Charitable negativity: working well or worthless?
How negative emotions in humanitarian communication
Influence people’s intentions to contribute to charity.
Master’s Thesis
Sophie Defaix
11864257
University of Amsterdam
Graduate school of Communication
Master’s programme Communication Science
Supervised by dr. Judith Möller
7322 words
June 28, 2019
Abstract
When is charitable advertising convincing enough to generate significant donations in time, money and material goods? This research looks into that question by studying the
effectiveness of negative emotions in humanitarian communication. The main purpose of this study is to present practical recommendations for not-for-profit organisations and to make a significant contribution to the literature on this topic. Earlier research on negative emotions in humanitarian communication is generally scarce and lacks conclusiveness with respect to its effectiveness. To fill this literature gap, this study aims to answer the question how negative textual emotions influence people’s intentions to donate or volunteer and how this
relationship is influenced by engagement and compassion fatigue. A survey experiment (N = 242) was conducted, where all participants were randomly assigned to one of the six
experiment groups and were shown a humanitarian advertisement. Afterwards, they were asked to answer several questions about their engagement, intentions to contribute and compassion fatigue. The results found in this study are twofold: on the one hand, no significant results were found for the influence of negative textual communication on contribution. On the other hand, people’s levels of engagement and compassion fatigue do play a significant role in their intention to contribute. Consequently, the importance of engagement and compassion fatigue when it comes to humanitarian communication should be considered in future research and in practice.
Keywords: Negative Emotions, Humanitarian Communication, Charity Advertising, Contribution, Sadness, Anger, Engagement, Compassion Fatigue.
Introduction
In June 2019, 43,232 not-for-profit (ANBI) organisations are registered with the Dutch tax authorities. Organisations are only classified as an ANBI organisation when they commit at least 90% of their resources to the common good (Belastingdienst, 2019). Each organisation may have its own definition of ‘common good’, but what they have in common is that they all rely on donations from the public. In 2015, het Centraal Bureau voor de Statistiek (CBS) published new data showing an increase in generosity towards these charitable causes; where households spent an average of 250 euros per year in the late 1990s, they now spend around 400 euros per year on charity. Even accounted for inflations, donations have
increased in the new millennium (CBS, 2015). However, in an article published by het NRC, these numbers are not as bright as one might think (Pama, 2017). According to their
independent research, people’s donations have declined significantly from a relative
viewpoint. In 1999, 0.96% of the gross domestic product (GDP) was spent on charity, but this percentage has dropped to 0.76% of the GDP in 2015 (Pama, 2017). Pama (2017)
concludes her news article with the statement that “Dutch citizens are less willing to sacrifice”.
Apart from the fact that a substantial number of ANBI organisations have to compete for a less generous audience, there are three other emerging trends in the not-for-profit sector: (i) organisations are reaching smaller audiences with their fundraising actions (Pama 2017); (ii) the culture of donating is changing from long term connections with the charity to incidental support; (iii) and overall trust in the good intentions of charities has declined (Vinkenvleugel, 2017). These trends together have created a serious issue for charities fighting for people’s support. This means that these organisations need to have a successful and efficient marketing strategy, to reach as many people as possible and convince them of their cause. The incorporation of negative emotions and shock appeals in advertising is the most widely used strategy among not-for-profit organisations (Chouliaraki, 2010). This thesis aims to investigate how effective negative emotions really are when it comes to triggering
people’s intentions to contribute. The purpose of this study is to add to the existing literature and to create practical recommendations for not-for-profit organisations fighting for people’s attention.
This study focuses on the charity appeals for organisations that are mostly concerned with humanitarian causes, such as poverty or hunger in developing countries. Very broadly, these appeals fall under the notion of humanitarian communication, which is defined as “the rhetorical practices of transnational actors that engage with universal ethical claims to
mobilize action on human suffering” (Chouliaraki, 2010, p. 108). When exploring the literature regarding negative emotions in humanitarian communication, several literature gaps are identified. First of all, there is relatively little research on negative emotions in not-for-profit communication (Burt & Strongman, 2005; Small & Verrochi, 2009), compared to contiguous areas of communication, such as corporate communication (Cotte & Ritchie, 2005; Dobele, Lindgreen, Beverland, Vanhamme & van Wijk, 2007; Eckler & Bolls, 2011) or political communication (Ansolabehere & Iyengar, 1995; Daignault, Soroka & Giasson, 2013; Landreville & LaMarre, 2011; Lau & Pomper, 2001).
