1
Feedback for Future
The effect of individually perceived
environmental impact on public blaming
2
Master Thesis
MSc. Marketing Management
Feedback for Future
The effect of individually perceived
environmental impact on public blaming
Elana Pfaller
University of Groningen Faculty of Economics and Business
Department of Marketing
First Supervisor: Dr. Mathilde van Dijk Second Supervisor: Dr. Sumaya Albalooshi
June 17th, 2019
Student number: S3510395
Augustenstraße 62, D-80333 München
3
Table of Contents
Preface 5 Abstract 6 1. Introduction 7 2. Theoretical Framework 102.1 Pro-Environmental blaming behavior - A conceptualization 10 2.2 Individually perceived environmental impact and its link to feedback 12 2.3 The effects of feedback on subsequent blaming behaviors 12
2.3.1 Feedback or the retirement from routine 12
2.3.2 Behavioral spillover: Feedback as alarm clock, feedback as mirror 13 2.3.3 Recycle, stand up, speak up: Feedback as catapult of psychological standing 15
2.4 The moderating effect of feedback privacy levels 18
2.4.1 The role of emotions 18
2.4.2 Conformity and Social Approval 19
2.5 Conceptual Model 20
3. Methodology 20
3.1 Research Design and participants 20
3.2 Procedure and manipulations of variables 21
3.3 Measures 25
3.4 Data Analysis Plan 26
4. Results 27 4.1 Overview 27 4.2 Scale Development 29 4.2.1 Correlations 29 4.2.2 Factor Analysis 30 4.2.3 Reliability Analysis 31 4.3 Hypothesis testing 32 4.4 Results summary 34 4.5 Additional analysis 34
4.5.1 Confounds and reasons why 34
4.5.2 Feedback and emotions: Three ANOVAs 35
4.5.3 Guilt within a Mediation Analysis 37
4
5.1 Main contributions 38
5.1.1 Causing harm to evoke good? Moral cleansing and goal-striving behaviors 38 5.1.2 Resignation or confrontation: Normative conformity and guilt 39 5.1.3. Psychological standing: Be moral to engage or engage to be moral? 40 5.1.4 I act, so I am: Behavioral spillover and self-perception 41
5
Preface
When I was a child, I found it hard to pause once I started reading. I was intrigued by the way authors set up sentences, created stories after stories and got me hooked right away. Ever since, I have had the greatest respect for words and their impact - when being used wisely. This respect turned into a desire which led me further to investigate feedback messages more closely. To find more evidence on how they affect public blaming behaviors and engagement in environmental action. An area where marketing is strongly required and actually can be used to do good. Where the true potential of marketing can be unleashed: To aid in tackling one of the most burdensome issue in the world. In times when doing good has potentially started to turn into a trend. In times of Greta Thunberg1, when the adults of tomorrow are starting big movements by themselves - it is our responsibility to act on opportunities. To respond to the rising needs of our planet and take collective action. To increase the scope of environmentally friendly behaviors and contribute to the change we want to see in the world. And this is where my research aims to dock on to.
At this point, I would like to express my deepest gratitude to all those who contributed to this work. Foremost, I would like to thank Dr. Mathilde van Dijk for all the effort, the constructive feedback and the goodwill (and sustainable mindset) of supervising me via Skype. For providing me with structure throughout the whole time, being there for friendly, supportive advice and granting me an encompassing, valuable learning opportunity. I would also like to thank my second supervisor, Dr. Sumaya Albalooshi for reading and reviewing my work. Another thank you goes to my boyfriend Sebastian, who constantly encouraged and challenged my work. My family, who were the first to test my survey on and who allowed me interesting insights into respondent behavior. My friends, who filled in my survey and helped me spreading it in order to acquire as many participants as possible. And finally, everyone involved, for engaging in fundamental discussions with me and shedding new light on different viewpoints about the environment and human nature.
1 Greta Thunberg is a 16-year old Swedish environmental activist who has recently initiated the worldwide
6
Abstract
General awareness for climate change is widely spread. Yet, too few public actions are taken to counteract environmental issues on a larger scale. This research proposes a way of how such civic engagement can be reinforced: By providing humans with feedback about their past environmental behaviors. Hence, this research investigates how an individual’s environmental self-perception, derived from such feedback, affects subsequent engagement in publicly blaming companies for their environmental misdeeds. Further examined is a potential moderating effect of receiving such feedback in public versus in private. The effect was tested within the frame of a survey experiment with 211 respondents from various countries and demographics. Albeit no statistical significance for the main hypotheses was yielded, an additional analysis suggested significantly higher levels of guilt among respondents who had received negative or public feedback. Feelings of guilt were further found to significantly predict blaming behavior. With these findings, this research adds pieces of clarity on the direction into which factors can increase public action. Further research is needed to examine the role of personality differences and emotions in using feedback as part of a campaign strategy to induce blaming behaviors.
7
1. Introduction
8
16-year old Swedish activist Greta Thunberg, on the other hand, stood up for the environment. She recently initiated a worldwide movement against climate change: Fridays for Future. It encourages students to not attend school on Fridays, and instead, demonstrate against climate change. The main reasoning behind this is: Why going to school to learn for the future whereas there might not be a future? She has been challenging politicians and opinion leaders, with a clear aim: to preserve the planet and achieve long-term, sustainable change in human environmental behavior. (Fridays For Future, 2018)
Where do her willpower to act and strength to proceed originate from? What makes her stand out from millions of people who only think positively about the environment; those, who intend to change their behavior in a favorable manner but do not actually do so? One may assume she feels a higher legitimacy to act upon her attitudes, for example through her perceived moral stake in the issue.
This concept of why and when one actually acts upon one’s attitudes and responds to outrage is rooted in the mechanism of psychological standing, investigated by researchers and scholars from different angles. It is, by definition, “the subjective judgment of legitimacy to perform an action”. (Sherf, Tangirala & Weber, 2017) Psychological standing is grounded in different phenomena. It can manifest itself through moralization of the issue at hand, through the existence of a material stake in that matter, through personal relevance and many other forms which will be discussed in more detail in the following chapters. (Miller, Effron & Zak, 2009; Effron & Miller, 2012) This research, however, goes beyond what is so far known and adds a new root of psychological standing, namely feedback. It investigates if and how receiving positive feedback about one’s private, everyday environmental behavior can provide an individual with psychological standing. – If it is through moralization, through personal relevance or through other, yet unknown, mechanisms. Taking it one step further, it aims to detect how such feedback can then increase subsequent public action, namely blaming companies for their misdeeds.
