Faculty of Economics and Business
Master of Science in Marketing Management MASTER THESIS
Author: Sirin Yildiz Student number: S2351447 E-Mail: s.a.n.yildiz@student.rug.nl
Supervisor: Dr. M. Moeini Jazani Co-assessor: prof. Dr. B.M. Fennis
13.01.2020
Word Count: 9005
TABLE OF CONTENTS
1. Introduction ... 4
2. Literature Review ... 6
2.1 Time framing and environmental concern ... 6
2.2 Environmental concern and behavioral intentions ... 12
2.3 Moderators ... 15
3. Conceptual Model ... 18
4. Methodology ... 19
4.1 Data Collection and Sample ... 19
4.2 Manipulation and Study Design ... 20
4.3 Measures ... 22
4.4 Results ... 24
4.5 Summary of Results ... 30
5. General Discussion ... 32
5.1 Theoretical implications ... 32
5.2 Managerial Implications ... 34
6. Future research and Limitations ... 36
7. Conclusion ... 38
8. References ... 39
9. Appendices ... 54
Appendix A ... 54
Appendix B ... 56
Abstract
Past research has shown the effect of time framing in different domains, such as the health and the food domain. This research examines the framing of time (i.e., granular vs. less granular expressions of duration) in a new domain, namely the environmental domain. In this research, it is proposed that individuals who are exposed to environmental risk expressed in days, as opposed to years, will display higher pro-environmental behavior.
Based on past research, we expected that day framing will result in more proximity, which will result in a higher perceived environmental concern, and therefore in a higher pro- environmental behavior. Accordingly, it is proposed that the relationship between time framing and pro-environmental behaviors is mediated by environmental concern.
Results of an extensive study largely support these predictions that time framing is an effective tool in communicating climate change risks. However, this effect is going in the opposite direction of what was expected. When the environmental risk is framed in years and therefore is less granular, individuals are more motivated to get engaged with
environmentally oriented behavior. Additionally, two moderators are included in this research; environmental self-efficacy and biospheric values, which both showed a non- significant effect on the relationship between time framing and environmental behavior.
Keywords: Temporal Framing, Granularity, Environmental Behavioural Intentions,
Environmental Concern, Biospheric Values, Environmental Self-Efficacy
1. Introduction
Pro-environmental behavior is a salient concern in contemporary society (Mahardika, Thomas, Ewing, & Japutra, 2019). Mounting scientific evidence suggests that human- induced climate change may pose a significant threat to humans and the wider environment (Whitmarsh, 2009).
Changing individuals’ environmental behavior can solve most of the environmental problems that we have encountered (Steg, Perlaviciute, & van der Werff, 2015). The current
generation's success in achieving sustainable patterns of living on the earth will determine whether the planet will retain a livable environment for future generations (Oskamp, 2007).
Therefore, there is a consensus that a change in behavior is needed to reduce environmental impacts and improve positive impacts on human health (De Young, 1993). The rapid development of the economy and increasing public exposure to environmental
lead to an increase in the expected number of people who will adopt pro-environmental behavior (Chen, De la Rosa, Peterson, Zhong, & Lu, 2016).
However, policies to achieve pro-environmental behavior have met with limited success, after decades of information campaigns and economic measures to encourage 'green' behaviors, the public is prepared to recycle, but few take action beyond this (Whitmarsh, 2009). Researchers and industry practitioners have developed different approaches to
promoting pro-environmental behavior, emphasizing different benefits. However, there have been mixed findings in terms of which approach is the most effective (Xu, Arpan & Chen, 2015). Research on energy consumption behavior highlights various psychological, social, economic, and physical barriers to fostering energy conservation (Whitmarsh, 2009);
therefore, environmental behavior cannot be achieved simply through information provision
and economic measures (Jackson, 2005).
Relevant here is, is that the way that information is framed is inevitable when communicating about any complex phenomenon (Graber, 2014), such as pro-environmental behavior.
Several researchers have conducted studies about the framing of information (Spence, Poortinga & Venables, 2010; Kees, 2011), a specific type of framing is temporal framing, which has been studied intensively by several researchers in different domains (Xu et al., 2015; Chandran & Menon, 2004). Nevertheless, there is high relevance to apply temporal framing specifically to the environmental domain and environmental risks in order to encourage pro-environmental behavior among individuals.