Second of all, research is inconclusive with respect to the effectiveness of emotions in humanitarian advertising. It is uncertain whether negative emotions (Baberini, Coleman, Slovic & Västfjäll, 2015), positive emotions (Ford & Merchant, 2010) or a combination of the two (Bennett, 2015) is most effective in advertising. Moreover, the field lacks research on the effect of specific emotions, such as anger, sadness, guilt, happiness, nostalgia, etc.
Furthermore, there is little research that measures effectiveness in terms of donations and volunteering (Burt & Strongman, 2005) or which psychological processes come into play when people are deciding to donate or volunteer. In conclusion, there are some studies that investigate the use of emotions and there are some studies that look at what moves people to donate or volunteer, but relatively little research has combined the two factors. It is the combination of these two variables that is most relevant for such not-for-profit organisations.
A third and final gap that was identified in the literature is the effectiveness of textual, emotional cues. All research regarding emotions has made use of visual emotions to test effectiveness. To my knowledge, there are no studies that have investigated such effect of textual negative emotional cues, whilst, nearly every humanitarian advertisement includes text in any form. Therefore, studying how textual emotions influence people would be a relevant addition to literature and practice.
In order to address the issue of declining donations to charity and aiming to fill some of the gaps in existing literature on negative emotions in humanitarian communication, the following research question is proposed: “How does the use of negative emotional textual
cues in humanitarian advertisements influence people’s intentions to contribute and to what extent is this relationship influenced by engagement and compassion fatigue?” To answer
this research question, a quantitative experimental study is performed. This study first explores the existing literature on the topic more extensively to define the key concepts of this study. Thereafter, the methodology of this study is clarified in terms of sample, research design, stimulus material and measures. Then, the results from the data analysis are
presented and interpreted, followed by a concluding answer to the research question. This study ends with several suggestions for future research and a practical recommendation for not-for-profit organisations in the field of humanitarian communication.
Literature Framework
The main aim of the literature framework is to create an overview of the research that has been conducted concerning negative emotions in humanitarian communication, after which possible gaps in the literature can be explored. To start, this section addresses the
relationship between negative emotions and humanitarian contribution. Then, the influence of engagement and compassion fatigue is examined. Finally, the effect of different types and intensities of negative emotions is discussed.
The concept of emotion in corporate communication has been studied extensively over the past three decades. In 1998, Brown, Homer and Inman conducted a meta-analysis of the relationship between emotion-evoking advertising and audience responses and found strong effects of negative emotions on brand and product attitudes. Moreover, in the case of viral marketing, multiple studies show that campaigns only go viral when they connect emotionally, because emotions enhance people’s intentions to share content (Dobele et al., 2007; Eckler & Bolls, 2011). Similarly to academia, the use of emotions has been studied extensively by advertising professionals in the field, which showed similar results in
effectiveness (Cotte & Ritchie, 2005). Besides corporate communication, there is also a vast body of research that studied the use of emotions in political advertising, focusing on how mostly negative emotions influence voters’ opinions, behaviour and turnout. There remain significant debates in the literature regarding the potential effects of negative advertising on voter participation. Some argue that negative political advertising cause higher attention levels (Daignault et al., 2013) and a higher intention to participate (Landreville & LaMarre, 2011; Lau & Pomper, 2001). Others suggest that negativity can trigger a ‘boomerang effect’ that significantly reduces voter turnout (Ansolabehere & Iyengar, 1995).
Even though significantly less research has been conducted on the impact of emotion in a humanitarian context, there are some studies that generated interesting results. Burt and Strongman (2005) studied charity advertising with emotional cues and donation requests and found that negative emotions in such ads generate significantly larger donations in items, time and money. Many other scholars have built on this research since to explain this significant effect and have come up with several theoretical frameworks. One theory is concerned with the concept of emotional contagion, which is defined as “the tendency to automatically mimic and synchronize facial expressions, vocalizations, postures and
movements with those of another person” (Hatfield, Cacioppo & Rapson, 1992, p. 153-154). In their study of charity advertising, Small and Verrochi (2009) demonstrated that this theory is driven by emotional convergence, which occurs when the viewer’s feelings align with the
emotion conveyed in the advertisement. Simply put, they propose that faces with emotional expressions cause sympathy and giving, because it makes the audience feel the same emotion. Later research on the topic shows that emotional contagion is a very primitive, automatic response, as it diminishes the moment people shift from looking at facial expressions to reading about the subject’s situation (Baberini et al., 2015). Moreover, Baberini et al. (2015) showed that emotional contagion is more effective on feelings of sadness than happiness, as the contagion effect lasted longer and generated greater sympathetic reactions for participants who were exposed to sad facial expressions. Other studies, however, have found that non-negative emotions can be effective as well, such as Ford & Merchant (2010), who demonstrated that feelings of personal nostalgia in charity appeals evoke higher levels of emotions and donation intention. Other scholars find that the use of mixed emotions is most effective, such as Bennett (2015), who argues that when an ad evokes negative emotions first and positive emotions later, the positive emotion dissipates the feelings of discomfort, which encourages a more favourable response towards the
proposed solution.