9
environmental conscious perception of the self. Viewing oneself as a person who cares about the environment results in increasing subsequent environmental private and public action. (Bem, 1972; Lanzini and Thøgersen, 2014; Lauren et al., 2018) Feedback can contribute to this mechanism by exerting a “mirror”-function: Receiving feedback about one’s environmental behavior is assumed to raise awareness for past, eco-friendly actions and activate subsequent cognitive processes. This, in turn, can lead to a moral, pro-environmental view of the self and result in a boost in subsequent public action.
Now, when scrutinizing feedback itself in its effect on subsequent blaming behaviors, it is important to consider the context it is received in. This leads to a crucial further investigation of the role of receiving this feedback in public versus in private. Within this context, normative conformity and the importance of maintaining status within a group serve as explanatory mechanisms. According to the concept of normative conformity, humans often adapt their behavior to match with the one of the group norm. In order to protect or enhance their self-esteem, they long for social approval from their peers. (Cialdini & Goldstein, 2004) Now, to what extent does the need for social approval affect the effectiveness of public feedback on social action? Is public, positive feedback itself enough to gain social approval and, hence, mitigates the need for further public sphere action? So, in order to reach higher participation levels, should private feedback then be the preferred measure? Hence, in addition to the effect of feedback on behavior, also the type of feedback has the power to contribute on literature. – On normative conformity and the power of social approval.
Results gained by this study allow Marketers to improve the creation of Social Marketing campaigns calling out for environmental action. If receiving positive feedback influences psychological standing, persuasive messages can be formulated by linking with such feedback. By this means, people who hold a certain belief can be given the legitimacy to act on the attitude they already have. (Miller, Effron & Zak, 2009) If private feedback is more influential than public one, different channels can be chosen for the implementation of such campaigns.
To sum it up - In present times, this research is relevant more than ever. To derive implications for sustainable change and to empower humans to act on their attitudes, like Greta Thunberg. To act now, while there is still time to prevent the worst from happening.
10
mechanisms will then be highlighted. Furthermore, psychological standing and its different predictors rooted in various literature will be presented. In addition to that, the concept of normative conformity and social approval will be highlighted as well as the role of emotions. Given all information together, the main hypotheses will be derived. The methodology chapter will provide insights about research design, procedures, measures and an analysis plan. It will give an insight into how feedback as such and the different conditions were manipulated. It will be followed by an extensive overview over all generated results before concluding by providing a general review over the whole study, coming up with limitations, managerial implications, suggestions for future research and a general conclusion.
2. Theoretical Framework
The following chapter provides an overarching framework over the context this research is embedded in and is rounded off by two established hypotheses, visualized by a conceptual model. At first, the dependent variable “blaming behavior” will be reviewed more closely. This is followed by a thorough depiction of feedback and its manifestation in the independent variable, the individually perceived environmental impact. The main topic of interest is the question of how to increase levels of the dependent variable, blaming behavior. For this reason, the potential effect of the individually perceived environmental impact on blaming behavior and the reasoning behind is outlined. Hereinafter, the moderating effect of privacy levels on the relationship between individually perceived environmental impact and blaming behavior is delineated.
2.1 Pro-Environmental blaming behavior - A conceptualization
11
this research docks on to, pro environmental blaming behavior will be classified and explained further.
Environmental blaming behavior is embedded in the multidimensional and diverse field of pro environmental behavior (Brick & Lewis, 2016). Pro environmental behavior can be further classified as heterogeneous, with an important distinction made between private and public-sphere behaviors (Ertz, Karakas, & Sarigöllü, 2016).
Private sphere behaviors, on the one hand, include all those actions an individual performs him or herself which directly exert impact on the environment; for instance, recycling, waste separation, the act of commuting by public transport or meat consumption. (Stern, 2000) Even though these behaviors are crucial in preserving the planet, the variable of interest, blaming behavior, is represented by the other group, namely public sphere behaviors.
Public sphere behaviors, on the other hand, are associated with different forms of environmental activism, such as engaging in boycotts, public protests, demonstrations, creating, signing or sharing petitions. They often involve blaming of companies or political institutions. (Hamilton et al., 1988; Skarlicki et al., 2017). The main difference over private sphere behaviors is that public sphere behaviors encourage pro environmental behavior on a societal, rather than an individual level. In addition, the influence on preserving the planet is rather indirect. Yet, it is often due to ego depletion, a lack of opportunities, contextual particularities or differences in motivation that engaging in such civic activism comes with lower participation levels. However, those fewer people that, nonetheless, decide to protest can, in contrast, reach many more people and affect their behaviors at once. Hence, in order to evoke long-term, sustainable change in the mindset and according behavior of society, public sphere behavior is crucial. (Lauren, Smith, Louis & Dean, 2018; Stern, 2000)
12
2.2 Individually perceived environmental impact and its link to feedback
Now, that the dependent variable has been closely investigated, the predictor variable, individually perceived environmental impact, will be expounded as a next step.In this research, feedback on the own, private-sphere environmental behavior will serve as a reflector of the independent variable, namely how one subjectively perceives her or his impact on the environment. In other words, if one receives positive feedback about his or her environmental behavior, it is assumed that this person perceives him or herself as eco-friendly.
This inference is made, firstly, due to the high transferability of feedback into practical contexts, such as the creation of environmental campaigns or new carbon footprint calculators. Secondly, feedback has proven to accompany powerful, strong effects on subsequent behavior, even beyond the area of environmental research (Hattie & Timperley, 2007).
The actual power of feedback and its potential effects on the dependent variable blaming behavior will be explored more in the following section.
2.3 The effects of feedback on subsequent blaming behaviors
Now, that both blaming behavior as the dependent and perceived environmental impact, reflected by feedback as the independent variable have been depicted, it is investigated how they are interconnected. For this purpose, three main mechanisms are presented which reveal ways of how feedback can affect according blaming behavior.
2.3.1 Feedback or the retirement from routine
Often, what hinders individuals from translating their attitudes into behavior and engaging in more environmental actions, are existing habits or daily routine.
13
Yet, this is not all it takes - In addition to that, motivation is required. Enough motivation is needed to break with the existing routine and stick to exerting the new behavior: Until it is transformed into an integral part of a new, updated routine. Now, one powerful way such motivation can be established on a long-term base is positive affirmation in the form of feedback. By frequently being demonstrated appreciation for one’s good performance in the desired behavior, efforts are increasingly perceived by the individual as “being worth it”. This enhances motivation to proceed, increases perceptual fluency for future pro environmental behaviors and slowly induces transition into this updated, habitual stage. (Bechtel & Churchman, 2003)
In sum, both due to the motivational and the awareness raising function of feedback, it has the power to break with existing routines and increase engagement in blaming behaviors.
2.3.2 Behavioral spillover: Feedback as alarm clock, feedback as mirror
A further way of how feedback can increase according blaming behavior is rooted in the behavioral spillover theory, conceptualized by an increasing number of authors.