According to Chandran and Menon (2004), when a hazard is presented in a day frame, the risk is construed as more concrete and proximal compared with a year frame, enhancing the effectiveness of the message. Therefore, in this master thesis, we will examine the effect of framing environmental risk expressed in two different levels of granularity. Environmental risk expressed in days and environmental risk expressed in years, where objectively, the time is the same, just expressed in different units. The findings aim to achieve results that can be used to communicate environmental risk effectively.
The paper is structured as followed: Firstly, the article will discuss relevant literature, based upon that, four hypotheses are created, which will be combined into a conceptual model.
Secondly, the thesis will provide the methodology to test the hypothesis. Thirdly, the results will be provided in the general discussion, consisting out of the theoretical and the
managerial implications. After this, future research suggestions and limitations will be
provided. Lastly, the paper will conclude with the conclusion.
2. Literature Review
Based on previous literature, this literature review will provide extensive description and analysis of the relevant variables used in this thesis. Throughout the literature review, we will propose the relevant hypotheses being tested in this research.
2.1 Time framing and environmental concern
Our subjective assessments of the time that passes may be inaccurate (Grondin, 2010) since the perception of time is subjective (Larson, Larson and Katz,1991). This argument is reinforced by Fredrickson and Kahnemen (1993), who state that “subjective perceptions of time are different from objective changes.” Therefore, individuals’ subjective estimates of duration do not accurately map onto objective time (Zauberman, Kim, Malkoc and Bettman, 2009). Individuals have difficulty thinking about time as an independent dimension and often misjudge the duration of events (Zauberman et al., 2009). Indeed, research shows that
consumers’ mapping of objective duration onto subjective time is nonlinear and characterized by insufficient sensitivity to changes in duration.
Since time is perceived differently by individuals (Droit-Volet & Meck, 2007), it is possible
to influence the subjective perceptions of time (Baker & Cameron, 1996). Time perception
can be altered by contextual factors (Antonides, Verhoef, & Van Aalst, 2002). One of these
factors is the temporal context, where framed outcomes influence the effects of the message
(Gerend, & Cullen, 2008). Chandran & Menon (2004) argue that the framing of time can
evoke a different psychological process, and the manipulation of time perceptions could
influence decision making (Kim & Jang, 2017).Churchill, Good & Pavey (2014) identify
temporal framing as “the application of a time frame to a particular outcome.”
According to Liberman and Trope (1998), the temporal frame of the message can influence how a message is construed. Perceptions of temporal distance systematically alter the way future events are construed and thus influence the evaluation and choices related to those future events (Liberman and Trope, 1998). Distal events seem more abstract and
decontextualized than proximal events and outcomes that occur in the distant future may have a less psychological impact than those expected to occur faster (Liberman & Trope,1998).
Relevant in the temporal framing literature is temporal discounting, which is the tendency of people to discount rewards as they approach a temporal horizon in the future or the past (Doyle, 2012). Temporal discounting can be expected to strongly affect people’s risk
judgments and their willingness to change appropriate behaviors (Gattig & Hendrickx, 2007).
These findings are reinforced by Nicolaij and Hendrickx (2003), in their study, they found that longer outcome delays result in lower average risk judgments and a decreased tendency to change appropriate behaviors. Assessments of temporal discounting often require
participants to choose between a smaller, immediate reward and a more substantial reward gained after a specified delay, such as £1 now or £5 after two weeks (Pavey & Churchill, 2017). There is a general tendency for diminished value to be placed on distant-future reward as oppose to near-future reward (Trope & Liberman, 2010). Humans discount outcomes that are uncertain, temporally delayed, spatially distant, and occurring to others (Gattig &
Hendrickx, 2007). The discounting processes involved appear to be similar across different outcome dimensions (Gattig & Hendrickx, 2007).
This information is highly relevant to the environmental domain since many environmental
problems are uncertain, have long-term consequences, and some have both immediate and
long-term adverse effects (Gärling & Schuitema, 2007). For instance, urban air pollution
from traffic may cause immediate odor annoyance as well as long-term health problems
(Gärling & Schuitema, 2007). Other environmental problems, such as the increase of
atmospheric greenhouse gas concentrations or the depletion of natural resources, such as oil, water, minerals, do not have severe immediate effects. However, in the long run, such risks may have catastrophic consequences (Gattig & Hendrickx, 2007).