Considering the studies mentioned above, a relationship between emotion and participation in a humanitarian context seems prevalent. However, they all employ different definitions of participation in humanitarian communication. To study how the use of negative emotions affects people’s behavioural intent, a clear and comprehensible definition is
required. Burt & Strongman (2005) argue that donating to charity is a unique social exchange, where “gifts are made with no tangible rewards” (p. 571). They propose that all charitable giving is a personal matter, involving emotions and value judgements. Even
though this is an interesting perspective on donation, it does not recognize the many different aspects that fall under the notion of charitable giving. However, such a distinction is made by Jones (2006), who studies giving and volunteering as two distinct forms in the context of civic engagement. The concepts of volunteering and donating together form the variable of
According to the literature, this intention to contribute should be influenced by exposure to a certain emotion. However, all studies mentioned earlier have looked at the effect of visual emotions. To my knowledge, there are no studies that have looked into the effects of textual emotions or words with an emotional connotation in a humanitarian context and therefore, this study adds to the existing literature. The following hypothesis is proposed: H1 = The use of negative, textual emotional appeals in humanitarian advertisements
positively influences people’s intention to contribute.
Engagement as a mediator
When looking at the relationship between emotions and humanitarian contribution, several psychological processes come into play. This study proposes that people’s engagement with the cause partially mediates the relationship between emotion and humanitariancontribution; the use of an emotion in humanitarian communication enhances people’s level of
engagement with the cause, which, in turn, enhances their intention to contribute.
Engagement, in this study, is seen as a feeling of connectedness to the cause. Fehrenbach & Rodogno (2015), for example, describe engagement as a specific effect that NGOs try to achieve among their audience, such as “stimulating emotion, like empathy or outrage” (p. 1125) and Joffe (2008) refers to engagement as “identification with suffering” (p. 86). In other words, engagement is not a very tangible concept with a clear definition, but it is concerned with feelings of sympathy, pity, compassion or outrage towards the situation, the ability to relate to the victim or cause and the urge to change the situation.
In the literature, engagement is often studied as an outcome of emotional advertising instead of a mediating process. Many studies have investigated how the use of emotions in advertising increases engagement and compassion (Hoijer, 2004; Lwin & Phau, 2014; Sciulli, Bhagat & Bebko, 2012; Small & Verrochi, 2009). For example, Sciulli et al. (2012) made use of eye tracking software to test levels of engagement with emotional print advertisement. They found that people unconsciously fixate on negative emotional cues in printed ads,
resulting in higher levels of attention and interest. Subsequently, there is also research supporting the proposed relationship between engagement and contribution. Warren, Sulaiman and Jaafar (2014) studied civic engagement on Facebook as a cue for internet activism. Similarly, Brandtzaeg and Haustveig (2014) looked at people’s motivation and reasons for ‘liking’ humanitarian causes on Facebook, which can also be seen as some form of contribution. They identified ‘emotional liking’ as one motivation to participate online. Even though no direct research has been found regarding engagement as a mediator in the relationship between emotion and contribution, there is some literature available to support the three processes separately. Therefore, besides the direct effect proposed by H1, the indirect effect of emotion on contribution is tested by the following hypothesis:
H2 = The relationship between the use of negative emotional appeals and intention to contribute is mediated by people’s level of engagement with the cause.