In a nutshell, the behavioral spillover theory suggests that the existence of one or few environmental behaviors leads to the adoption of more pro-environmental behaviors. In specific, some authors have found positive spillover effects from private sphere to public sphere behaviors: Encouraging private sphere behaviors verbally or monetarily raised according engagement levels in public sphere behaviors. (Lanzini and Thøgersen, 2014; Lauren et al., 2018; Truelove, Yeung, Carrico, Gillis, & Raimi, 2016) Transferred to this research context, receiving positive feedback about one’s previous private sphere environmental impact acts as a reminder or mirror of past behaviors and, by this, raises the likelihood to engage in future public sphere pro environmental blaming. This provides proof on the effect of feedback on blaming behaviors.
In the following, there will be presented several mechanisms expounding the why and the how of this theory.
Principle of consistency
14
(1966) added an important contribution based on this consistency: Once committing to a small request and performing the according behavior, the likelihood of complying with ensuing, larger requests increases. (Truelove, Carrico, Weber, Raimi & Vandenbergh, 2014) Feedback, in this context, acts as a reminder or “alarm clock” to the individual that the smaller scale behavior had taken place.
Hence, the principle of consistency supports the assumption that feedback on private sphere behaviors with a generally smaller reach mirrors individuals their good deeds and, in turn, has the power to subsequently increase the probability of engaging in public sphere behaviors.
Self-perception theory
Another explanation for the behavioral spillover theory is rooted in the, so called, self-perception theory which is based on the assumption that an individual’s self-perception influences the spillover of past and future behaviors (Lauren et al., 2018).
According to the self-perception theory, individuals constantly derive their attitudes, beliefs, and other internal states from observing their behavior (Bem, 1972). Thus, by observing the times they engaged in pro environmental behaviors in the past, they will subsequently view themselves as being a person that cares about the environment. Hence, by providing respondents with positive feedback regarding their past private sphere behaviors, they are being mirrored their good deeds and encouraged for further observation. Due to this reminder, such positive self-perceptions are activated and subsequently act as a facilitator of future, other environmental behaviors. Thereupon, people will more likely engage in behaviors in line with their self-perceptions of an environmentally friendly human and, by that, the probability to perform public sphere blaming behaviors in the future is increased. (Lauren et al., 2018) This insight provides further evidence for a positive effect of the independent variable feedback on the dependent variable blaming behaviors.
Learning Theories
self-15
confidence and perceptual fluency levels to proceed with similar, future behaviors. (Lanzini & Thøgersen, 2014)
Now, what is the role of feedback in this context? Such a learning-by-doing process can be created by individuals exerting certain environmental behaviors and receiving according feedback on their performance. This subsequently increases awareness, knowledge and skills on environmental issues and leads to a boost in self-confidence. That way, fluency to continue with pro environmental behaviors in the future is created. (Lanzini & Thøgersen, 2014) Hence, learning theories provide another, powerful implication on how feedback on past environmental behavior can affect future eco-friendly engagement.
Summing it up, all three mechanisms have proven how feedback can be used to create a spillover effect between previous and future environmental behaviors. By this, the effect on pro environmental blaming behavior is confirmed.
2.3.3 Recycle, stand up, speak up: Feedback as catapult of psychological standing
One further mechanism which portrays how receiving positive feedback can increase blaming behaviors, is the construct of psychological standing. It will be explained in the following.
Individuals not directly invested in the issue often do not engage in public action because of the system’s (political) complexity. They further do not want unnecessary attention to be drawn upon them and, thus, even be target of negative emotions like anger. They feel like their action is not legitimate once it is not directly driven by self-interest. Such uninvested individuals may even see themselves as “having less to gain by taking action”. Hence, among them, the presence of an avoidance motivation is common as well as the saying “It’s not my place”. (Ratner & Miller, 2001) And albeit individuals may feel outrage, this emotion alone is often not enough for them to feel entitled to actually stand up and act on their attitudes. What they lack can be traced back to the feeling of being ‘their place to act’ on that outrage, a justification for doing so. (Miller et al., 2009) And this is what is referred to as psychological standing or the “subjective judgment of legitimacy to perform an action” (Sherf et al., 2017).
16
perceptions of legitimacy to stand up and blame companies for their misdeeds. - And how feedback can serve in enabling psychological standing.
As portrayed, one determinant of psychological standing is the degree to which one is personally invested in the topic, for example, through one’s personal characteristics, relevance, history or experiences with the issue at hand.
Additionally, perceived legitimacy to act can be increased if sufficient grounds are available to support one’s blames with.
Furthermore, feelings and attitudes will likely translate into blaming behavior if the extent to which one is materially affected by the issue is high. (Miller et al., 2009; Effron & Miller, 2012) In support for the next antecedent, it is essential to understand that, to most people, morality is key to create and maintain a positive self-concept and identity. Viewing themselves as moral and fair is crucial while self-interest impairs such moral norm. (Oc, Bashshur & Moore; 2015) In addition to that, several researchers have highlighted the motivational power of possessing moral identity. They found moral judgements and identity to result in subsequent moral behavior. (Reynolds & Ceranic, 2007) Hence, even individuals who do not possess personal relevance, direct interest or material stakes in the issue, will nonetheless act on issues if they perceive them as morally reprehensible. Thus, in order to increase civic engagement, moralization of the issue at hand has to take place and, by this, makes up another powerful antecedent of psychological standing. (Miller et al., 2009; Effron & Miller, 2012)
Here is where feedback comes into play: Feedback about a strikingly moral topic, namely individuals’ environmental behavior, increases awareness about the moral character of pro environmental blaming behavior. By this, it constitutes the described moralizing function of the issue at hand. Following the above logic, this activates individuals’ moral self-identity, strengthens psychological standing and, hence, intensifies the perceived legitimacy to act. It transforms the issue from only those financially or otherwise invested have a say in into a matter that everyone is entitled to stand up for. (Miller et al., 2009; Effron & Miller, 2012)
17
One other antecedent feedback is linked with are personal characteristics. As previously proven, receiving positive feedback about one’s past environmental behavior can evoke or strengthen an environmentally friendly self-image. This, in turn, can increase the perceived entitlement to engage in pro environmental blaming behaviors (Miller et al., 2009).
The table below illustrates existing research about psychological standing and its antecedents and pictures how this study complements previous findings (Table 1).
Table 1: Antecedents of psychological standing
Authors Relevant findings Antecedents Sherf, Tangirala
& Weber (2017)
Men reported lower psychological standing than women regarding gender-parity initiatives. This, in turn, led to lower participation levels among men.
Directly invested in the issue due to a demographic characteristic (gender)
Miller, Effron & Zak (2009)
Experiencing outrage alone does not
necessarily lead to acting upon one’s values. People further need to feel entitled to engage in protest.