To make individuals more aware of these environmental problems, framing information can be applied since framing can influence attitudes and behaviors by highlighting certain aspects of the situation or by changing how one might think and consider the presented information (Chan, 2019). Besides, temporal framing can be used to encourage pro-environmental behavior, since, for pro-environmental behaviors, it is that both negative and positive consequences are often somewhat uncertain and occur over the long-term and this restricts the ability of an individual to realize the benefits of behaving pro-environmental (Grønhøj,
&Thøgersen, 2011). Consequently, interventions designed to change the subjective perception of future time can be used as a strategy (Zauberman & Kim, 2011) in the
environmental domain. Subtle changes in the message could alter how long or short a future event might seem (Zauberman & Kim, 2011).
Temporal framing has been heralded as a promising strategy to make risks appear more
proximal and concrete and thus counter temporal discounting effects (Lo, Smith, Taylor,
Good & von Wagner, 2012). In order to use temporal framing to affect individuals their risk
judgments and their willingness to change behavior, it is relevant to mention that individuals
perceive differences between two values when those values are communicated with the usage
of different units (Wertenbroch, Soman, & Chattopadhyay, 2007). Therefore, in this research
paper, temporal framing is operationalized by presenting temporal distance of environmental
risk in different time units, namely in days versus years (Chandran & Menon, 2004). This
manner of temporal framing is that the same quantity of time can be expressed at different
levels of granularity, for example, “1 year,” “12 months,” or “365 days.” (Zhang & Schwarz,
2011). Although there is no rational reason to suppose that a week might differ from 7 days, it does (Adaval, Pandelaere, Briers, Lembregts, Bagchi, Li, & Coulter, 2013). This effect leads to errors in estimation that affect not only how we make progress in goals that are set but also affect, at a more fundamental level, how we perceive things (Adaval et al., 2013).
Therefore, we frame the risk of an environmental disaster happening in a similar subjective time yet expressed in different units. Namely, the time expressed in days and the time expressed in years. Temporal framing of objectively neutral reference periods leads to
differential subjective perceptions of the distance (Chandran & Menon, 2004). Therefore, the persuasive effect of a message can differ even when the content is the same, depending on how the contents are framed and delivered (Kahneman & Tversky, 1979).
Also, how messages are framed influences the extent to which they are attended to, how much knowledge the audience gain from them, and how positively such messages are evaluated (Graber, 2004; Lakoff, 2010). So, using the time frame as a unit influences the perception of an individual (Pandaleare, Briers and Lembregts, 2011), and the size of the unit is used to make an inference about the proximity of an event (Monga, & Bagchi, 2011). Fine- grained utterances increase confidence that the information is accurate (Zhang & Schwarz, 2011), resulting in environmental risk perceived as more accurate. We can state that expressing an attribute in a different unit leads to a more significant perceived difference if the unit is on an expanded or finer-grained scale (Pandelaere, Briers, & Lembregts, 2011).
Examples are that a donation seems smaller when it is framed using a per-day rather than a per-year format Gourville (1998), and individuals underestimate when they budget for a month rather than for a year (Ülkümen, Thomas, & Morwitz, 2008).
Once a future event is understood as having a greater relevance to the present, individuals
should be more willing to take action (Lewis and Oyserman, 2015). For the environmental
domain, if the environmental risk can have greater relevance to the present, as a result of how
time units are expressed, individuals should be more willing to take action in favor of the environment. Chandran & Menon (2004) show in their research that comparisons can be made between messages containing the sentence “every day” or “every year.” Objectively, the time frame does not convey information about the temporal distance. However, it has been argued that the day and year frames trigger different subjective temporal perceptions (Chandran & Menon, 2004). That systematically affect judgments of risk’ associated with events that are closer in time, more proximal, and thus more concrete and probable than those associated with a year frame (Chandran & Menon, 2004). Additionally, the study of Lewis and Oyserman (2015) showed that people say they will start saving four times sooner if told how many days rather than how many years they have until their child goes to college or until they want to retire. Changing the metric used to frame when the future begins should make the future feel psychologically relevant to the current situation by making it connected to and congruent with the present self (Lewis and Oyserman, 2015).
Individuals experience temporal granularity, the farther away a future event is, the fewer details can be imagined, the closer a future event is, the more details can be imagined (Lewis and Oyserman, 2015).
Additionally, compared with closer events, further events are typically considered using a more gross-grained time metric, such as years rather than days (Lewis and Oyserman, 2015).
So, the use of temporal reference frames, such as a day or year, is likely to trigger subjective
associations with a near or a distant future, respectively, that results in temporal distance
effects (Chandran & Menon, 2004). A day frame suggests that an event is closer in time and,
therefore, more proximal and thereby replicating a near-future event (Chandran & Menon,
2004). Contrastingly, a year frame moves the event further in time and is, therefore, less
proximal, replicating a distant future event, evoking a different psychological process from
the temporal of every year (Chandran & Menon, 2004).