Compassion fatigue as a moderator
Another concept that should be included in the study of effectiveness of emotions in
humanitarian communication is ‘compassion fatigue’. Compassion fatigue is a phenomenon that interferes with the effectiveness of humanitarian campaigns. It is also called psychic numbing, referring to when people “relate to the suffering of one as a tragedy, but tune out the loss of thousands as statistic” (Maier, 2015, p. 702). In other words, compassion fatigue makes people not care about tragedies anymore, unless it is made personal and the tragedy has a face. Chouliaraki (2010) explains that humanitarian communication long relied on representing suffering to reach audiences. Organisations often made use of the ‘shock effect’, which are appeals that seek to turn grand emotions into action by blaming ‘the Western Perpetrators’ for human suffering (p. 111). The popular resistance to this shock effect is compassion fatigue and has two consequences: the ‘bystander effect’, which refers to the idea that negative emotion ultimately leaves people feeling powerless; and the
sender for making people feel miserable about the suffering instead of doing something about it (p. 112). These effects should strongly influence people’s intention to contribute, as it interferes with the effect that emotions have on engagement. Therefore, compassion fatigue is a moderating variable in this study, resulting in the following hypothesis:
H3 = The effect of negative emotional appeals on intention to contribute is weaker when people display high levels of compassion fatigue.
Intensity of emotion
The relationship between the frequency or intensity of the emotion and its effect on people is a complex one. If the emotion is too subtle, it might not be picked up by the audience and therefore have no effect. In their research, Burt & Strongman (2005) looked at whether the intensity of the emotion differs in persuading people to donate money, items and time. Only for monetary donations, they found significant results between intensity, meaning that the more intense the emotion, the higher people’s intentions to donate. However, this theory is not entirely supported by Cotte and Ritchie (2005), who propose that there is no linear relationship between intensity and effectiveness. They argue that weak appeals may be ineffective if they fail to convince people of the urgency, while strong appeals may fail if they deter people from the message. This theory was confirmed by Coulter and Pinto (1995). Therefore, this study proposes that the relationship between the intensity of the emotion and its effectiveness in getting people to contribute is curvilinear; as the intensity increases, the effectiveness increases as well, up to a certain point, after which effectiveness decreases as intensity keeps increasing. Unfortunately, to my knowledge, there are no studies that have measured intensity of emotion in terms of textual cues. Therefore, there are no benchmarks as to what constitutes to a low, medium or high frequency. In earlier research, intensity of emotion was determined by means of a pre-test to determine how intense an emotion is perceived (Cheshin, Rafaeli & Bos, 2011; Hancock, Landrigan & Silver, 2007). In order to prove this curvilinear relationship of textual cues, the following hypotheses will be tested:
H4a = The use of moderately intense appeals is most effective.
H4b = The use of highly intense appeals is more effective than low intense appeals.
Different types of emotions
Even though this study mainly tests the effect of emotions on humanitarian contribution as a whole, it is also examined how different types of emotions play a role in contribution;
specifically, whether the portrayal of anger and sadness have different effect on people’s intentions to contribute. As mentioned earlier, Jones (2006) makes a clear distinction
between giving and volunteering in his definition of contribution. He states that “volunteering reflects direct engagement in community life and an active community-based civil society. In complement, giving, which is more contingent on personal resources, is indicative of an interest in public concerns through supporting the actions of others” (2006, p. 250). When translating this into a humanitarian context, direct engagement with the cause by actively joining constitutes volunteering and indirect engagement by donating towards initiatives of others constitutes donating. It is important to make this distinction, as this study hypothesizes that the strength of the intention to donate or volunteer depends on the type of emotion to which the participant is exposed.
When looking into the study of specific emotions in humanitarian communication, most of the research is concerned with the effect of guilt appeals (Coulter & Pinto, 1995; Hibbert et al., 2007; Lwin & Phau, 2014). However, according to Ortony (1987) guilt is not a basic human emotion, as people do not always feel guilty when they are. In order to create a new perspective to the research on specific emotions in humanitarian communication, this study will specifically investigate the effects of the two most intrinsic negative emotions: anger and sadness. DeSteno, Petty, Rucker, Wegener and Braverman (2004) investigated the persuasiveness of these two emotions and found that the message becomes persuasive when the framed emotion matches the nature of the message.
In other words, sad messages are most persuasive when framed with sadness cues and angry messages when framed with anger cues. Even though this study is very relevant, no distinction was made in the behavioural outcomes. According to the literature, anger is known for its encouraging effect on people to do something and gives them an incentive to act (Weber, 2012). Hence, anger should increase volunteering intentions. This hypothesis is based on a well-known political theory, called the mobilization effect. From a cognitive point of view, the mobilization theory proposes that angry emotions trigger risk-seeking and problem-focused coping behaviour and therefore, people who experience anger are more likely to act on it (Valentino, Brader, Groenendyk, Gregorowicz & Hutchings, 2011). On the other hand, sadness is known to generate sympathy among its audience and sympathy leads to increased donating (Small & Verrochi, 2009; Burt & Strongman, 2005). To test whether these different emotions indeed cause different behaviour, the following hypotheses are proposed:
H5a = The use of sadness appeals has a stronger effect on people’s intention to donate than anger appeals.