This research defines different determinants of psychological standing.
Material stake, Personal cost Personal characteristic, history, relevance
Other forms of self-interested stake
Availability of sufficient grounds
Moralization
Ratner & Miller (2001)
Individuals tend to not engage in public sphere behaviors unless they are manifested in self-interest. This is due to the
complexity of public action and the perception of having “less to gain” by undertaking action.
Self-interest
Effron & Miller (2015)
Linked with psychological standing: Advisors who had suffered for their misdeeds in the past were viewed as more legitimized to give advice and felt more self-righteous to do so.
Past (negative) experience with the issue at hand
Effron & Miller (2012)
The more participants related
sociopolitical issues with moral values, the higher their indicated willingness was to act upon those issues. Material stake and moralization were found to be
interchangeable factors in explaining what drives individuals to act upon their values.
Material stake Moralization
Current research Assumption:
Granting individuals with psychological standing through positive environmental feedback. This, in turn, increases
subsequent public action, namely blaming. behaviors. The effect is moderated by feedback privacy levels.
18
Concludingly, various theories on feedback, motivation, behavioral spillover and psychological standing demonstrate ways on how positive feedback is able to induce pro environmental blaming behavior. Hence, taking all explained mechanisms and reasonings together, the following hypothesis can now be derived:
H1: A positively (relative to negatively) individually perceived environmental impact increases blaming behaviors towards companies for their environmental misdeeds.
2.4 The moderating effect of feedback privacy levels
Receiving feedback on one’s past environmental behaviors in public versus privately may affect the relationship between feedback and blaming behaviors. In the following, it will be depicted why this effect may influence the above relationship and how it is assumed to work.
2.4.1 The role of emotions
Emotions were found to be felt stronger when receiving feedback in public relative to private settings. (Belschak & Den Hartog, 2009) The general research stream on emotions suggests that positive feedback will likely result in feelings of pride and satisfaction. However, when receiving such feedback in public, both positive feelings plus feelings of embarrassment will arise or be enhanced. This can cause emotional ambivalence (Coulter & Pinto, 1995) and, thus, may diminish the effect of feedback on subsequent blaming behaviors.
Negative feedback, on the other hand, will induce negative emotions such as disappointment or anger, leading to absenteeism or negative turnover behaviors. (Lazarus, 1991) Simultaneously, receiving negative feedback in public will reinforce feelings of disappointment or anger which, in turn, pose a threat to one’s self-concept and will lead to lowering one’s efforts towards future behaviors, such as blaming behaviors. (Belschak & Den Hartog, 2009)
19
2.4.2 Conformity and Social Approval
Another mechanism complementing this finding is rooted in the principle of conformity. It encompasses the “act of changing one’s behavior to match the responses of others.” In order to enhance, protect or repair their self-esteem, individuals constantly strive for a positive self-concept. One way of enhancing their self-concept is normative conformity, namely conforming to behaviors of the group norm. It is driven by the desire to be accepted, engage in meaningful relationships and gain social approval from others. (Cialdini & Goldstein, 2004)
Following this logic, positive feedback received in public grants one social approval by peer groups and, hence, maintains and enhances the individual’s positive self-concept. Hence, social approval is already achieved and further engaging in public sphere behavior will not lead to any additional public acknowledgment. Differently put, once having received positive affirmation for one’s pro environmental behavior, the probability to engage in subsequent public action will be mitigated. In contrast, receiving private feedback is assumed to have the opposite effect. After being granted feedback privately, one’s behavior is yet not approved by his or her peer groups. Hence, it is assumed that the urge to gain approval by engaging in blaming behaviors which are commonly used as a strategy to establish status within a group, is higher than in the previous condition.
Taken these mechanisms together, strong support is yielded for receiving feedback in private in order to increase public action, for both negative and positive feedback conditions. Thus, it is hypothesized:
20
2.5 Conceptual Model
The following conceptual model visualizes both above stated hypothesis and constitutes the very foundation for the ongoing research (Figure 1).
Figure 1: Conceptual Model
3. Methodology
3.1 Research Design and participants
The goal of this research is to measure the effect of different types of feedback on subsequent blaming behavior, moderated by a public or private setting in which this feedback is received. For this reason, a 2 (valence of feedback, serving as the independent variable) x 2 (public versus private feedback, serving as the moderator variable) factorial, between-subjects design was set up. The dependent variable, blaming behavior, was put into practice by confronting respondents with a hypothetical scenario in which they were asked to indicate their willingness and likelihood to sign such a petition.
21
feasible by the online software Qualtrics which is an efficient way to spread surveys via online communication channels. Since it allows participants to fill in the surveys at their own convenience, the online experiment was chosen as the preferred type of experiment. Plus, it made it possible to question people from different demographics and regions all across the world within a relatively short period of time. Since several questions posed in the survey are of highly moral character, the sensitivity of respondents towards such topics is assumed as high. Hence, social desirability bias of the respondent is diminished when the experiment is conducted online, compared to offline (Duffy, Smith, Terhanian & Bremer, 2005).
In total, 160 valid responses were generated. Participants were acquired mainly by means of my personal network while applying the non-probability snowball sampling technique. This method was chosen due to its generally low sampling variance and costs. Furthermore, with this technique, it is more probable of “locating the desired characteristic in the population”. (Malhotra, Hall, Shaw & Oppenheim, 2006) Participants were reached in person, via WhatsApp and Facebook. In addition to that, some respondents were collected by means of LinkedIn- and Facebook- Research Groups. The data was collected in April 2019.
3.2 Procedure and manipulations of variables
At first, individuals were provided with important information about the experiment such as anonymity of data, retraction without consequences and the necessity to answer all questions until one has reached the very end. After having indicated their consent, participants were randomly assigned to one of the four conditions.
At first, they were posed a form of control variable questions namely gender, age, education level and income. Due to the fact that income is widely perceived as a sensitive information, there was an option of skipping the question in which the exact amount was asked for. However, all respondents were still obligated to indicate whether they lived comfortably with their current income. Those control variables were placed at the beginning in order to establish familiarity towards the research and prepare respondents for subsequent questions of a more pervasive character.
22
their carbon footprint2 had to be calculated. As a precondition for the feedback manipulation they were soon to receive, they were asked to fill in a questionnaire about their private sphere environmental behaviors. Topics included food consumption behaviors, such as meat consumption, acquisition of locally produced food as well as food waste and recycling behaviors. Use of public transport versus the own vehicle as also frequency and distance of air travel were subject of further questions. An exemplary item is portrayed in the appendix (Exhibit 1). Another topic was electricity consumption: Respondents had to indicate how or to how many degrees they heated their homes, or how often they turned off their lights and appliances when not in use. The last questions within that block centered around monetary consumption levels of different product types such as clothing, footwear or beauty products.