As mentioned, individuals construe proximal events in more concrete and contextualized terms, while distal events are construed in abstract, decontextualized terms (Trope &
Liberman, 2000). Message framing focused on a day could be used to increase the perceived temporal proximity to the outcome (Chandran &Menon, 2004). Chandran and Menon (2004) argue that a hazard seems more proximal when it is expressed in terms of a per-day rather than a per-year basis; the per-day formulation makes the hazard seem closer in time because a day is shorter than a year.
If we apply this theory for the environmental risk, we can, therefore, state that:
H1: Temporal framing will affect environmental concern such that these judgments are
higher when the duration to environmental risks is expressed in a day frame as opposed to a
year frame.
2.2 Mediator and Dependent Variables: Environmental concern (ME) and pro-environmental behavior (DV)
Many environmental problems are related to human behavior. Therefore individual action in the marketplace can remedy environmental problems (Paavola, 2001). Environmentally significant behavior can reasonably be defined by its impact: the extent to which it changes the availability of materials or energy from the environment or alters the structure and dynamics of ecosystems or the biosphere itself (Stern,1997). Particular behavior, such as clearing forest or disposing of household waste, directly or proximally, causes an
environmental change (Stern, Young, & Druckman,1992). Where other behavior is
environmentally significant indirectly, by shaping the context in which choices are made that directly cause environmental change (Rosa & Dietz, 1998).
A more thorough understanding of the relationship between concern and behavior is
necessary if marketers and public policymakers are to reduce environmentally harmful
behavior (Kilbourne & Pickett, 2008). Kilbourne & Pickett (2008) found in their study that
when concern increases, consumption behavior might become more environmentally
sensitive, and consequently, consumers may change purchase patterns, tilting them toward
going greener. Minton & Rose (1997) reinforce this argument and state that "what makes
people more inclined to behave in environmentally friendly ways is an awareness of various
environmental problems and the consequences of their behavior in response." "This may
result in a willingness to relocate their resources such as time, money, and attention, to make
personal behavior more environmentally friendly" (Minton & Rose, 1997). Additionally,
individuals become engaged in environmental issues and perform pro-environmental
behavior since they are more concerned about the adverse consequences of environmental
problems for themselves, others, or the biosphere (Hansla, Gamble, Juliusson & Gärling,
2008).
Thus it has generally been found that there is a positive correlation between environmental concern and environmentally friendly behavior (Straughan and Roberts, 1999). Therefore, mitigating environmental problems requires behavioral changes by individuals all over the world. Encouragingly, studies have shown that the majority of the world's population is aware of environmental problems and supports environmental protection (Milfont & Schultz, 2016). If the majority of individuals are concerned about the environment, the collective concern could be manifested by individual pro-environmental behavior (Sapci & Considine, 2014). Pro-environmental behavior best can be described as 'behavior that is undertaken to change the environment (Stern, 2000) and is a mixture of self-interest and concern for other people, the next generation, other species, or whole ecosystems (Bamberg, & Möser, 2007).
Since environmental concern is an important issue ( Diekmann & Franzen, 2019),
influencing individuals, their environmental behavior should be of interest. Subjective risk plays a role in influencing individuals their behavior (Rundmo, 1999), which is also applicable to the environmental risk domain. Since environmental concern significantly affects environmental behavior (Rundmo, 1999). Kilbourne and Pickett (2008) argue that when concern increases, consumption behavior might become more environmentally sensitive. An example of a particular behavior is that individuals who perceive that
environmental problems constituted a severe threat to health were more likely to engage in environmental practices, such as recycling, conserving water, buying environmentally friendly products, and limiting their driving (Baldassare & Katz, 1992). Therefore, we can state that personal environmental threats had a significant role in predicting overall
environmental practices (Baldassare & Katz, 1992).
Another example is that environmentally concerned households tend to be more conservative
in their use of energy (Sapci & Considine, 2014). Hines, Hungerford, and Tomera (1987)
found a significant and positive correlation between general measures of environmental
concern and environmental practice. Individuals who demonstrate strong environmental concerns are more likely to undertake waste recycling, green purchasing behavior, and lower energy consumption (Lin & Huang, 2012). Consumers who have environmental concerns are willing to behave in a more environmentally friendly way to help environmental protection (Sapci & Considine, 2014). Lin and Huang (2012) found in their study that when presented with certain conditions or situations, such as perceived worsening environmental threats, or the availability of subsidies or discounts for green products, consumers with severe
environmental concerns are more willing to behave in a pro-environmental way. Therefore, consumers who are concerned more with the environment are more willing to make efforts to change (Lin & Huang, 2012).
