How can environmental charity be promoted to the public?

(1)

How can environmental charity be promoted to

the public?

The role of message framing, nature imagery, and emotions in

environmental charity advertising

Aurora Duci

(2)

How can environmental charity be promoted to

the public?

The role of message framing, nature imagery, and emotions in

environmental charity advertising

Master Thesis Marketing Management

Author: Aurora Duci

Address: Via Ludovico Muratori 13, 24030, Mozzo (BG) Email: a.duci@student.rug.nl

Phone number: +39 3473532383 Student number: S4101634

Department: Faculty of Economics and Business Master: Marketing Management

Supervisor: Dr. J.I.M. de Groot Second supervisor: Dr. J. Berger

(3)

Acknowledgements

Since I started my bachelor’s in business administration in Bergamo, I hoped I could study and live abroad at least once in my life. I loved my studies in the business area and I knew I wanted to expand my knowledge in the marketing field. Once I moved to Groningen, I knew this Master was something special to me and I wanted to bring out the best from this experience. It has been a tortuous path of new information, experiences, lectures, and exams. It has been overwhelming at times, especially during the Covid-19 measures. Not being able to see my family and friends but on the other hand seeing every day all of the negative information about Bergamo and Italy was heartbreaking. However, this experience taught me a lot, not only in terms of theoretical knowledge gained but also in terms of knowing myself better and of coping with difficult situations. Now, this chapter of my life has almost come to an end, and I can surely say that it is a positive end, also because of those who were next to me during this year. I would like to express my sincere gratitude to all of the people that made this experience possible and positive.

First, I would like to thank my parents that never discouraged me from moving abroad and, more importantly, decided to invest in my education. Second, my supervisor, who has been present and very understanding of my personal situation. I am glad that she was my supervisor as she showed humanity and interest. Moreover, her guidance during the thesis process was fundamental to learn new things and to improve my skills. Thanks to her feedback delivered the work I have done. Third, but not of least importance, I would like to thank all of the beautiful people I have known in Groningen, especially my boyfriend Lennard and my friend and classmate Nicola. They are those I spent most of my time with and they made me feel like I was at home and always supported me.

(4)

Abstract

This study investigated the effect of message framing (gain vs. loss) and nature imagery (pleasant vs. unpleasant) on emotions and donation intention. More specifically, the research looked into the joint effect of message framing and nature imagery on emotions, into the mediating role of emotions on donation intention, and into whether a loss-framed message would increase donation intention more than a gain-framed one. The results of this 2 by 2 between-subjects experiment, run with a survey on 269 participants, showed that nature imagery moderates the effect of message framing on emotions. However, emotions do not have any relation to donation intention. Therefore, the indirect mediating effect hypothesized was not supported. On the other hand, a direct effect of message framing on donation intention was identified, such that gain-framing influenced the intention more positively than loss-framing.

(5)

In the last decades, the development of the so-called green economy and sustainable consumption has grown consistently (Nidumolu et al., 2009). Climate change and increasing awareness about the capitalist economy’s dark sides are making consumers more conscious and shifting their lifestyles toward green ones (Cherian & Jacob, 2012). As a result, profit-driven businesses have responded with new marketing strategies such as green advertising (Atkinson & Kim, 2015; Hartmann & Apaolaza-Ibáñez, 2009) and cause-related marketing (Patel et al., 2017). On the other hand, the non-profit sector has seen a development of those charity causes that involve the environment, such as global warming, wildlife extinction and deforestation.

A NPO is “any organization whose main objective is not for profit” (Baraldi, 1998). Thus, non-profit organizations (NPOs) are non-profits driven entities that advocate to a given cause and dedicate their income to it. They rely on donations of individuals, companies and governments. The scopes that these charities face can be several and regard, for instance, human services, education, health, religion, environment and animals. Even if the first aim of a NPOs is not to benefit the owners or the donors, promoting effectively the mission of the organization is fundamental to raise funds and fulfil the cause.

(8)

Environmental and conservation organizations are those NPOs that protect habitats, wildlife, flora and fauna. Even if the environment and animal non-profit sector is one of the smallest non-profit subsectors, for example, in the U.S., it keeps growing since 2005. In the decade 2005-2015 it saw the largest growth rates in the number of organizations, increasing 14.7 percent from 12,721 to 14,591 organizations (McKeever, 2018). The environmental and conservation organizations revenue grew from $13.0 billion to $19.7 billion (McKeever, 2018). In this case, the revenues and assets of the NPOs grew faster than the U.S. GDP, hence the market flourishes and presents opportunities that can be exploited. A way for the NPOs to draw attention to them is through marketing and communication.

Two communication strategies have been commonly used to increase donations and they include message framing (C. T. Chang & Lee, 2009) and nature imagery (Schmuck et al., 2018). Message framing refers to the way the information is shown, it is either positive or negative (gain vs. loss terms). Choosing the right frame increases message persuasiveness (Martin & Marshall, 1999). On the other hand, nature imagery, also called “virtual nature experiences” (Hartmann & Apaolaza-Ibáñez, 2012), consists of the use of pictures of nature in advertising. It is defined as the visual representation of nature, such as unspoiled landscapes background in the advertisements and it can activate feelings. Likewise, message framing can induce positive or negative emotions depending on whether the message is gain- or loss-framed.

Message framing (Levin et al., 1998; Martin, 1995; Pelletier & Sharp, 2008) and nature imagery (Banerjee et al., 1995; Hartmann et al., 2013; Matthes et al., 2014; Matthes & Wonneberger, 2014) have always been investigated within two separate fields of research. The first with regards to communication research, the other to environmental psychology. However, in order to convey a message, all advertisements do it through a combination of verbal and visual elements (Houston et al., 1987). Therefore, the present study will combine these two popular communication strategies because the combination of verbal and visual cues is important to influence people (Spack et al., 2012). In this study, the aim is to examine how these two strategies can encourage donation intentions to environmental NPOs.

(9)

understanding donors’ motivation and intention through additional research is important to develop successful marketing campaigns. Nowadays higher competition due to the increasing number of NPOs, demanding consumers, and decreasing funds from the governments are relevant threats for the non-profit sector. On the other hand, attention on environmental issues is lately growing due to the climate change impact. This study focuses on how NPOs can increase their public donations on their charity cause. More specifically, it investigates which emotions are triggered by message framing and nature imagery and which condition(s) are the most effective in increasing donations for an environment charity cause.

This research looks for answering the following question:

(10)

2. Literature background

2.1 Donation intention of environment charities

As the goal of a non-profit organization is to raise funds, in this study, the donation intention is considered to evaluate the effects of communication strategies that are used in advertising. From past research, the theory of planned behavior (TPB; Ajzen, 1991) claims that an accurate prediction for different types of behaviors can be obtained considering the behavioral intentions. ‘Behavioral intention’ is a concept that has been interpreted differently over the years, but with the conceptualization of the TPB, Ajzen (1991) interpreted the behavioral intention as an indicator of ‘how hard people are willing to try’ and ‘how much of an effort they are planning to exert’. Ajzen (1991) reflects the effort people are going to invest toward a certain behavior. Low intention is translated into low effort, while high intention into high effort.