H5b = The use of anger appeals has a stronger effect on people’s intention to volunteer than sadness appeals.
In conclusion, the hypothetical model tested in this study is displayed in figure 1.
Methodology
Sample
In order to test the hypotheses, a survey experiment was conducted. To this end, data were collected from a convenience sample (N = 242) drawn from the research population and all participants were asked to answer several questions regarding the effect of an emotional advertisement. Data were collected between April 19th and May 12th of 2019. In order to
ensure a heterogenous sample, participants over 18 years old were recruited from different nationalities, genders, ages and levels of education. To enhance variety in the sample, the survey was published in English and people from different demographic groups were asked to share the survey in their networks. The hyperlink to the survey was mostly distributed online through, among others, WhatsApp, Messenger, e-mail and various social media platforms. After cleaning the data, respondents with more than three missing values (e.g. most of them dropped out the study after seeing the advertisement) were deleted. There was a non-completion rate of 13.6% among all participants who entered the survey. Respondents with three or less missing values were kept in the study; missing values in the scale items were replaced with the mean score of the rest of the scale and missing values in the demographic questions were left missing. A sample of 242 participants was used for analysis. The distribution of participants over the six experiment groups is approximately equal: low anger (n = 41), medium anger (n = 34), high anger (n = 40), low sadness (n = 44), medium sadness (n = 37) and high sadness (n = 46). In terms of gender, nationality and age, the sample is predominantly female (67.4%), consists of primarily Dutch respondents (71%) and is relatively young of age (M = 31.17, SD = 13.32). Moreover, the sample is highly educated with 76.4% respondents indicated to have obtained a bachelor’s degree or higher.
Research design and procedure
To answer the hypotheses, a 2 (anger vs. sadness) x 3 (high, medium, low frequency)
several questions were included in a survey experiment. Participants were assigned randomly to one of the six experimental groups and were shown the corresponding
advertisement (see appendix A). All ads were manipulated in terms of type and intensity of emotion. In order to ensure that participants read the entire text in the ad, the ‘next’ button appeared only after 30 seconds. Afterwards, respondents were directed to a page with the open question ‘please, describe what you just saw?’ as a manipulation check. The
manipulation worked as intended, as 99.1% of the respondents answered this question correctly, except for two participants who did not answer at all. Then, all respondents were asked to answer the same questionnaire, including questions about their feelings of guilt, engagement and compassion fatigue, their intentions to contribute and some demographic questions. Afterwards, people were debriefed in order to tell them that the study was actually about the influence of emotion in advertisements and to give them an opportunity to leave questions or remarks about the study. Participation in the survey took approximately ten minutes in total. In order to enhance quality and maximize completion of the survey among respondents, a couple of measures were taken. Besides the ethical brief in the beginning, where people’s anonymity and privacy were guaranteed, a time indication and a progress bar were also included. Moreover, each new set of questions was briefly introduced and
explained.
Stimulus material and pre-test
As mentioned earlier, all participants were randomly assigned to one of the
experimental groups and saw one of the six advertisements congruent with this group (anger high, anger medium, anger low, sadness high, sadness medium, sadness low). The cause in the advertisement was kept very general in order to appeal to a wide audience and included several fixed and variable elements. The ad asked people to help improve poor sanitary conditions in Sub-Saharan Africa. All six ads included the same title, call to action and picture of a mother with her son in front of a primitive tent in a desolate area. All of these were kept
as neutral as possible, to make sure that it could match either with the sadness emotion or with the anger emotion. The variable elements in the study were a short phrase about the picture and a text of 95 to 120 words explaining the consequences of these poor conditions in Africa. These textual elements were manipulated for type and intensity of emotion, using emotional word lists from the Linguistic Inquiry and Word Count (LIWC). The LIWC word lists for negative emotions consists of words that trigger a certain type of negative emotion when people read it in a text (see Appendix A). This is based on the psychological meaning of a word and its arousal among people (Tausczik and Pennebaker, 2010). The high condition contained twenty emotional words, the medium condition contained eleven emotional words and the low condition only one. Except for the emotional cues, the text in both conditions were very similar (see appendix A).