We assume that in a day framing, an environmental risk will be perceived as more severe, which will lead to individuals being environmentally concerned in comparison when the environmental risk would be communicated in a year frame. From the theory altogether, we can state that individuals who are environmentally concerned behave in a more pro-
environmental responsible way.
Therefore, we propose:
H2: Perceived environmental risk will mediate the effect of temporal framing on pro- environmental behavior, such that in a day (vs.year) frame, consumers show increased pro- environmental behavioral intention.
2.3 Moderators
Two moderators were included in this research, environmental self-efficacy and biospheric values, to examine the relationship between framing and behavioral intention more deeply.
Environmental Self-efficacy
Self-efficacy can be defined as the belief in an individual's capacity to organize and guide the courses of action required to tackle certain situations in the immediate future (Wood &
Bandura, 1989). Perceived self-efficacy is a significant determinant of intention, which further influences performance (Bandura, 1997). A higher level of self-efficacy in specific tasks indicates more excellent capabilities and confidence in performing the tasks and is likely to increase an individual behavioral intention to take action (Huang, 2016).
Several researchers argue that self-efficacy or effectiveness can affect people's environmental behavior (Lee, 2011; Lee et al., 2014; Tabernero & Hernández, 2011).
Environmental self-efficacy has a positive direct effect on pro-environmental behavior (Huang, 2016). Bandura (1997) argues that self-efficacy in the environmental domain is strengthened when people master the specific skills required to engage in pro-environmental behavior and are verbally persuaded of their ability to perform such behavior. An example is from the research from Huang (2016), he found that when individuals have higher levels of self-efficacy in mitigating global warming, they are likely to be more interested in obtaining updated information for possible solutions and in acting. Additionally, Meinhold & Malkus (2005) argue that pro-environmental attitudes and behaviors are stronger among individuals with high levels of self-efficacy. Individuals engage in and feel better about pro-
environmental acts, and therefore their self-efficacy in their lives will increase (Meinhold &
Malkus, 2005). Therefore, environmental self-efficacy may influence the relationship of
temporal framing on environmental behavioral intentions since individuals who are high in
environmental self-efficacy are involved in pro-environmental behavior. Accordingly, those individuals are less likely to be influenced by the time frame in comparison with individuals who score low on environmental self-efficacy.
Therefore, we propose:
H3: The effect of framing will be more significant among people with low environmental self- efficacy.
Biospheric Values
Individuals differ in their amount of protecting the environment (Chan, 2019), and personal values can explain people's pro-environmental actions and in which these actions may be more aligned with nature, the biospheric values (de Groot & Steg, 2008). Biospheric values emphasize the quality of nature and the environment independently from the benefits it provides for human beings (Steg and De Groot, 2012). Individuals who endorse biospheric values, therefore, assign importance to the intrinsic value of the ecosystem, which translates into a moral imperative to act pro-environmentally (Stern and Dietz, 1994).
Biospheric values are considered to be important antecedents of environmental behavior (Werff, Steg, Keizer, 2013). Schultz, Gouveia, Cameron, Tankha, Schmuck, & Franek (2005) relate pro‐environmental intention and behavior positively to altruistic and biospheric
concern. Furthermore, several researchers have shown that individuals with strong biospheric
values are more like to have pro-environmental intentions and to act pro-environmentally
(Steg & Groot, 2012). Previous research showed that biospheric values had been related to
pro-environmental behavior (Schultz & Zelezny, 1998), acceptability of climate change
policies (Nilsson, von Borgstede & Biel, 2004) and environmental activism (Steg et al.,
2011). Therefore, biospheric values may moderate the relationship between temporal
framing and environmental behavioral intentions. Individuals who are high in biospheric
values are already related to pro-environmental behavior and therefore are less likely to be influenced by the time frame in comparison with individuals who carry low biospheric values.
Therefore, we propose:
H4: The effect of framing will be more significant among people with low biospheric values.
3. Conceptual Model
Based on the literature review and the two proposed hypotheses, a conceptual model is constructed. In this conceptual model, temporal framing and duration are the independent variables, which will influence environmental behavioral intentions, where this effect is mediated through the environmental concern. Furthermore, the moderator's biospheric values and environmental self-efficacy influence the relationship that temporal framing has on environmental behavioral intentions.