After participants had completed the questionnaire, they were told to wait some seconds during their carbon footprint was calculated. A visualization is presented in the appendix (Exhibit 2). As soon as this short time period had passed, participants in the public condition received the message: “Your answers stand out! Imagine receiving your result during a public radio interview”. This message was phrased alternately and repeated, in order to ensure the participants did not overlook this for the experiment crucial information. For illustration purposes, a graphic underlining the scenario was incorporated. In order to view the full instruction, consult the appendix (Exhibit 3).
For participants in the private condition, on the other hand, it was not necessary to be provided with an additional scenario since privacy of feedback was assumed to serve as the status quo. Instead, they were again reminded of the anonymity of data during the time they waited for their results to load.
By means of a bogus carbon footprint calculator, they were subsequently provided with randomized feedback as a manipulation. Hence, unbeknownst to the participants at that point, this feedback did not reflect their real private sphere environmental behaviors and was not based on their original answers. Thus, the result they received was twofold: Either respondents were a major carbon contributor, or they were not.
2Wiedmann & Minx (2008) have thoroughly reviewed common terminologies of the carbon footprint to come to
23
Participants in the “Positive feedback” condition were provided with the statement: “Based on your individual behaviors, you are not a major carbon contributor”, graphically highlighted (Exhibit 4).
Respondents assigned to the “Negative Feedback” condition were led to believe the opposite: “Based on your individual behaviors, you are a major carbon contributor.” which was again visually portrayed (Exhibit 5). In order to add additional power to the manipulation, another statement was added: Participants were told how many earths would be needed if everyone behaved like them which is another common measure for calculating one’s carbon footprint. (Toner, Gan & Leary, 2014)
With this feedback, each respondent now had a perception of her or his individual environmental impact. This built the ground for the independent variable and the subsequent procedure. Before continuing with the next scenario, participants in the public condition were asked to fill in a short manipulation check question, namely to indicate if it was true or false that everyone participating in the radio interview had heard the respondent’s individual feedback (Exhibit 6). This was a necessary step in order to control if participants had understood the scenario they had been confronted with. Participants in the private condition were presumed to be aware of the fact that their result was conveyed in private since they had not been instructed otherwise. Thus, it was decided to forgo a manipulation check and, hence, avoid any redundant confusions.
Next, in order to test the effects of the feedback received on the dependent variable, namely blaming behavior, the last part of the survey was conducted. Since participants were about to be opposed with another scenario which required mental effort to read and understand, they were at first given a quick attention check question. They were specifically asked to forgo several multiple choice options and instead enter the present day’s date below. This attention check question served as a directed query and was used due to its facility of understanding, its high objectivity and down-to-earth implementation (Abbey & Meloy, 2017).
24
climate change. In addition, an eye-catching flyer was crafted and added which summarized the main message and highlighted the case’s exigence (Exhibit 7). Below, the text “You are now well informed on the topic of their misbehavior regarding climate change” was included. This was highlighted in order to reassure that lack of information about the topic at hand was not a factor which had to be further controlled for.
In the following, participants were asked to indicate their willingness to sign. In order to test the underlying mechanisms that explain the independent variable, questions referring to psychological standing, moralization, self-perception and behavioral spillover were posed. Thereafter, it was controlled for existing public sphere behaviors of participants. Inclusion of this variable follows the logic that among those who already engage in public sphere behaviors on a frequent basis, behavioral spillover had already taken place and, hence, has the power to mitigate or even eliminate the effect of feedback on signing petitions. Next, another variable controlled for was perceived power: Several studies demonstrate the positive effect of perceived environmental efficiency (PEC) on pro environmental engagement levels, in order to advance the social common good. (Lee, Kim, Kim & Choi, 2014) Hence, the lack of belief that petitions as such can combat climate change may influence respondents’ willingness to sign the petition. Familiarity with the “real” carbon footprint survey on the website of WWF UK could have further created awareness bias and, thus, affected results. Hence, it had to be controlled for as well. Finally, intrinsic normative goals of participants were tested. They may possibly had suspected that, being willing to sign the petition had resembled the norm and was viewed as “the right thing to do” (Steg, Bolderdijk, Keizer & Perlaviciute, 2014). Due to their desire to comply with the norms, they may had answered more in favor of petitions than they would have in real life settings.
As the final block, manipulation checks about the positive and negative feedback condition were placed. Respondents were asked about their feelings about their carbon footprint results, including guilt, regret, surprise and satisfaction. These questions were set up in the end in order to give participants time to process those thoughts and feelings more and, hence, increase precision.
25
that their anonymity of data was still reassured. In the end, they had the option to participate in a lottery and insert their email addresses.
3.3 Measures
Several studies provide evidence that acting environmentally friendly is commonly stereotyped with higher degrees of femininity. This often influences men’s willingness on adopting more pro environmental behaviors. (Wilkie, Jingjing, Isaac, & Gal, 2016) Hence, gender differences could impact the dependent variable, namely the likelihood to sign petitions and blame companies for their misdeeds. The options “Male”, “Female” and “other/Don’t want to say” were given.
Models such as the Needs-Opportunity-Ability model of consumer behavior present the mere presence of opportunities, as well as possessing enough financial, temporal, cognitive and physical abilities as decisive factors in influencing levels of environmental engagement. (Howell, 2014) Due to this, age, education and income levels were tested by means of self-reported and multiple-choice questions.
Self-reported and multiple-choice questions on the participants’ private-sphere environmental behaviors were partly adopted and adapted from the WWF UK carbon footprint survey. (WWF UK, n.d.) These measures of private-sphere environmental behavior are further supported by various literature who used similar items for their studies (Kaiser & Wilson, 2004; Lanzini, & Thøgersen, 2014).
26
environmentally conscious. Response scales were based on a 7-point Likert scale and ranged from 1 = not at all to 7 = very much.
Measures of perceived consumer efficiency used in this study are rooted in items of the PCE scale commonly used in literature. (Lee et al, 2014) For this purpose, three items measured the perceived efficiency of petitions as such in combating climate change. One further item was included in order for consumers to rate the effectiveness of different forms of environmental activism. This was aimed to improve understanding of consumers’ efficiency perceptions of petitions. For all items, a 7-point Likert scale (1 = not at all, 7 = very much) was used.
In concordance with intended public sphere behaviors, existing public sphere behaviors were based on the environmental activist behavior (intention) scale. (Lee et al., 2014) Included were two items containing frequency of previous petitioning behaviors and frequency of other forms of environmental activism. Both items were measured on a 5-point Likert scale (1 = never, 2 = almost never, 3 = 1-3 times a year, 4 = 4-10 times a year, 5 = at least every month).