In this specific case, the intention focuses on whether a person donates to an environmental charity after exposure to an environmental cause advertisement. Donation intention is considered as an important donation predictor (Cheung & Chan, 2000; Fishbein & Ajzen, 2005). Moreover, the relation between donation intention and actual behavior is strong (Kashif et al., 2015). Therefore, considering the donation intention allows to predict actual behavior. As a consequence, intention is the closest proxy to understand how message framing and nature imagery play a role to predict actual donation behavior, and the extent to which emotions might mediate these relationships.

2.2 Message framing in charitable causes: Goal framing

(11)

Message framing research sees its roots in the principles of the prospect theory (Kahneman & Tversky, 1979; Tversky & Kahneman, 1981). The prospect theory explains how people make decisions under risk, and how they estimate the perceived likelihood of a certain option. Kahneman and Tversky (1979) did this by running a series of studies, including one called ‘Asian disease problem’. By this, the researchers found that when participants were exposed to discrete choices that had the same expected value, but different levels of risk (i.e. uncertainty), the choice they made depended on how the options were described. In fact, in one case the terms used in the text were positive (i.e. lives saved) while in the other they were negative (i.e. lives lost). More specifically, when the task was positively framed, people tended to select the choice with a certain outcome (i.e. 100% chance to save a certain number of people instead of a lower chance of saving a higher number of people). Conversely, when the task was negatively framed participants chose the risky option. For instance, participants preferred to go for one-third chance of losing nobody and a two-thirds chance of losing everybody instead of 100% chance of a loss of two-thirds of the people. The findings imply that people tend to take risks when the focus is on the losses and to be risk-averse when the message is focused on the gains (Kahneman & Tversky, 1979). Because people tend to dislike losses more than they like gains, the results showed that the final outcome can be influenced by using message framing to manipulate the reference point people base their choice on.

People tend to be risk- or loss-averse, thus they tend to dislike losses more than equivalent gains. Loss-aversion biases our choices outweighting the probabilities of uncertain events, hence people tend to avoid losses more than they like achieving equivalent gains. Kahneman and Tversky (1979) used the prospect theory to explain the risk decision-making process and showed it in a S-shaped utility function. First, people do not think in terms of absolute values, but of expected utility compared to a reference point which is subjective (e.g. current wealth of the individual). The reference point coincides with a null value of losses and gains. Secondly, the shape of the curve, which is concave for the gains and convex and steeper for losses, shows how making decisions with regards to losses involves a bigger difference in value than it does for gains.

(12)

prospect theory (Fagley & Miller, 1987, 1990; Frisch, 1993; Kühberger, 1995; Schoorman et al., 1994). Levin, Schneider, and Gaeth (1998) elaborated on the cause of these mixed results that was identified in the wrong fit between type of framing manipulation considered and theory used to explain the results. In fact, they debate that past studies were actually using different types of message framing and choices, but they were trying to explain the results only by means of the prospect theory, which in turn is mainly focused on risk and loss aversion. Levin et al. (1998) argue that different framing manipulations and relative underlying processes exist, and they are characterized by different effects that cannot always be explained by the prospect theory. Levin, Schneider, and Gaeth (1998) distinguished among three types of framing manipulations: risky choice, attribute, and goal framing. This distinction allows to better know and locate the type of message framing used in this study in one of these categories, thus better interpret the results.

First, the so-called risky choice framing is the one studied by Kahneman and Tversky (1979) and that is explained above. It involves using frames to influence a choice between a set of independent options which have different levels of risk. In this case the overall effect consists of being more likely to take risks when the choices focus on the chance to avoid losses than when the focus is on the chance of obtaining gains. Second, attribute framing is when “some characteristic of an object or event serves as the focus of the framing manipulation” (Levin et al., 1998). Differently from risky choice framing, it is related to an evaluation process about a characteristic of an object or event and either evaluation is complementary to the other. For example, the evaluation includes favorability ratings such as evaluating through scales from bad to good, or express yes or no judgements (i.e. being in favor or not of a certain program). Levin et al. (1998) identified two main reasons of the why attribute framing differs from risky choice. First, a single characteristic is framed instead of more independent options. Second, attribute framing choices do not involve riskiness. Indeed, the effects of attribute framing differ from the prospect theory predictions, such that positive attributes produce more favorable evaluations than negative ones (Krishnamurthy et al., 2001; Levin et al., 1998). The last identified category of message framing is goal framing, which is also the one that fits best with charitable messages.

(13)

simplest version of goal framing considers as positive condition a message that focuses on the potential gains of involving in the behavior, while as negative a condition that focuses on the losses that can be avoided by involving in the behavior (Levin et al., 1998). Both of the framing conditions try to convince to involve in the same behavior by either presenting the gains (positive framing) or the avoided losses (negative framing). Moreover, Levin et al. (1998) identified additional variations of goal framing. For instance, in a study about breast self-examinations run by Meyerowitz and Chaiken (1987), the positive framing included either the gains or the avoided losses of taking action. On the other hand, the negative condition included the losses or the missed gains of not taking action. Meyerowitz and Chaiken (1987) did not define the message framing used as ‘goal framing’, but it was later identified by Levin et al. (1998). What Meyerowitz and Chaiken (1987) found is that when people are facing eventual losses they try to avoid them. This finds foundation in the utility function of the prospect theory and the fact that people are generally loss averse. Furthermore, Levin et al. (1998) argued that goal framing variations do not change the final outcome: loss framed messages are more persuasive than gain framed ones because people tend to avoid the losses, that are shown to them, by involving in the behavior.

(14)

Negative frames revealed to be more effective on donation intention. Exposure to negative appeals (i.e. loss-framing or something unpleasant) is of overall greater impact on the observer’s psychological state as supported by the negativity bias theory (Martin, 1995). Research with regards to donations provide support to the greater impact of negative framing in the charity domain (C. T. Chang & Lee, 2009, 2010; Chou & Murnighan, 2013; Erlandsson et al., 2018; Sarstedt Marko & Schloderer Matthias Peter, 2010). For instance, Chang and Lee (2009, 2010) dig into the charitable framing effects of child poverty causes. Erlandsson et al. (2018) studied actual donation behavior. Also, Chou and Murnighan (2013), in their study about blood donations, showed how people volunteered more by donating blood when the message was promoting the behavior as a way to prevent deaths (negative frame) rather than saving lives (positive frame). As a result, the hypothesis with regards to message framing and donation intention is the following:

H1. Advertisements with loss-framed message influence the donation intention for an

environmental charity more positively than advertisements with gain-framed message.