In order to check whether the stimulus material worked as intended, a small pre-test was conducted (N = 8). Six people were all asked to rate one of the conditions and two people were asked to rate all conditions. First, they were asked to describe how the advertisement made them feel and then they were asked to rate each ad for both sadness and anger on a scale from 1 to 10. In terms of intensity, the ads were rated exactly as
intended (e.g. anger high got the highest anger rating, anger low the lowest, etc.), which was important, as there was no theoretical precedent found for number of textual cues to
represent intensity of emotion. The results of the pre-test (Appendix C) were good enough to proceed with these materials.
Measures
After cleaning the data, multiple reliability analyses were performed to test whether the different items could be combined into new variables. The newly created variables, the corresponding Cronbach’s Alpha scores and the items they consist of are displayed in table 1. The measurement of engagement was based on Kang’s (2014) thirteen level scale, consisting of three dimensions: affective commitment, positive affectivity and empowerment. According to Kang, people feel engaged when they score high on these three dimensions. However, since this study makes use of negative emotions, positive affectivity seemed to be redundant to measure. Therefore, only the factors accounting to empowerment and affective commitment were used and adapted to the specific case in the study (see Appendix B). People were asked to what extent they agreed with the statements on a 5-point Likert scale. Without positive affectivity, the scale consisted of eight items and was found to be highly reliable (α = .833), so all items were computed into a new variable with the mean for total engagement score. The scores for total engagement were normally distributed across the data.
In this study, intention to contribute consists of two factors: donating and volunteering. This is based on Jones’s (2006) study regarding the effect of civic participation. After seeing the advertisement, participants were asked to rate on a 5-point Likert scale how likely they were to participate in each of the ten activities. Six items fell under the notion of donating (α = .770), and four activities fell under the notion of volunteering (α = .829), (see Appendix B). For both scales, new variables were computed for scores that indicate total intention to donate and total intention to volunteer. The total contribution scale of all ten items was found to be highly reliable (α = .850), resulting in a newly computed variable to indicate mean contribution intent score. The scores for intention to contribute were normally distributed across the data.
Finally, people’s level of compassion fatigue was measured by asking them several questions regarding their overall feelings towards charity and charitable causes. The survey
consisted of ten statements that would indicate either the bystander effect or the boomerang effect (Chouliaraki, 2010, p. 112) and asked people to what extent they agreed on a 5-point Likert scale. They were also asked to rate whether they felt indifferent, distant, neutral, compassionate, sympathetic and indignant when seeing such ads (see Appendix B). Together, these sixteen items received a Cronbach’s Alpha score of .763. This score could be improved to an Alpha of .775 by deleting one item from the compassion scale (I feel indignant). Therefore, the newly computed variable of the mean for total compassion fatigue score consisted of fifteen items. The scores for compassion fatigue were normally distributed across the data.
Method of analysis
In order to test the regressive part of the hypothetical model proposed earlier, a conditional process analysis was performed. To this end, Andrew Hayes’ PROCESS modelling software for SPSS was used (Hayes, 2013). Model 10 fit the hypothetical model from this study exactly and included type of negative emotion as independent variable (X), contribution intention as dependent variable (Y), engagement as mediating variable (Mi), and compassion
fatigue and intensity of emotion as moderating variables (W and Z, respectively). This model tested the first three hypotheses from this study, which proposed that the use of negative emotions in humanitarian advertising has a direct effect on people’s intentions to contribute (H1) and an indirect effect, mediated through engagement (H2). And, according to the model, both relationships are moderated by people’s levels of compassion fatigue (H3). In order to test whether the mean scores for contribution intent varied significantly between the different types of emotions and the intensity levels of emotion (H4 and H5, respectively), an
Results
In order to explore how people’s intention to contribute differs between the six experiment conditions, a one-way ANOVA was conducted. Figure 2 displays the mean scores and standard deviations of intention to contribute per experiment group. However, the ANOVA showed that the differences between the groups are not significant, F(5, 236) = 0.50, p = .780.
