva.nl).
Should you have any complaints or comments about the course of the research and the procedures it involves as a consequence of your participation in this research, you can contact the designated member of the Ethics Committee representing ASCoR, at the following address:
ASCoR Secretariat, Ethics Committee, University of Amsterdam, Postbus 15793, 1001 NG Amsterdam; 020‐525 3680; ascor‐secr‐fmg@uva.nl. Any complaints or comments will be treated in the strictest confidence.
We hope that we have provided you with sufficient information. We would like to take this opportunity to thank you in advance for your assistance with this research, which we greatly appreciate.
Kind regards, Linde Mouw
Appendix 2 – conditions