2.3 The role of emotions

Ads’ elements can influence observers’ emotions. Emotion phenomena is “an organized psychological reaction to ongoing relationships with the environment” (Lazarus, 2000). The communication strategies used in advertising can arouse emotions in the public and they also have been found influencing decisions (Faseur & Geuens, 2012). For example, positive framing can arouse emotions such as happiness, contentment, and hopefulness, while negative framing appears to stimulate other emotions such as sadness, guilt, anger, and fear (Cho & Boster, 2008; Erlandsson et al., 2018; Shen & Dillard, 2007; Urbonavicius et al., 2019; Van ’T Riet et al., 2010). In advertising, emotions can mediate the effects of the communication strategies (Albouy, 2017; Holbrook & Batra, 1987) and it is difficult to find scenarios concerning donation decisions where emotional elements are not present (Polonsky & Sargeant, 2007).

(15)

(Erlandsson et al., 2018). Negative appeals can induce more negative emotions, which are in general more easily induced (Baberini et al., 2015) and, in turn, are also shown influencing donations more than positive emotions (Albouy, 2017; Burt & Strongman, 2005; Erlandsson et al., 2018). Indeed, the prospect theory (Kahneman & Tversky, 1979) explains that it is more important to avoid negative consequences than equally positive ones. In fact, the willingness to avoid negative emotions increases people’s intention to donate (Stets & Carter, 2012). As a result, in this study, the hypothesis that is going to be tested is the following:

H2. Emotions mediate the relationship between message framing and intention to donate to

an environment charity cause, such that negative emotions will affect the donation intention more positively.

2.4 Message framing and nature imagery: The congruency effect

Not only message framing can affect people’s behaviors, additional variables should be considered as they can affect the influence of message framing (Latimer et al., 2007). Marketing communication rely on the combination of both verbal (e.g. written messages) and nonverbal (e.g. imagery) elements to convey a message and to attract people’s attention (Houston et al., 1987). In fact, advertising can also benefit from visual imagery such as persuading people better (Seo & Dillard, 2019) and enhancing the memorability of an ad (Graber, 1990; Nelson et al., 1976). Thus, combining verbal and visual cues can affect positively the effectiveness of an advertisement.

(16)

greater when the image valence and the message of the ad are congruent and especially when the frame is negative (i.e. loss condition)(Chang & Lee, 2009). Thanks to congruency between an environment cause and the imagery, the emotional response of the ad can be increased. The association of message and image to the environment can influence the public’s cognitive response more (Hartmann et al., 2016).

In this study, a specific type of imagery called nature imagery is going to be used. In the last decades, thanks to the spread use of green advertising this type of imagery has seen a continuous growth (Banerjee et al., 1995; Matthes & Wonneberger, 2014; Matthes et al., 2014). Green advertising is a marketing strategy that promotes products through environmental claims (Prakash, 2002). For firms, the relevance of this strategy has increased over the years (Nidumolu et al., 2009). However, green advertising is not the only domain in which nature imagery is used. For example, images of nature are commonly used in today’s the environmental non-profit sector. In fact, NPOs use nature and climate change pictures to convey their environmental messages (O’Neill & Smith, 2014). Nature imagery is a visual representation of nature such as unspoiled landscape backgrounds in the advertisements and it can activate feelings that the watcher would experience in a real situation. Hartmann and Apaolaza-Ibáñez (2012) also call it “virtual nature experiences”.

(17)

with vegetation, but savanna, evoked the same emotions felt in real nature. On the other hand, the rocky desert and the city scenery results were significantly different from the real nature condition ones. Overall, the results showed that pleasant nature imagery aroused positive emotions in the observers, while unpleasant nature imagery negative ones.

H3. Emotions are more influenced by advertisements that show congruent message framing

and nature imagery (i.e. pleasant nature with gain framing, unpleasant nature with loss framing).

2.5 Conceptual framework

(18)

3. Methodology 3.1 Research design

In order to investigate whether message framing and nature imagery influence people’s emotions and donation intention for an environment charity cause, a 2 (gain- vs. loss-framed messages) x 2 (pleasant vs. unpleasant nature imagery) between subjects’ experiment run in the form of an online questionnaire. Four different ads were developed so that all of the combinations of message framing and nature imagery could be considered. The conditions were as following: (1) gain-framed message and pleasant nature image, (2) gain-framed message and unpleasant nature image, (3) loss-framed message and pleasant nature image, and (4) loss-framed message and unpleasant nature image. The participants were randomly assigned to one of the four experimental conditions (see Table 1).

Table 1: Ad conditions and number of participants

Condition Message framing Nature imagery Participants (n)

1 Gain Pleasant 67 2 Gain Unpleasant 66 3 Loss Pleasant 70 4 Loss Unpleasant 66

3.2 Participants

Two hundred and seventy participants completed the survey. The questionnaire was shared with the authors’ contacts through personal social media profiles (i.e. Facebook, LinkedIn, Instagram, and WhatsApp). Additionally, it was asked the participants to share the link of the questionnaire with their contacts. The study did not require a pre-defined target population; thus, it was opted for two nonprobability sampling techniques to reach a larger number of participants: the convenience sampling and the snowball technique (Malhotra, 2010). In fact, the techniques allowed to gather the sample for the study in two days since the survey was opened, therefore the analysis of the dataset started.

(19)

and because it was the only demographic question where they could not choose the ‘I’d rather not to say’ answer. Also, information about age was gathered with the aim of describing the sample, thus there was no reason of excluding these participants from the survey. Additionally, with regards to age, a value lower than 18 was found (age = 9), so the whole observation was deleted. No other missing values or outliers were found because the survey was constructed in a way that answers were either Likert-scales or multiple choices subjected to force response in order to submit the survey. Therefore, the final number of participants considered in this study was 269. This size was appropriate according to some rules of thumb that are presented below.

An ideal sample size was calculated beforehand so that it was possible to know whether the study would incur in power issues. In the case of t-test or ANOVA, the minimum number of participants required for each condition was thirty people (Cracraft, 1988), hence a total of 120 for this study. However, because the study was also including covariates, additional 30 participants per covariate were added in the estimation, thus reaching a minimum of 150 participants since the study was going to include one covariate. Additionally, in order to run a factor analysis, a sample of around 50 participants per factor is considered appropriate (Day et al., 1994). As a result, it was foreseen that a maximum of four factors would be possible based on the items of the survey and what they were referring to, hence at least 200 participants should be reached. Finally, in the case of regression the minimum sample size was way lower than the sized mentioned above. Therefore, the predictions were met and overcame with a total of 269 participants and around 67 participants per condition (see Table 1).

(20)

employees. In conclusion, the sample obtained was mainly European, relatively young, and well-educated.

3.3 Pre-test

A pre-test was run in order to select the two pictures (i.e. a pleasant and an unpleasant one) that were going to be used in the actual survey. Also, the pre-test was done to understand if the message was properly conveyed by the ad, and to find possible mistakes. Pre-testing refers to ‘the testing of the questionnaire on a small sample of respondents to identify and eliminate potential problems’ (Malhotra, 2010, p.354). Following what Malhotra (2010) suggests, the tested aspects include ‘the question content, wording, sequence, form and layout, question difficulties and instructions’ (Malhotra, 2010, p.354). As a result, the pre-test was created with Qualtrics and was available in two languages such as the final survey. Also, it was sent to a small number of Italian, Dutch and International author’s friends because the participants should be as similar as possible to the actual population of the survey (Malhotra, 2010, p.354).