Measures of psychological standing were derived from Effron & Miller (2012), precisely from their fourth study, asking participants about perceptions of legitimacy and appropriateness on a 7-point Likert scale (1 = not at all, 7 = very much). Measures of moral identity were based on a summarized version of the Self-Importance of Moral Identity Measure according to Aquino & Reed (2002).
Since engagement pro environmental behavior can be influenced by normative goals or the desire to comply with the norm, one further item was included (Stern, 2000). This item encompassed the question: “Be honest with yourself - do you think the topic of this survey has affected your choice to sign the petition?”. It was measured on a 7-point Likert scale (1 = not at all, 7 = very much).
3.4 Data Analysis Plan
In the following, an exploratory factor analysis will be conducted, in order to reduce dimensions of control variables and items explaining underlying constructs. After factors have been determined, they will be tested on their reliability, by viewing their Cronbach’s Alpha.
27
group coded as dummy variables and blaming behavior as the dependent variable will be carried out.
Within the frame of an additional analysis, at first, different variables will be tested for potential covarying effects. Secondly. separate ANOVAs will be conducted, in order to detect the role of emotions across conditions. Thereafter, a simple mediation analysis with the variable guilt as a potential mediator will be conducted.
4. Results
4.1 Overview
In total, 211 respondents participated in the experiment out of which 161 completed the whole survey and 3 only missed the last few questions. Hence, 164 respondents were decided to be included at first. Next, 4 more respondents were excluded who had failed more than one attention or manipulation check or did not demonstrate coherency regarding remaining results. This led to a total of 160 respondents for subsequent analysis.
Out of the 160 respondents, 58.8% were female and 40.6% male3. While participants were between 16 and 65 years old, the average age was 28 (27.75) years old, hence slightly younger than the average population. An improved visualization of the sample distribution among the variable age is presented in the appendix (Exhibit 8). 42.5% had a graduate degree, 35.0% an undergraduate degree and 15.6% had their high school diploma. The mean level of education was 2.9, meaning an undergraduate degree. This leads to the general assumption that all participants had sufficient cognitive abilities and English proficiency to properly understand the survey. In addition, the mean of respondents expressed living comfortably or coping on their current income which is represented by a horizontal bar chart in the appendix (Exhibit 9). In addition, more details on those demographics are portrayed in the following table (Table 2).
28
Table 2: Descriptive Statistics
Variable Mean
Gender Count Percentage
Male (=1) 65 40.6
1.60
Female (=2) 94 58.8
Age Count Percentage
27.75 16 to 24 63 39.5 25 to 34 78 48.8 35 to 44 9 5.7 45 to 54 4 2.5 55 to 65 6 3.7
Education Count Percentage
Doctorate degree*(=1) 4 2.5
2.90
Graduate degree ** (=2) 68 42.5 Undergraduate degree***(=3) 56 35.0 “Abitur”/High school (=4) 25 15.6 Less than high school (=5) 7 4.4
Income Feelings Count Percentage
Living comfortably (=1) 96 60.0
1.52
Coping (=2) 47 29.4 Finding it difficult (=3) 15 9.4 Finding it very difficult (=4) 2 1.3 *(PhD/MD/Other) (=1)
**(MA/MSc/MPhil/“Staatsexamen4“/Other) (=2)
***(BA/BSc/Other)
Respondents from 22 different countries across the world took part in the research. 56.9% of participants reside in a German speaking country and 13.8% in one of the BeNeLux countries. The remaining 29.4% are diverse. This worldwide sample spread is visualized in the appendix (Exhibit 10).
The carbon footprint survey which each participant filled in at the beginning, served as a profound description of general environmental behaviors and greenness of the sample. It yielded, depending on the topic, below or above average levels of green, pro environmental behavior. Regarding commuting behaviors, respondents demonstrated higher levels of greenness whereas the majority of participants described their diet as “meat in most meals” or “meat in some meals”.
29
A more thorough representation of respondents’ pro environmental behaviors is depicted in the appendix (Exhibit 11).
4.2 Scale Development
4.2.1 Correlations
As a next step, a correlation matrix (Bivariate Pearson Correlation) was designed, in order to get a first feeling for the variables and the relationship between them (Table 3). Two strong and several moderate and weak correlations could be detected. No association was stronger than 0.9 and, hence, no probability of multicollinearity existed.
Positively associated were one’s perceived appropriateness to sign and the dependent variable, namely the willingness to sign. This already provides a first indication of supporting the concept of psychological standing.
Table 3: Pearson Correlations among willingness to sign and underlying constructs
1 2 3 4 5 6 7 1 DV Willingness to sign - 2 Perceived efficacy .473** - 3 Environmental self-perception .319** .195* - 4 Moral self-image .322* * .163* .571* * - 5 Appropriateness to sign .624** .400* * .366** .379* * - 6 Guilt .197* .157* .088 .069 .083 - 7 Existing public sphere behaviors .323** .332* * .403** .230* * .302* * .095 - ** Correlation is significant at the 0.01 level (2-tailed).
* Correlation is significant at the 0.05 level (2-tailed).
In addition, moral self-image was moderately correlated with psychological standing, providing a glimpse of evidence for moralization as an antecedent of standing.
30
The control variables perceived efficacy and existing public sphere behaviors yielded moderate correlations with the dependent variable which means that, in further parts of this chapter, these factors will have to be tested on their covarying effect on the dependent variable.
A weak significant correlation between guilt and willingness to sign the petition was detected which will be elaborated more in the last part of this chapter, the additional analysis.
4.2.2 Factor Analysis
Due to high intercorrelations between items of different constructs of interest (environmental self-perception, moral self-image, appropriateness to sign), three separate factor analysis were conducted.
In the first factor analysis, items related to psychological standing and other items associated with perceived power of petitions as such were inspected. Now, in order to test if the data was suitable for a factor analysis after all, 3 checks were conducted. The KMO measure of sampling adequacy resulted in a meritorious value of 0.822 and, thus, indicated that data were likely to factor well. In addition, the Bartlett’s test of sphericity indicated a significance level of .00 and so, H0, that all variables were uncorrelated, could successfully be rejected. As a last assumption considered, communalities were examined. Around 76%, so more than 60% of variance in each variable were explained by all extracted factors which made all items eligible for factor analysis. This led to the conduction of a Principal Component Analysis. The aim hereby was to determine the optimal number of factors of all variables used for the analysis. A total of 2 components yielded higher eigenvalues than 1, explained more than 60% of variance and represented an inflection point of the scree plot which suggested 2 or 3 factors. Using Varimax, the component matrix was orthogonally rotated and represented factor loadings with each higher than 0.7 loading on the retained 2 factors.