Figure 1: Conceptual Model
4. Methodology
This part of the thesis will provide information about the methodology that is used to test the four proposed hypotheses. Besides, it will provide information about the effect of the
influence of the mediator's environmental concern and the two moderators' biospheric values and environmental self-efficacy. The methodology will contain the data collection and sample, manipulation of the study design, measures, results, and will conclude with a summary of the results.
4.1 Procedure Data Collection and Description Sample
To acquire information about the relationship between time framing and personal
environmental behavioral intention, where this relationship is mediated by perceived
environmental concern, a questionnaire was conducted. In December 2019, data was
collected through a survey via the online survey platform Qualtrics. Participants received a
payment for their participation. The study initially contained 363 participants. Several
attention questions were incorporated into the questionnaire, and those who failed these
quality checks (N=53) were excluded from the analysis, resulting in a final sample of 310
participants. The participants consisted for 42,6% (N = 132) out of male, 57,1% (N = 177)
out of females and 0,3% (N = 1) was specified as “other”. The average age of the participants
in the study is 37, with a standard deviation of 12.134 and a mean of 37,35. The minimum
age is 19 years, and the participant with the maximum age was 74 years. Regarding the
participants their political orientation, 18,4 % (N = 57) identified themselves as Republic,
41,6% (N=129) as Democrat, 32,6 % (N = 101) as Independent, 1,9% (N =6) as other and
lastly 5,5% (N = 17) of the respondents did not have a political preference.
With reference to employment, 45,5 % (N = 141) was full-time employed,18,1 % (N= 56) was part-time employed, 15,8% (N= 49), self-employed, 7.1% (N=11) unemployed but looking for a job, of the respondents 3,9% ( N=12) was retired, 4,5% ( N=14) identified themselves as housewife/househusband 14 4,5 and lastly, 5,2% (N= 16) was unable to work
& others.
4.2 Manipulation and study design
In order to test the hypothesis, a 2 (temporal frame: day vs. year) x 2 (objective duration: 27 vs. 41 years) design was used. In this research, a 2-factor between-subject experimental design was conducted. This thesis consists of 2 independent variables, namely, time framing, and objective duration. Both variables are categorical and consist of 2 levels. First, the perception of time was manipulated by using different expressions of time, namely day versus year. The duration was manipulated, consisting of the two different levels, 27 years and 41 years. The reason we included two different durations of time was to examine the generalizability of our framing effect across different objective times.
The participants were randomly assigned to one of the four different conditions. In which they read a news article on climate change's impacts occurring in 27 years expressed in days (N = 76), 27 years expressed in years (N =81), 41 years expressed in days (N =83) and 41 years expressed in years (N =70). The news article is as followed: “What makes this report particularly unique is that it provides concrete estimates of the so-called point of no return based on our current trajectory of emissions,” says Thomas Wirth, the IPCC spokesperson.
According to the report, scientists forecast that in (9,855 days/14,965 days/27 years/41
years)
We will reach a crucial threshold beyond which it will be impossible to recover the planet back to a normal state. In the US alone, the increased frequency and variety of natural disasters such as ultra-intense hurricanes Harvey, Florence, and Dorian as well as massive and extreme wildfires, floods and droughts across the country are pointed as harbingers of a warming globe and irreversible climate change threshold we are reaching in the next (9,855 days/14,965 days/27 years/41 years). "It is possible to limit global warming to 1.5 degrees Celsius but doing so would require unprecedented changes in our lifestyles and major transformations in the ways our global society functions and interacts with natural
ecosystems," the report says. Lowering emissions to this degree would require widespread transitions in energy, urban, transportation, infrastructure, agricultural, and industrial systems in the next (9,855 days/14,965 days/27 years/41 years). "International cooperation, as well as individuals' adoption of preventive measures, are imperative to limit emissions and, therefore, global warming and its devastating impacts. Decisions made today will have lasting consequences for future generations", says Miriam Perkins, a scientist at the Climate Change Research Center at Yale. The next (9,855 days/14,965 days/27 years/41 years) will be critical in the evolution of these efforts.
The news article highlighted environmental risk occurring in the future, and the article was
inspired by the scenario used by Schuldt, Rickard, and Yang (2019) in their study about how
departure dates influence climate change engagement. The stories across all four conditions
were identical. The time framing and duration were different. In the article, there was an
emphasis on framing the risk, and the article highlighted multiple times the moment when
environmental risk would occur.