First, the six pictures were selected for the pre-test (see Appendix A). Three pictures were considered ‘pleasant’ and three ‘unpleasant’ based on Hartmann et al. (2013). More specifically, the three pleasant pictures represented a lake and forest, a mountain stream, and a forest. The first two pictures mentioned were chosen because of the direct resemblance of the pictures used in the Hartmann’s et al.’s Study 1 that had the highest rates of positive emotions. Indeed, these pictures were considered as ‘pleasant’ by the researchers. Instead, the ‘forest’ picture was added because of the lack of other pictures from the Hartmann et al.’s Study 1 of imageries that could allow to test three pleasant and three unpleasant pictures. In fact, other pictures considered by Hartmann et al. (2013) were representing the savanna or the Mediterranean coast which did not make sense with the deforestation cause. Therefore, a photo of a forest was added.

(21)

In order to create a pre-test that was as similar as possible to the actual survey, the pictures tested included the logo of the fictious non-profit organization (i.e. Nature Fund) and the message of the campaign (see Appendix A). These elements were positioned following the layout of the ads used in the study of Hartmann et al. (2013), namely a logo positioned in the bottom right corner of the nature picture, and the text in the bottom part of the ad. Also, the position of the picture, the text, and the logo was the same for all of the ads to reduce possible different effects of these variables. With regards to the messages used, they were inspired to the study of Chang and Lee (2009) and were a gain- and a loss-framed message. First, the gain-framed message was as follows: “With your help, Nature Fund will be able to plant

new trees worldwide and save forests from deforestation. In this way, the precious nature, animals and the climate of the world we live in will be preserved”. Conversely, the

loss-framed message was as follows: “Without your help, Nature Fund won’t be able to

fight against the aggressive deforestation worldwide. The planet we live in will be destroyed by irreparable climate changes and environmental damages”. The loss-framed

message was matched with the ads showing the unpleasant pictures, while the gain-framed message with the pleasant ones. This choice aimed at avoiding showing too many conditions to each participant. In fact, this study was a full-factorial within-subjects design and each participant had to watch and answer questions about six ads. The survey’s questions are shown in Appendix B.

(22)

congruency was perceived. Afterwards, demographic questions such as gender, nationality, and age followed. Also, an additional open question was added to ask participants whether they noticed any mistake and/or they wanted to provide any suggestion.

Twenty-three participants finished the pre-test survey. No missing values and outliers were found. The 30.43% were men and the 69.57% were women. The age ranged from 22 to 69 years old (M=32.13, SD=12.08). Italians were the 60.87% of the sample, while Dutch participants followed with the 26.09% and 13.04% were of other nationalities. The minimum time of 15 seconds provided to watch the ad was considered enough by the 78.26% of the sample, while only the 21.74% thought it was too much. Therefore, this information was used to set a minimum time of 15 seconds to view the ad in the survey. Three paired-samples t-tests were run to investigate the mean differences of message framing, nature imagery, and emotive response to the ads. The results of the paired-samples T-tests (see Appendix C) showed that the pictures, the message frames, and the emotive responses were different between the three ‘positive’ ads (i.e. for pleasant imagery and gain-framed message the rates to the statements were higher) and the three ‘negative’ ads (i.e. unpleasant imagery and loss-framed message the rates to the statements were lower). As a result, the pictures for the survey were chosen. Soil erosion was the least pleasant nature imagery and with a more negative emotive response (pleasantness, M=2.00, SD=0.93; happiness, M=2.04, SD=1.20). On the other hand, the ‘lake and forest’ imagery was the most pleasant and with a more positive emotive response (pleasantness, M=6.26, SD=0.74; happiness, M=5.61, SD=1.13).

3.4 Materials

As anticipated in the previous section, the materials were first tested, analyzed, and selected from the pre-test. The two advertisements used in the experiment were the picked from the pre-test and they were created with Canva.com.

3.4.1 Charity cause

(23)

message (Bendapudi et al., 1996). In order to confirm this, in the survey the participants were asked to rate on a 7-points Likert scale (1=strongly disagree, 7=strongly agree) the following statements:

- ‘Nature Fund' is likely to be the name of a real charity; - The 'Nature Fund' logo is likely to be the logo of a charity.

3.4.2 Manipulation of independent variables

The independent variables were manipulated, thus the conditions that are shown below were created and participants assigned to one of them. The variables manipulated in this study were message the framing and nature imagery. First, the text was framed either positively (i.e. gain-framing) or negatively (i.e. loss-framing). The message was inspired to the two conditions used in the study of Chang and Lee (2009), hence using “With your help” followed by positive consequences of donating in the gain-frame, and “Without your help” followed by negative consequences of donating in the loss-frame. The two messages were constructed in a way that they would be similar in the form such as the length, in the structure (i.e. two sentence), and in the goal promoted. However, they differed in the framing used, therefore in whether the stress was on the positive (negative) consequences of donating (not donating). The messages were the same used in the pre-test as they were perceived as significantly different from each other and in the way that was expected depending on the framing used. Second, nature imagery was manipulated so that it had two levels: pleasant and unpleasant. As explained in the previous section, a pre-test was run and data about six pictures gathered and analyzed. The four conditions’ materials consisted in the following ads:

(24)

3. Gain framing & unpleasant nature 4. Loss framing & unpleasant nature

3.5 Procedure

The experiment was run with the use of an online survey created on Qualtrics.com (See Appendix D). The questionnaire was shared using a link that was distributed on the author’s social medias such as Facebook, LinkedIn, WhatsApp, and Instagram. Also, the people contacted were asked to share the link with their network. Because the author’s network was mainly Italian, but also Dutch and international, the survey was provided into two languages: Italian and English. The two languages allowed to share the questionnaire with more people and gather information from more countries about the same topic.

(25)

3.6 Measures

3.6.1 Dependent variable: Donation intention

The scale used to measure the donation intention is based on Merchant et al. (2010) and Ranganathan and Henley (2008). It was measured with a 7-point Likert scale (1=strongly disagree, 7= strongly agree) and consisted of three items, namely:

1. ‘Seeing this advertisement makes me willing to donate to the charity’. 2. ‘If asked I would certainly donate to this charity’.

3. ‘I definitely want to donate to this charity in the future’.

3.6.2 Manipulation check: Message framing, nature imagery, and emotions

Participants had to answer questions about message framing and nature imagery in order to identify whether the manipulation was successful. Because message framing and nature imagery were manipulated, a manipulation check of those variables was included in the survey. The manipulation check of message framing was drawn from the study of Chang and Lee (2009) and partially adapted to this study. Hence, respondents were asked to rate on a 7-point scale (1=very negative, 7=very positive) the tone of the text of the ad. Additionally, the nature imagery manipulation check consisted of asking participants to rate the pleasantness of the nature picture showed in the background on a 7-point Likert scale (1=very unpleasant, 7=very pleasant).