The regression model with contribution intention as dependent variable and negative emotions, engagement and compassion fatigue as independent variables is significant, F(6, 235) = 31.19, p < .001. The regression model can therefore be used to predict contribution intent. The strength of the prediction is moderately strong, as 44 per cent of the variances in contribution intent can be explained by the independent variables (R2 = .44). Engagement, b
= 0.60, t = 10.42, p < .001, 95% CI [0.49, 0.71], and compassion fatigue, b = -0.35, t = -3.14,
p = .002, 95% CI [-0.57, -0.13], have a significant, strong association with contribution intent,
while type of negative emotion is not a significant predictor for the dependent variable, b = -0.59, t = -1.32, p = .188, 95% CI [-1.46, 0.29]. 2 ,5 2,6 1 2 ,7 6 2 ,5 5 2 ,5 7 2 ,6 3 0 0,5 1 1,5 2 2,5 3 3,5 4 4,5 5 A N G E R H I G H S A D N E S S H I G H A N G E R M E D I U M S A D N E S S M E D I U M A N G E R L O W S A D N E S S L O W IN T EN T ION T O C ON T R IBU T E EXPERIMENT GROUPS Mean
Exploring the conditional direct effects of negative emotions on contribution intent shows no significant results. Therefore, H1 is not supported by the data and in the
population, it cannot be proven that seeing negative emotional cues in advertising affects people’s intention to contribute. Moreover, when looking at the conditional indirect effects of negative emotions on contribution intent, there are no significant results either. However, due to the overall model statistics, indicating that engagement is a significant predictor for
contribution intent, b = 0.60, t = 10.42, p < .001, 95% CI [0.49, 0.71], H2 is partially
supported. In other words, the data show no significant indirect effect of negative emotions on contribution intent through engagement, but the effect of engagement on humanitarian contribution is statistically significant (p < .001). Finally, when looking at the effects of the moderating variable of compassion fatigue, it is a statistically significant predictor for both engagement, b = -0.51, t = -4.18, p < .001, 95% CI [-0.75, -0.27] and humanitarian
contribution, b = -0.35, t = -3.14, p = .002, 95% CI [-0.57, -0.13]. Therefore, H3 is supported and compassion fatigue is a significant moderator in the model. When looking at the
hypothetical model proposed in the literature section, the following paths were found significant (see figure 3):
Then, an independent samples t-test was conducted to compare volunteering and donating scores in the anger and the sadness conditions. H4 proposes that anger should have a stronger effect on volunteering and sadness should have a stronger effect on donating. Levene’s test for equality of variances was insignificant (F = 0.15, p = .696), which means that equal variances were assumed. According to the results, there is no significant
difference in donating intentions for anger (M = 2.86, SD = 0.78) and sadness (M = 2.83, SD = 0.83); t(240) = 0.28, p = .781. Neither is there a significant difference in volunteering intentions for anger (M = 2.22, SD = 0.88) and sadness (M = 2.24 , SD = 0.97); t(240) = -0.19, p = .850. Therefore, H4a and H4b are both rejected and for people’s intentions to donate or volunteer, the type of emotion displayed in the advertisement does not make a significant difference.
Furthermore, a one-way analysis of variances was conducted to compare the overall contribution intention scores among the three levels of intensity in emotion. H5 proposes a curvilinear relationship between intensity and intention to contribute. According to the results, there is no significant difference in contribution intention between the low (M = 2.60, SD = 0.77), medium (M = 2.65, SD = 0.75) and high (M = 2.55, SD = 0.77) conditions, F(2, 239) = 0.28, p = .756. Even though the mean contribution score was highest for the medium
condition, the ANOVA test indicated insignificant results and therefore, H5 is not supported. After examination of the results, it can be concluded that the variable of negative emotions is not related to any statistically significant effects or differences in the data. However, there seems to be a strong correlation between the variables of engagement, humanitarian contribution and compassion fatigue. A Pearson’s correlation test was conducted to confirm this assumption and results indicate a strong positive association between engagement and contribution intention, r(242) = 0.64, p < .001, as well as moderate negative associations between compassion fatigue and engagement, r(242) = -0.44, p < .001, and between compassion fatigue and contribution intention, r(242) = -0.43, p < .001. These results are also discussed in the next section.
Discussion
The purpose of this paper is to study how negative textual emotions in humanitarian communication influence people’s intention to donate to or volunteer for the cause. By manipulating charity advertisements for two types of negative emotions in various intensities, an effect on participants’ engagement and intention to contribute could be measured. After analysis of the results, it can be concluded that there is no simple answer to this question. The results found in this study are twofold: on the one hand, no significant results were found for the influence of negative textual communication on contribution. On the other hand, people’s levels of engagement and compassion fatigue do play a significant role in their intention to contribute. These results should be generalizable to the population for several reasons. To start, the experiment included a control group, who were shown an
advertisement without any emotional cues. Therefore, any significant effect of negative emotions could have been measured in the treatment groups. Moreover, the manipulations in the experiment were fully controlled, as the advertisements were kept exactly the same in terms of visuals and tone of voice. The only aspect that differs is the number and type of words used, which means that a difference in effect would most likely be caused by the manipulation. However, even though the sample was fairly large, it was not selected randomly, which is a requirement for generalizability. This concluding chapter of the paper elaborates on the implications of these results, reflects on possible limitations of the study and provides suggestions for future research and a practical recommendation.