4.3 Measures
After exposure to the article, participants were provided with a questionnaire; we state the questions in order of their presentation. Furthermore, it is stated from which articles the questions and the scales are derived, and information is provided on which scale is used.
Perceived Concern (Mediator). After exposure to the scenario, perceived concern, the mediator, was assessed. Using a total of 12 questions derived from Van der Linden (2014 &
2015), McRight (2010), and Hornsey et al. (2015). Participants responded on a 7-point Likert scale (1 = Not at all agree, 7 = Very much agree). The questionnaire included items such as “I feel the consequences of climate change are inevitable for future generations” and Climate change is an important issue.”On average, the participants were quite concerned about environmental risk after reading the news article (M=5,56, SD=1.58).
Behavioral intentions (DV). On a 7-point Likert scale (1 = Strongly disagree, 7 = Strongly agree), personal behavior intentions were measured. The scale that measured environmental behavioral intention contained 19 items and was derived from Gifford & Comeau (2011) and Tobler et al. (2012). The questionnaire included items such as “I am willing to save
electricity (e.g., by switch off my lights when not in use or by using energy-efficient light bulbs” and “I am prepared to greatly reduce my energy use to help tackle climate”. In
general, participants were also quite intended concerning their behavioral intentions (M=5,38, SD=1,07).
Self-Efficacy (Moderator). Environmental self-efficacy was measured with four items on a
5-point Likert scale (1=Strongly disagree, 5=Strongly agree). These four items were derived
from past research (Heath & Gifford, 2006; Benjamin, Por, and Budescu, 2017). The
difference to alleviate the negative effects of global warming” and “There is very little I can do to mitigate the negative effect of global warming." Generally, participants scored
moderately on environmental self-efficacy (M=3,18, SD=0,59).
Biospheric Values (Moderator). Lastly, on a 5-point Likert scale (1 = Strongly disagree, 5
= Strongly agree), 6 items were used to measure participants ‘Biospheric Values'. This scale was derived from Dunlap (2000), Ziegler (2017), and Whitmarsh (2011). The questionnaire included items such as “Humans were meant to rule over the rest of nature," and "Plants and animals have the same right to exist as humans." Participants in general were in favor of these values (M = 3, 86, SD=0,79).
Demographics. The questionnaire concluded with several questions to identify the
demographics of the sample size. Items included participants to specify their age, gender,
income, political orientation, education, ethnicity, and description of their job.
4.4 Results
Personal Environmental Behaviour. To test Hypothesis 1 and to analyze the impact of time framing and duration on personal environmental behavior, a two-way ANOVA analysis was conducted. The two-way ANOVA not only aims at assessing the main effect of each
independent variable but also identifies if there is any interaction between them (Casella, 2008). Therefore, the two-way ANOVA analysis tested for an interaction effect between the two factors. The result from this two-way ANOVA (appendix table B.1) shows that there was no interaction between time framing and duration on environmental behavior (F (310) = 0.025, p =0.874, η2p = .020. Additionally, the main effect of the factor duration is highly insignificant on environmental personal intentions. (F (310) = 0.074, p=0,786, η2p = .020 and therefore has become irrelevant as a factor. The 2-way ANOVA analysis support that there is a significant main effect of time framing on environmental personal behavioral intentions (F (310) = 5.752, p <0.017 η2p = .020. In figure 1, it is observed that when the environmental risk occurs in a year frame, expressed in blue, personal environmental behavior is
significantly higher, compared to when the risk was communicated in a day frame, expressed in red. These results show that the time framing is effective in motivating behavioral
intentions, regardless of the duration (27 or 41 years). However, the direction of the effect is the opposite of what we had initially hypothesized. That is when the durations were
expressed in years (vs. days) people showed a higher tendency to engage in behavioral
changes.
27 years 41 years Figure 1
As noticed, the results of the two-way ANOVA analysis showed that there was no interaction effect and that the factor framing has a significant main effect on personal environmental behavior. Therefore, we collapsed our data across durations and analyzed our data only using the framing (year vs. day) as the IV in the subsequent analysis.
To assess the effect of temporal framing on the personal environmental behavior in both framing conditions, two separates one-way ANOVA analysis was conducted. Namely, the effect of time framing in the 27 years condition and the 41 years condition.
Framing 27 years. To test the effect of time framing on personal environmental behavior for 27 years, a one-way ANOVA analysis was conducted. According to Malhotra (2009), the one-way ANOVA provides an understanding of the influence of one nominal independent variable on an interval dependent variable. The results of this analysis can be found in
appendix table B.2 The result of the one-way ANOVA showed a marginally significant effect (F(310) = 2.585, p = 0.110),)η2p = 0.10). Participants in the 27 years conditions showed that personal behavior significantly was influenced when environmental risk was framed in a year as opposed to a day frame.