3.6.3 Emotions

Emotions measure was derived from the study 1 of Hartmann et al. (2013). Participants had to rate their emotional response to the ad on six 7-point bipolar scales. The sixth item pair was added since the role of guilt is relevant in charity communication research (C. Chang, 2014). The six items included were: aroused – relaxed, unpleasant – pleasant, sad – happy, oppressed – free, unsecure – secure, hopeless – hopeful, guilty – innocent.

3.6.4 Demographics

(26)

3.6.5 Confounding variable: Donation experience Donation experience

(27)

4. Results

4.1 Plan of analysis and pre-hypotheses testing results

4.1.1 Manipulation check: independent t-test and assumptions

Two independent samples t-tests were conducted to assess whether the manipulations of message framing and nature imagery were successful. One test was conducted for message framing and another for the manipulation check of nature imagery. However, first, there were several assumptions that the data had to meet to run the independent t-test:

1. The scale of measurement of the data gathered had to be continuous or ordinal. This assumption was met because the variables for the manipulation check were measured with a 7-point Likert scale, which is ordinal.

2. All of the observations had to be independent from each other. This assumption was met since the participants were answering the survey individually and each observation was considered individually.

3. Data had to be normally distributed. A test of normality on both variables was run. Both the Shapiro-Wilk and Kolmogorov-Smirnov tests showed that the data was not normally distributed (see Appendix E), thus the assumption was violated for both variables.

4. The last assumption was the homogeneity of variances. This was tested with the Levene’s test of equality of variances. The assumption was met by message framing (p=.209), but not by nature imagery (p=.006).

(28)

4.1.2 Principal component analysis

A principal component analysis (PCA) was run to reduce the large set of data into a smaller one by finding common variance. The aim was to summarize a number of items into a smaller number of ‘principal factors’ such that the analysis and interpretation would be easier. The final result would be a certain number of uncorrelated factors that are composed by two or more items correlated to each other. The fact that factors would be uncorrelated allows to reduce multicollinearity in the further analyses.

First, in order to proceed with the PCA it was necessary to check whether the data were adequate. Therefore, the Kaiser-Meyer-Olkin (KMO) measure and the Barlett’s test were calculated (see Appendix F). The KMO, that was higher than 0.05, showed that the data were appropriate to proceed with the analysis. Second, the value of the Bartlett’s test of sphericity was significant (p<.05), hence it was possible to reject the null hypothesis (H0: variables are uncorrelated) and proceed with the analysis. Lastly, communalities were also checked and a common rule of communalities > .4 was followed. It has to be noted that all of the items but the item ‘guilty – innocent’ (communalities = .408) showed values way above the cut-off point of .4 (see Appendix F, Communalities). However, even if the value of ‘guilty – innocent’ was very close to .4, it was kept in the analysis as the value was still over 0.4.

(29)

Table 2: Rotated component matrix

Component

1 2 3 4

Seeing this advertisement makes me willing to donate to the charity.

.014 .863 .072 .202

If asked I would certainly donate to this charity.

-.003 .894 .164 .165

I definitely want to donate to this charity in the future.

.017 .880 .160 .128 Aroused - Relaxed _.759 _-.089 _-.145 _-.129 Unpleasant - Pleasant _.820 _.032 _-.065 _-.053 Sad - Happy _.760 _-.060 _.080 _.015 Oppressed - Free _.786 _.079 _-.005 _.116 Unsecure - Secure _.731 _.024 _.002 _.004 Hopeless - Hopeless _.710 _.262 _.039 _-.019 Guilty - Innocent _.619 _-.118 _.078 _.068

I frequently donate to charities. _.060 _.232 _.893 _.053

I've donated money to one or more charities in the past 12 months.

-.048 .109 .915 .019

'Nature Fund' is likely to be the name of a real charity.

.027 .231 .019 .864

The 'Nature Fund' logo is likely to be the logo of a charity.

-.013 .185 .051 .883

4.1.3 Reliability analysis and variables computation

(30)

results, it was chosen to calculate and compare the Composite Reliability of each composite score as well. A general rule of thumb for Composite Reliability and Cronbach’s Alpha is a score higher than 0.70. The results in Table 3 show the scores that were calculated for all of the items of ‘factor’. The names of the new variables are already shown in the table. All of the variables could be computed because of good reliability scores.

Table 3: Reliability statistics

Component N of Items Composite Reliability Cronbach's Alpha Difference Internal consistency Donation intention 3 .911 .894 .017 ✓ Emotions 7 .896 .864 .032 ✓ Donation experience 2 .899 .808 .091 ✓ Credibility of the fictious brand 2 .866 .770 .096 ✓

4.1.4 Moderated mediation: Assumptions and analysis

Because this researched aimed to study a moderated mediation, it was decided to run the analysis using the Model 7 of PROCESS v3.5 by Hayes (2018). More specifically, the model 7 allowed to understand whether there was a moderation of ‘nature imagery’ on the relation between ‘message framing’ and ‘emotions’, but also to understand whether there was an indirect mediation effect of ‘emotions’ between message framing and donation intention. However, there are some assumptions of regression that need to be met to run an analysis with PROCESS. A more detailed output of the assumptions’ tests is shown in Appendix G. The assumptions were as follows:

1. Continuous or interval scales of the dependent variables.

This assumption was met as emotions were measured on a 7-point bipolar scale, and donation intention on a 7-point Likert scale.

2. Independence of observations.

(31)

3. No outliers.

This assumption was met for all of the data. The Cook’s distance was used to understand whether outliers were present. The Cook’s distance was largely lower than one in both emotions (Cook’s=.026) and donation intention (Cook’s=.023) (see Appendix G).

4. Normal distribution of the residuals of regression.

This assumption was tested using the Shapiro-Wilk test, such that a not significant value (Sig.>.05) meant the distribution was normal. Emotions met the assumption, but donation intention met the assumption only in the case of gain framing.

5. Homoscedasticity of the residuals.

This assumption was difficult to test by only looking at the regression screeplot of the binary nature of the predictor. Therefore, it was decided to proceed with the Levene’s test. The test showed that donation intention met the assumption, while emotions not.

6. Absence of multicollinearity.

The VIF scores showed that multicollinearity was not an issue. The VIF scores were both largely lower than 10 (VIF=1).

Afterwards, the main analysis without the covariate was run (see Appendix H). A second analysis was run using the same model but including donation experience as a covariate (see Appendix I).

4.2 Hypotheses testing

In order to test the hypotheses of this study, a moderated mediation was run (see Appendix H). More specifically, the model used for the analysis was the model 7 (bootstrap 5000 samples) of PROCESS macro (Hayes, 2018). The variables included in the model 7 are listed below:

- Predictor – Message framing. - Moderator – Nature imagery. - Mediator – Emotions.