31
After having confirmed the eligibility for the next factor analysis (KMO: .813, Bartlett’s test of sphericity: .000, communalities >6) with items reflecting efficacy and environmental self-perception, a PCA was conducted. Scree plot, eigenvalues and accumulated percentage of variance (77.5%) indicated 2 factors as an optimal number of factors which were used in the following. Rotated component loadings were no lower than .784.
Presented below are all established factors including their loadings (Table 4). On a side note and relevant for the discussion, as an indication of how high, on average, each factor was rated, factor means are presented in the appendix (Exhibit 12).
Table 4: Rotated component loadings Perceived efficacy Psycholog ical standing Moral self-image Environme ntal self-perception To what extent do you think petitions are an
effective way to combat climate change? .903 To what extent do you think that, by signing this petition, you can make a change? .866 How effective do you rate different types of environmental activism? - Online Petition .856 To what extent do you think petition like
this can be successful? .827
How effective do you rate different types of environmental activism? - Paper Petition .786 To what extent do you feel that you have the
right to sign this petition? .917
To what extent do you feel that it is
appropriate for you to sign this petition? .824 To what extent would you consider yourself
a moral person? .910
To what extent is acting morally important
to you? .900
To what extent would you consider yourself
environmentally conscious? .936
To what extent do you think you are a
person that cares about the environment? .895
N=160
4.2.3 Reliability Analysis
Established components
32
was ensured. Their α could not have been improved by deleting any items. Ergo, for each of the factors, their variables are internally consistent and underly the same construct.
Dependent Variable
In order to make one factor out of the dependent variable, a reliability analysis for the items “signing_likelihood” and “signing_willingness” was conducted. It resulted in a high score (α=0.873) and, hence, demonstrated internal consistency. Thus, those items were summed up into one factor, by taking their average score into account. This new factor’s (“DV_Blaming-behavior”) mean was 4.50, indicating a slightly above average tendency to sign the petition, all groups taken together.
4.3 Hypothesis testing
Now, in order to compare means between the groups “Positive versus negative feedback” and “Private versus public group”, a two-way ANOVA was carried out.
Hereby, descriptive statistics provided first results. This is reflected by the following table which presents overall means across groups (Table 5).
Table 5: Dependent Variable: Blaming Behavior (Overall means)
Condition Mean Negative Feedback 4.37 Positive Feedback 4.28 Group Private 4.40 Group Public 4.24 N=160
As it can be identified, means did not differ substantially from each other across groups and did not reflect assumptions stated in the hypothesis of this thesis.
33 Figure 2: Between-group means
Heights of the different bars almost resembled each other and, again, demonstrated non-significant differences between the different groups. As a next step, results of the ANOVA were inspected. In line with these findings, no significant effects were detected. The valence of feedback did not significantly affect one’s tendency to sign the presented petition, F (1, 156) = 0.034, p = .855. In addition, group privacy levels did not exert a significant impact on respondents’ willingness to sign, F (1, 156) = 0.149, p = .700. Furthermore, an interaction between feedback and group could not be detected, F (1,156) = 0.161, p = .689.
34
Table 6: DV Blaming behavior
β Feedback (1 = Positive, 0 = Negative) .042 Group (1 = Public, 0 = Private) .106 Observations: 159 R²: .001 Adjusted R²: -.012
4.4 Results summary
Based on the conducted analysis and the above results, it can be concluded that hypothesis 1 and 2 can be rejected. Receiving positive compared with negative feedback does not imply higher willingness to sign petitions and engage in blaming behaviors. In addition, being confronted with such feedback in public versus privately did not yield any additional effects. For a better visualization, consult the hypothesis summary (Table 7):
Table 7: Hypothesis Results.
Hypothesis Formulation Result
H1
Individual perceived environmental impact increases blaming behaviors towards companies for their environmental misdeeds.
Not supported
H2
Feedback privacy levels enhance the effect of one’s perceived environmental impact on according blaming behavior.
Not supported
4.5 Additional analysis
Since both main hypotheses were not supported, as a next step, potential reasons, further explanations or derivations were aimed to detect.
4.5.1 Confounds and reasons why
35
Firstly, it was important to investigate if appropriate theoretical constructs had been used in order to explain the assumed effect. The components Psychological Standing (F (1, 154) = 43.577, p = .000), Environmental Self-Perception (F (1, 154) = 13.631, p = .000) and Moral Self-Image (F (1, 154) = 28.705, p = .000) yielded significant covarying effects on the relationship between Feedback, Group and the dependent variable. This means that the mechanisms psychological standing and its antecedent moralization, as well as the self-perception theory had been used properly in explaining the assumed relationship. They will be elaborated more in the discussion.
Secondly, potentially confounding variables were aimed to detect, in order to be able to expound insignificances of the hypotheses. The component Perceived Efficacy (F(1,154) = 37.827, p = .000) and the variables normative goal (F (1, 154) = 28.874, p = .000), signing frequency (F (1, 154) = 24.408, p = .000) and familiarity with the WWF survey F((1, 153) = 7.324, p = .008) were found to significantly confound the above hypothesized relationship and, thus, can serve as an explanation for their insignificances.
4.5.2 Feedback and emotions: Three ANOVAs
36
Table 8: 2x2 ANOVA with the Dependent Variables Guilt, Regret, Satisfaction
F (1, 155) p-value DV: Guilt Group 8.223 .005 Feedback 57.050 .000 Group*Feedback 0.169 .681 DV: Regret Group 3.749 .055 Feedback 89.965 .000 Group*Feedback 0.611 .436 DV: Satisfaction Group 0.701 .404 Feedback 274.694 .000 Group*Feedback 1.069 .303 N=159
As presented, across all dependent variables, effects of feedback were significant: Negative feedback resulted in higher levels of guilt and regret; positive feedback led to higher satisfaction levels.5 More information on means across different groups can be found in the appendix (Exhibit 13a-c). However, no significant interaction effects of feedback and group were yielded for neither of the three dependent variables.
In addition, only for the dependent variable guilt, there were significant differences between receiving the feedback in public versus private. For this reason, it was of particular interest to take a closer look at guilt. Hence, significant differences between means for the dependent variable guilt, are visualized below (Figure 3).
5 The finding that negative feedback led to significantly increased levels of guilt and regret can additionally serve as
a manipulation check. By this, it is assured that respondents in general read, understood and believed the feedback manipulation, and that the manipulation meets the criteria of internal and external validity.
37
Figure 3: Mean effects on guilt between conditions.
4.5.3 Guilt within a Mediation Analysis
Due to the proven significances across conditions, it is important to scrutinize the variable guilt even more thoroughly and detect, if it potentially served as a mediator in the relationship between the conditions (Feedback, Group) and the dependent variable, blaming behavior.