Blue = Day Frame
Red = Year Frame
Framing 41 years. Secondly, a one-way ANOVA Analysis for 41 years was conducted. The findings are presented in appendix table A.3. The result of the one-way ANOVA showed a significant difference (F(310) =3,173 p= 0.077., η2p = .050), Participants in the 41 years conditions showed that time framing significantly influenced personal behavior when environmental risk was framed in a year opposed to a day.
Based on both analyses, we can conclude that there was a main effect of framing on each duration. Additionally, if we compare both durations, The One-way Anova analysis was, in both cases, significant. Yet the effect of time framing in the condition of 41 years is stronger compared to the 27 years condition (although this effect did not qualify a significant
interaction). In the following analyses, we continue exploring data while the data is collapsed across the duration conditions.
Mediation Analyses. Therefore, in order to test Hypothesis 2, a mediation analysis was conducted by using the PROCESS Macro model 4 in SPSS (Hayes, 2013). Meditation test if the relationship between time framing and personal environmental behavior can be explained by environmental concern.
1
From the mediation analysis, Appendix B table 7, we can derive that the relationship
between the independent variable, time framing, and the mediator, the concern is significant.
1 1Firstly, we conducted a mediation analysis for both durations separately. 27 years was marginally significant mediated, and 41 was significantly mediated. These can be found in Appendix A 5 and A 6. Considering both mediations 27 and 41 years, have the same pattern of results, the durations are combined in the analysis, to have a larger sample size and have higher predictive power. For each duration separately, the results can be found in the appendix.
(B=0,4557, SE=0.1178, p=0.0108<0.05). The effect of the mediator's concern on the independent variable's environmental behavior is highly significant (B=0.4528, SE=0.0288, p=0.000<0.05), concern is highly significant in predicting personal behavioral intentions.
The direct effect of the independent variable, time framing on the dependent variable is not significant. (B=0.0819, SE=0.097, P=0.3678 >0.05). The indirect effect of X and Y is significant, which is confirmed by a bootstrap analysis (b = .2063, SE = .0819, 95% CI = .0449 to .370. Based on the analyses, we can state that there is an indirect significance since this 95% confidence interval is not including any zero values. Therefore, the entire effect of time framing is being conveyed to the dependent variable, trough the mediator concern, which is a sign of mediation. The result of the mediation analysis provides support for the second hypothesis. This mediation result supports our hypothesis 2.
Moderation Analysis. This research consisted of 2 moderators that may have influenced the relationship of the independent variable time framing on the dependent variable personal environmental behavior. To explore if the moderator's environmental self-efficacy and biospheric values separately are moderating the relationship between the independent variable and dependent variable, two separate moderation analysis were conducted.
Environmental Self-Efficacy. To identify whether environmental self-efficacy moderates the relationship between time framing on pro-environmental behavior, moderation analysis was conducted. This was done by using the PROCESS Macro model 1 in SPSS (Hayes, 2013).
The output can be found in Appendix B, Table 8. From the analysis we can conclude that the interaction is not significant within environmental values, (F (310) F= 0,1192 P>0,05 P = 0.7302. From figure 2, we can derive that among those who score low and average on environmental self-efficacy, the effect of framing is significant. The framing of time does not influence individuals that score above average.
Effect SE T P
Below Average .3128 .1594 1.9621 .0507
Average .2739 .1126 2.4327 .0156
Above Average .2350 .1593 1.4752 .1412
Figure 2
Biospheric values. The second moderator, biospheric values, showed a similar pattern.
Repeatedly we used the PROCESS Macro model 1 in SPSS (Hayes, 2013). The output can be found in Appendix B,Table 9. From the analysis we can conclude that the interaction is not significant (F (310) F= 0,3904 P>0,05 P = 0.5326. Although the interaction was not
significant, we looked into the effect of framing on the DV at different levels of the
moderator. These results are provided in figure 3; we can derive that among those who have low environmental values, the effect of framing is marginally significant. Such that a year has led to more pro-environmental behavior as opposed to a day frame. Among individuals who have high environmental values, framing has not further increased their pro-
environmental behavior, since this is not significant.
Effect SE T P Below Average .2419 .1415 1.7097 .0883
Average .1795 .0990 1.8135 .0707 Above Average .1172 .1397 .8389 .4022