(32)

H1: Advertisements with loss-framed message influence the donation intention for an environmental charity more positively than advertisements with gain-framed message.

The main analysis, run with PROCESS (Model 7), showed that there was a direct relation between message framing and donation intention. The p-value of message framing was significant (see Table 4). In order to visualize the means thus the effect better for each level of the independent variable, a one-way ANOVA was run. The result was significant (see Appendix I) and confirmed the result of the regression, namely: donation intention was lower in the case of loss-framing (see Table 5 and Figure 2). Concluding, the hypothesis 1 could not be supported.

Table 4: Direct effect of message framing on donation intention (main analysis) Dependent variable: Donation intention

Model: R-sq.=.0238, p=.0408

Predictor Coefficient p

Message framing .4177 .0139

Emotions .0098 .8830

Table 5: One-way ANOVA results Condition message framing Mean donation intention SD N Loss 3.598 1.42530 136 Gain 4.020 1.28953 133

Figure 2: Means of donation intention per message framing condition 1.00 2.00 3.00 4.00 5.00 6.00 7.00 Loss Gain

(33)

Additionally, a second analysis was run with PROCESS (Model 7) but including the covariate ‘donation experience’. The covariate showed a significant effect only on donation intention as was expected (see Table 6).

Table 6: Effect of donation experience on donation intention Dependent variable: Donation intention

Model: R-sq.=.1170, p=.000

Donation experience .2243 .000

H3: Emotions are more influenced by advertisements that show congruent message framing and nature imagery (i.e. pleasant nature with gain framing, unpleasant nature with loss framing).

The results of the moderated mediation run with PROCESS showed a significant interaction effect of message framing and nature imagery. In Table 7, the list of the effects on emotions are shown. However, since the interaction was significant (p=.0355), only this effect was considered instead of single effects. In Figure 3 the emotions mean scores are shown for each of the four conditions. It can be seen that the congruency of message framing and nature imagery resulted in the lowest score in case of loss and unpleasant elements, while in the case of gain and pleasant elements resulted in the highest emotions score. Therefore, the hypothesis could be confirmed.

Table 7: Moderation of nature imagery on emotions (main analysis) Dependent variable: Emotions

Model: R-sq.=.2649, p=.000

Nature imagery .9139 .0000

Interaction: message framing

(34)

Figure 3: Emotions responses per condition (mean scores)

H2: Emotions mediate the relationship between message framing and intention to donate to an environment charity cause, such that negative emotions will affect the donation intention more positively.

In order to assess the presence of mediation, two effects needed to be significant. First, the interaction effect of message framing and nature imagery on emotions. Second, the effect of emotions on donation intention. As shown in the previous results, first, the interaction effect of message framing and nature imagery on emotions was significant (see Table 7). However, a direct effect was detected between message framing and donation intention while testing the first hypothesis (see Table 4). This already suggested the lack of a significant indirect effect. In fact, the effect of emotions on donation intention was not significant (p=.8830). Therefore, there is no mediation of emotions. An overview of the effects and hypotheses is shown in the following page.

1.00 2.00 3.00 4.00 5.00 6.00 7.00 Loss Gain

Mean scores of emotions per condition

(35)

4.3 Overview of the effects

4.4 Overview of the (not) supported hypotheses

Hypotheses Results

H1: Advertisements with loss-framed message influence the donation intention for an environmental charity more positively than advertisements with gain-framed message.

Not supported

H2: Emotions mediate the relationship between message framing and intention to donate to an environment charity cause, such that negative emotions will affect the donation intention more positively.

Not supported

H3: Emotions are more influenced by advertisements that show congruent message framing and nature imagery (i.e. pleasant nature with gain framing, unpleasant nature with loss framing).

(36)

5. Discussion 5.1 Conclusions

The present study aimed at answering the following question: “How do message framing

(gain vs. loss) and nature images (pleasant vs. unpleasant) influence the emotions and the consequent donation intention of an environmental cause promoted by an NPO?”. The study

examined the effect of message framing (gain vs. loss) and imagery (pleasant vs. unpleasant) on emotions and donation intention of an environmental charity cause. The expectations were that loss-framed message would be more effective than gain-framed message in increasing donation intention, congruency between type of message framing and nature imagery would arouse stronger emotional responses, and finally, that emotions would mediate the relationship between message framing, nature imagery, and donation intention. These hypotheses were only partially confirmed, in fact, the findings supported only the stronger effect on emotions of ads that were composed of congruent message framing and nature imagery (i.e. gain with pleasant, loss with unpleasant). Also, a direct effect of message framing on donation intention was found. However, the mentioned effect was contrary to the one hypothesized. Lastly, emotions mediation was not found, therefore no indirect effect could be assessed between message framing and donation intention.

Based on the loss-aversion theory (Kahneman & Tversky, 1981; Tversky & Kahneman, 1979) and on literature about message framing (C. T. Chang & Lee, 2009, 2010; Chou & Murnighan, 2013; Erlandsson et al., 2018; Sarstedt Marko & Schloderer Matthias Peter, 2010) the expectation was that loss-framed messages would be more effective at influencing the donation intention. People exposed to a loss-framed message should tend to feel worse about the possible loss and tend to involve in the donation behavior to avoid a possible loss. However, this was not the case. One possible explanation to this can be that the loss-framed message discouraged the public from believing that saving the forests was possible. Also, it is possible that the loss-framed message backfired because perceived as manipulative.

(37)

intention is not possible. An evaluation of the final effect on donation intention would have given information about whether the message was conveyed better with congruent ads.

5.2 Contributions

This study contributed to research in different ways. First, the hypothesis about the major effect of loss-framing on donation intention was not supported, however, the result suggested that gain framing could be more effective in promoting environmental charity. It contributed by expanding specifically the research about environmental charity as it has not been extensively researched. Additionally, the results about the combination of message framing and nature imagery brought to an interesting remark: the two ads that included pleasant nature imagery showed the highest rates of the emotions studied, even in the case of loss-framed message. When promoting an environmental cause, NPOs can take into consideration that gain-framing worked better in increasing donation intention, but also, that pleasant nature imagery is willing to receive a positive emotional response from the public. This could be relevant depending on the intent of the organization when promoting a certain cause.

5.3 Limitations and future research

This study comes with some limitations. The first limitation regards the sampling technique used. The sampling technique used was convenience sampling, which implies that the results cannot be generalized to a whole population. Also, it could bring to biased results. Therefore, there is a chance that the results of this study could not be reliable. Because of time limitations, in this case convenience sampling was appropriate, however, future research could assess this by using a nonrandom sampling technique, so that the results could be generalized and more reliable.

(38)

Hartmann’s studies focus mainly on the effect of pleasant imagery on people’s emotional responses and consequently on products’ advertising effectiveness. Future research could investigate on the effects of more and different unpleasant nature imagery pictures, but also on their effects specifically in charity advertising.