In respect thereof, a simple mediation analysis was carried out (Hayes, 2017). Results6 revealed the following:
Firstly, a significant effect of feedback on guilt (a) (β = 2.064, S.E. = .280, p = .000) ; and a significant effect of guilt on blaming behavior (b) (β = 0.210, S.E. = .077, p = .007) was detected. However, no significant indirect (c’) (β = -.352, S.E. = .312, p = .261) and direct effects (c) (β = .044, S.E. = .275, p = .874) of feedback on blaming behavior were yielded. Thus, no mediation could be detected.
Secondly, the condition “group” was found to significantly affect guilt (a) (β = -.777, S.E. = .319, p = .016); guilt, in turn, had a significant impact on blaming behavior (b) (β = .179, S.E.
6 For the following analysis, feedback is coded as 1 = Positive, 0 = Negative, and group is coded as 1 = Public, 0 =
38
= .067, p = .009). This time, also no significant indirect (c’) (β = .281, S. E. = .274, p = .307) and direct effects (c) (β = .106, S.E. = .276, p = .700) of group on the dependent variable, were found. Hence, even though no significant mediations could be detected for both variables, significances for a and b each nonetheless illustrate a degree of importance of the received results. What this exactly means will be elaborated more in the next chapter.
5. Discussion
5.1 Main contributions
5.1.1 Causing harm to evoke good? Moral cleansing and goal-striving behaviors
Contrary to the hypothesis, neither positive feedback was found to result in higher engagement levels, nor did negative feedback lead to lower overall petitioning behaviors. Hence, with the obtained results, other mechanisms than assumed must underlie which will be examined in the following.
The finding that negative feedback on the perceived environmental behavior evokes guilt, which, in turn, results in taking increased action in public sphere behaviors is potentially reflected by the effect of moral cleansing, widely discussed by different scholars. There is a broad research stream which has investigated moral cleansing and moral balance theories in general, beyond the field of environmental research. Several concepts and principles have been defined within this context, such as the closely related guilt reduction or the warm glow effect. These mechanisms are highly relevant for implications across domains. (Tiefenbeck, Staake, Roth & Sachs, 2013)
39
moral licensing: Engaging in good deeds acts as a buffer for future moral violations. (Mazar & Zhong, 2010; Miller & Effron, 2010)
Now, the theory of moral cleansing can serve as an explanation of why guilt, evoked through negative feedback, results in higher subsequent positive environmental actions: Through feedback, the internal scoring system is activated and turns negative. Moral identity and self-worth are threatened, resulting in higher levels of guilt. Now, in order to regain pieces of lost self-worth and overcome levels of guilt, subsequent willingness of respondents to engage in pro environmental behaviors increases. (Sachdeva, Iliev & Medin, 2009, Yu, Hu, Hu & Zhou, 2013)
Another reasoning stressing positive outcomes of negative feedback can be viewed as complementary to the theory of moral cleansing. Within the context of work settings, it has been shown that negative feedback enhances task-learning, directs behavior and facilitates goal-striving behaviors. If applied to this research context, it can serve as another explanation why respondents who felt guilty, after the exposure to negative feedback, indicated higher willingness to sign after being exposed to negative feedback. (Belschak & Den Hartog, 2009)
5.1.2 Resignation or confrontation: Normative conformity and guilt
Results partly confirm the phenomenon of normative conformity, mentioned in earlier chapters, as well as the role of public feedback in increasing emotions. As illustrated in the literature review, the intensity of all emotions experienced is enhanced when feedback is given in public, rather than private settings (Belschak & Den Hartog, 2009). This is confirmed by the finding that, in the case of negative feedback, guilt was experienced even stronger in the public condition. Now, instead of the anticipated withdrawal from efforts, as a consequence, the intensified emotion of guilt led to increased compensation efforts in form of higher willingness levels to sign. This finding contributes to the theory of moral cleansing (Mazar & Zhong, 2010) and posits evidence for the power of alternative, positive turnover reactions to guilt, other than resistance and neglect.
40
strategy to make people feel seen by others and signals to others that they are environmentally conscious and moral (Bauman & Mullen, 2014). Hence, in order to repair their self-esteem and achieve or maintain status within a group, people who had received negative feedback in public will even indicate higher willingness levels to sign petitions than respondents in the private condition.
5.1.3. Psychological standing: Be moral to engage or engage to be moral?
This research contributes to literature on psychological standing in two ways. First, results demonstrate that people who feel it is appropriate and they have the right to perform a certain action will subsequently engage in such action. This finding is in line with several scholars who state that, in order to perform social action, people need to feel a certain legitimation or entitlement to act upon their attitudes (Effron & Miller, 2012; Miller et al., 2009).
This brings us to the second contribution of this research, to the mechanism of psychological standing. There are several ways to derive such entitlement to act from which had been discussed in earlier chapters. One way is by individuals being materially affected by the issue at hand. However, due to the research design, it was impossible for respondents to have a material stake in this matter. Other factors such as personal experience were further assumed to be generally low, due to low mean scores for items relating to previous activism and petitioning behaviors. Respondents who had rarely engaged in previous public sphere behaviors were not expected to derive entitlement from relevant past experiences.
Nevertheless, individuals felt that it was appropriate for them to sign, judging by their high overall rating on the factor “psychological standing”. Now, what other reasons could have entitled them to indicate willingness to sign the petition?
41
standing. In other words: People with a higher moral self-image also perceive a higher appropriateness for themselves to sign petitions.
Secondly, one other form of entitlement granting respondents with psychological standing was found to be personal relevance. Participants who indicated high levels of care for the environment also reported higher perceived appropriateness and right to sign the petition.
This contributes to findings of Miller et al. (2009) who detected that personal relevance and familiarity with the issue enhances perceptions of appropriateness to act.
5.1.4 I act, so I am: Behavioral spillover and self-perception
Within this research, no significant spillover effects from private onto public sphere behaviors were detected. Simultaneously, no spillover effects from past public sphere onto future public sphere behaviors were found.
However, what was partly proven, was the self-perception theory. As mentioned in previous chapters, self-perception mediated the effect between previous and future pro environmental behavior (Lauren et al., 2018). This was explained by the self-perception theory, namely the act of constantly deriving one’s attitudes and beliefs from previous behaviors. In line with these findings, viewing oneself as environmentally conscious and as caring about the environment significantly affected respondents’ tendency to sign the petition. Also, respondents who had engaged in public sphere behaviors in the past were found to view themselves as environmentally conscious. Here, the mediating effect of environmental self-perception was confirmed. The lack of significant differences between positive and negative feedback regarding their effect on one’s environmental self-perception and reasons why will be discussed more in the next part.