Third, the scale used to measure emotions was taken from Study 1 of Hartmann et al. (2013), but the study mainly focused on the emotional response to nature imagery, and considered scales that regarded two opposite emotions put together. This could have brought to a more unprecise measurement of the emotions, and possibly to a lack of mediation. Past research showed a mediating effect of emotions, therefore future research could narrow down the consideration of emotions either by dividing them into positive and negative emotions or by selecting a number of emotions that are relevant to message framing and charity.

(39)

References

Ajzen, I. (1991). The theory of planned behavior. Organizational Behavior and Human

Decision Processes. https://doi.org/10.1016/0749-5978(91)90020-T

Albouy, J. (2017). Emotions and prosocial behaviours: A study of the effectiveness of shocking charity campaigns. Recherche et Applications En Marketing, 32(2), 4–25. https://doi.org/10.1177/2051570716689241

Arora, R. (2000). Message Framing and Credibility. Health Marketing Quarterly. https://doi.org/10.1300/j026v18n01_03

Atkinson, L., & Kim, Y. (2015). “I drink it anyway and i know i shouldn’t”: Understanding green consumers’ positive evaluations of norm-violating non-green products and misleading green advertising. Environmental Communication.

https://doi.org/10.1080/17524032.2014.932817

Baberini, M., Coleman, C. Lou, Slovic, P., & Västfjäll, D. (2015). Examining the Effects of Photographic Attributes on Sympathy, Emotions, and Donation Behavior. Visual

Communication Quarterly. https://doi.org/10.1080/15551393.2015.1061433

Balliet, D., Parks, C., & Joireman, J. (2009). Social value orientation and cooperation in social dilemmas: A meta-analysis. Group Processes and Intergroup Relations. https://doi.org/10.1177/1368430209105040

Banerjee, S., Gulas, C. S., & Iyer, E. (1995). Shades of green: A multidimensional analysis of environmental advertising. Journal of Advertising.

https://doi.org/10.1080/00913367.1995.10673473

Baraldi, S. (1998). Management Control Systems in NPOs: An Italian Survey. Financial

Accountability and Management, 14(2), 141–164.

https://doi.org/10.1111/1468-0408.00056

Barnett, J., Hammond, S., Taking, R., Threats, F. O. R., Seo, K., Dillard, J. P., & Shen, F. (2013). The Effects of Message Framing and Visual Image on Persuasion. Journal of

Community and Applied Social Psychology, 61(4), 309–314.

https://doi.org/10.1002/(SICI)1099-1298(199907/08)9:4<309::AID-CASP515>3.0.CO;2-7

Bendapudi, N., Singh, S. N., & Bendapudi, V. (1996). Enhancing helping behavior: An integrative framework for promotion planning. Journal of Marketing.

https://doi.org/10.2307/1251840

Bennett, R. (2003). Factors underlying the inclination to donate to particular types of charity.

International Journal of Nonprofit and Voluntary Sector Marketing.

https://doi.org/10.1002/nvsm.198

Burt, C., & Strongman, K. (2005). Use of images in charity advertising: Improving donations and compliance rates. International Journal of Organisational …, 8(8), 571–580.

http://www.usq.edu.au/extrafiles/business/journals/HRMJournal/InternationalArticles/V olume 8/Burt Vol 8 no 8.pdf?origin=publication_detail

Chang, C. (2014). Guilt regulation: The relative effects of altruistic versus egoistic appeals for charity advertising. Journal of Advertising, 43(3), 211–227.

(40)

Chang, C. T., & Lee, Y. K. (2009). Framing charity advertising: Influences of message framing, image valence, and temporal framing on a charitable appeal. Journal of Applied

Social Psychology, 39(12), 2910–2935.

https://doi.org/10.1111/j.1559-1816.2009.00555.x

Chang, C. T., & Lee, Y. K. (2010). Effects of message framing, vividness congruency and statistical framing on responses to charity advertising. International Journal of

Advertising, 29(2), 195–220. https://doi.org/10.2501/S0265048710201129

Cherian, J., & Jacob, J. (2012). Green marketing: A study of consumers’ attitude towards environment friendly products. Asian Social Science.

https://doi.org/10.5539/ass.v8n12p117

Cheung, C. K., & Chan, C. M. (2000). Social-cognitive factors of donating money to charity, with special attention to an international relief organization. Evaluation and Program

Planning. https://doi.org/10.1016/S0149-7189(00)00003-3

Cho, H., & Boster, F. J. (2008). Effects of gain versus loss frame antidrug ads on adolescents.

Journal of Communication. https://doi.org/10.1111/j.1460-2466.2008.00393.x

Chong, D., & Druckman, J. N. (2007). A theory of framing and opinion formation in competitive elite environments. Journal of Communication, 57(1), 99–118. https://doi.org/10.1111/j.1460-2466.2006.00331.x

Chou, E. Y., & Murnighan, J. K. (2013). Life or Death Decisions: Framing the Call for Help.

PLoS ONE, 8(3), 1–6. https://doi.org/10.1371/journal.pone.0057351

Chowdhury, R. M. M. I., Olsen, G. D., & Pracejus, J. W. (2008). Affective responses to images in print advertising. Affect integration in a simultaneous presentation context.

Journal of Advertising. https://doi.org/10.2753/JOA0091-3367370301

Coeterier, J. F. (1983). A photo validity test. Journal of Environmental Psychology. https://doi.org/10.1016/S0272-4944(83)80034-6

Cracraft, J. (1988). Early evolution of birds. Nature, 331(6155), 389–390. https://doi.org/10.1038/331389a0

Day, S., Pedhazur, E. J., & Schmelkin, L. P. (1994). Measurement, Design and Analysis: An Integrated Approach. The Statistician. https://doi.org/10.2307/2348155

Dunlap, R. E., Van Liere, K. D., Mertig, A. G., & Jones, R. E. (2000). Measuring

endorsement of the new ecological paradigm: A revised NEP scale. Journal of Social

Issues. https://doi.org/10.1111/0022-4537.00176

Erlandsson, A., Nilsson, A., & Västfjäll, D. (2018). Attitudes and Donation Behavior When Reading Positive and Negative Charity Appeals. Journal of Nonprofit and Public Sector

Marketing, 30(4), 444–474. https://doi.org/10.1080/10495142.2018.1452828

European Commission. (2014, January 15). Major European event to shape the future for social businesses. European Commission press release. Retrieved from

https://ec.europa.eu/commission/presscorner/detail/en/IP_14_20

Fagley, N. S., & Miller, P. M. (1987). The effects of decision framing on choice of risky vs certain options. Organizational Behavior and Human Decision Processes.

https://doi.org/10.1016/0749-5978(87)90041-0

Fagley, N. S., & Miller, P. M. (1990). The Effect of Framing on Choice. Personality and

(41)

Faseur, T., & Geuens, M. (2012). On the effectiveness of ego- and other-focused ad-evoked emotions: The moderating impact of product type and personality. International Journal

of Advertising. https://doi.org/10.2501/IJA-31-3-529-546

Fishbein, M., & Ajzen, I. (2005). The Influence of Attitudes on Behavior. The Handbook of

Attitudes. https://doi.org/10.1007/BF02294218

Frisch, D. (1993). Reasons for Framing Effects. Organizational Behavior and Human

Decision Processes. https://doi.org/10.1006/obhd.1993.1017

Giving USA. (2019). Giving USA 2019: Americans gave $427.71 billion to charity in 2018 amid complex year for charitable giving. Giving USA. Retrieved from

https://givingusa.org/giving-usa-2019-americans-gave-427-71-billion-to-charity-in-2018-amid-complex-year-for-charitable-giving/

Glance, N. S., & Huberman, B. A. (1994). The Dynamics of Social Dilemmas. Scientific

American, 270(3), 76–81. https://doi.org/10.1038/scientificamerican0394-76

Graber, D. A. (1990). Seeing Is Remembering: How Visuals Contribute to Learning from Television News. Journal of Communication.

https://doi.org/10.1111/j.1460-2466.1990.tb02275.x

Hartmann, P., & Apaolaza-Ibáñez, V. (2009). Green advertising revisted: Conditioning virtual nature experiences. International Journal of Advertising, 28(4), 715–739. https://doi.org/10.2501/S0265048709200837

Hartmann, P., & Apaolaza-Ibáñez, V. (2012). Consumer attitude and purchase intention toward green energy brands: The roles of psychological benefits and environmental concern. Journal of Business Research. https://doi.org/10.1016/j.jbusres.2011.11.001 Hartmann, P., Apaolaza-Ibáñez, V., Levin, I. P., Schneider, S. L., Gaeth, G. J., Cheng, T.,

Woon, D. K., Lynes, J. K., Bu, O. B., Go, A. S., Sargeant, A., Kao, T. F., Du, Y. Z., Mo, Z., Liu, M. T., Liu, Y., Chang, C. C. T., Oosterhof, L., Heuvelman, A., … Kuiper, J. (2008). Improving the effectiveness of fundraising messages: The impact of charity goal attainment, message framing, and evidence on persuasion. International Journal of

Advertising, 18(2), 300–314. https://doi.org/10.1002/mar

Hartmann, P., Apaolaza, V., & Alija, P. (2013). Nature imagery in advertising: Attention restoration and memory effects. International Journal of Advertising, 32(2), 183–210. https://doi.org/10.2501/IJA-32-2-183-210

Hartmann, P., Apaolaza, V., & Eisend, M. (2016). Nature Imagery in Non-Green

Advertising: The Effects of Emotion, Autobiographical Memory, and Consumer’s Green Traits. Journal of Advertising, 45(4), 427–440.

https://doi.org/10.1080/00913367.2016.1190259

Holbrook, M. B., & Batra, R. (1987). Assessing the Role of Emotions as Mediators of Consumer Responses to Advertising. Journal of Consumer Research, 14(3), 404. https://doi.org/10.1086/209123

Houston, M. J., Childers, T. L., & Heckler, S. E. (1987). Picture-Word Consistency and the Elaborative Processing of Advertisements. Journal of Marketing Research.

https://doi.org/10.2307/3151383

Hull IV, R. B., & Stewart, W. (1992). Validity of photo-based scenic beauty judgments.

(42)

ISTAT. (2014). Nonprofit institution profile based on 2011 census results. Rome: Istituto nazionale di statistica. Retrieved from http://www.istat.it/en/files/2014/10/Nonprofit-

Institution-Profile-based-on-2011-Census- results_EN_definitivo.pdf?title=Nonprofit+institutions+profile+-+9+Oct+2014+-+Full+text.pdf

ISTAT. (2019). Structure and profiles of the non-profit sector. Rome: Istituto nazionale di statistica. Retrieved from: https://www.forumterzosettore.it/files/2019/10/Struttura-e-profili-settore-non-profit-2017_del11.10.2019.pdf

Kahneman, D., & Tversky, A. (1979). Prospect Theory: An Analysis of Decision under Risk Daniel Kahneman; Amos Tversky. Econometrica. https://doi.org/10.1111/j.1536-7150.2011.00774.x

Kashif, M., Sarifuddin, S., & Hassan, A. (2015). Charity donation: Intentions and behavior.

Marketing Intelligence and Planning, 33(1), 90–102.

https://doi.org/10.1108/MIP-07-2013-0110

Krishnamurthy, P., Carter, P., & Blair, E. (2001). Attribute Framing and Goal Framing Effects in Health Decisions. Organizational Behavior and Human Decision Processes,

85(2), 382–399. https://doi.org/10.1006/obhd.2001.2962

Krosnick J.A. (2018) Questionnaire Design. In: Vannette D., Krosnick J. (eds) The

Palgrav Handbook of Survey Research. Palgrave Macmillan, Cham.

Kühberger, A. (1995). The Framing of Decisions: A New Look at Old Problems.

Organizational Behavior and Human Decision Processes.

https://doi.org/10.1006/obhd.1995.1046

Latimer, A. E., Salovey, P., & Rothman, A. J. (2007). The effectiveness of gain-framed messages for encouraging disease prevention behavior: Is all hope lost? In Journal of

Health Communication. https://doi.org/10.1080/10810730701619695

Lazarus, R. S. (2000). How emotions influence performance in competitive sports. Sport

Psychologist, 14, 229–252.

Levin, I. P., Schneider, S. L., & Gaeth, G. J. (1998). All Frames Are Not Created Equal: A Typology and Critical Analysis of Framing Effects. Organizational Behavior and

Human Decision Processes, 76(2), 149–188. https://doi.org/10.1006/obhd.1998.2804

Lochbuehler, K., Mercincavage, M., Tang, K. Z., Dana Tomlin, C., Cappella, J. N., & Strasser, A. A. (2018). Effect of message congruency on attention and recall in pictorial health warning labels. Tobacco Control. https://doi.org/10.1136/tobaccocontrol-2016-053615

Malhotra, N. K. (2010). Marketing research: an applied orientation (6th ed.). Pearson. Martin, B. (1995). Is the glass half-empty or half-full? How the framing of advertising

messages affects consumer response. New Zealand Journal of Business, 17(1), 95–105. Martin, B., & Marshall, R. (1999). The interaction of message framing and felt involvement

in the context of cell phone commercials. European Journal of Marketing. https://doi.org/10.1108/03090569910249247

Matthes, J., & Wonneberger, A. (2014). The skeptical green consumer revisited: Testing the relationship between green consumerism and skepticism toward advertising. Journal of

How can environmental charity be promoted to the public?

How can environmental charity be promoted to

the public?

The role of message framing, nature imagery, and emotions in

environmental charity advertising

How can environmental charity be promoted to

the public?

The role of message framing, nature imagery, and emotions in

environmental charity advertising

Table of Contents

Mean scores of emotions per condition