Interpretation of the results
First of all, the effect of compassion fatigue and engagement on humanitarian contribution is discussed. Both hypothesis two and three are confirmed by the data in this study. When looking at the literature, it has become evident that there is an interplay between compassion fatigue and engagement. Maier (2015) defined compassion fatigue as “when people relate to suffering of one as tragedy, but tune out the loss of thousands as a statistic” (p. 702). This is
very simply explained by the fact that people feel engaged when that one person who suffers has a face, which is clearly reflected in the data. The significant effect of compassion fatigue and engagement on intention to contribute adds to the literature with respect to these
concepts, as relatively little is known about these phenomena in a humanitarian context, even though they are very present in a practical context. The data show that people who experience high levels of compassion fatigue feel less engaged and are less inclined to contribute. On the contrary, people who experience weaker feelings of compassion fatigue indicate a higher level of engagement and intention to contribute. Furthermore, the literature has demonstrated a clear connection between engagement and contribution (Brandtzaeg & Haustveig, 2014; Warren et al., 2014). This study confirms earlier results and shows that, especially in a humanitarian context, people want to contribute to the cause when they feel engaged.
Second of all, the lack of significant effects of negative textual emotions is important to discuss. Hypothesis one, four and five could not be confirmed by the data in this study. It can be concluded that the way negative emotions were manipulated in the advertisements did not have any significant interaction with people’s level of engagement or intention to contribute. There are two possible explanations for this outcome: either there was a problem with how the manipulation worked in this study or textual emotions do not trigger people’s feelings or intentions. The latter explanation is actually very interesting and should be discussed more specifically.
Overall, the literature shows that negative emotions have an impact on people, whether it is in a marketing context (Brown et al., 1998; Dobele et al., 2007; Eckler & Bolls, 2011), a political context (Daignault, Soroka & Giasson, 2013; Landreville & LaMarre, 2011; Lau & Pomper, 2001) or a humanitarian context (Burt & Strongman, 2005; Small & Verrochi, 2009). It can be concluded that when people connect emotionally, they feel engaged and want to contribute somehow. However, all this research has used visual emotions as the stimulus material, which is explained by the concept of emotional contagion. Baberini et al.
(2015) showed that this effect is so primitive and automatic that it diminishes the second people read information on the situation. Therefore, according to the emotional contagion theory, it should be impossible to convey emotion through text, because the reading diminishes the emotional effect. This matches the results from this study, as the textual emotional cues did not have any effect on people. The implications of this conclusion are crucial, as it means that people could only be impacted by emotion through visually
stimulating material. The stimulus material in this study did include an image, however, only the text was manipulated. Consequently, in this study, the picture could have had an effect on people’s engagement and intentions to contribute, but it did not differ significantly across the experiment groups, as they all saw the same picture.
Similarly, hypotheses four and five are both rejected; there is no significant difference in mean scores for contribution between the different experimental groups. This could be because the manipulation did not work correctly or because of the fact that textual emotions do not trigger people, as explained previously. Regarding the type of emotion, there is a third possible explanation, as it could be that most people do not make a clear distinction between the two emotions. It is very likely that the case portrayed in the advertisement makes people feel both anger and sadness. This also became evident from the pre-test (see Appendix C).
Limitations and future research
This study has several limitations that should be considered with the interpretation of the results. As mentioned multiple times before, it is unclear how well the manipulation of the emotions worked in the experiment. First of all, due to the nature of the cause displayed in the advertisement, people could be feeling both anger and sadness emotions. Second of all, there was no theoretical precedent for manipulating the intensity of the emotion, so it had to be based on the results of a pre-test with a relatively small sample (N = 8). Another limitation is concerned with the socially desirable nature of the questions. People were asked to what extent they felt engaged with the cause, how likely they were to contribute in order to
improve the cause and to what extent they felt compassion fatigue. All these topics are subject to social desirability. Even though participants’ anonymity was guaranteed, it is likely that it has still unconsciously influenced their answers. This could have influenced the results when people felt less engaged to the cause, but did not want to answer that they were not likely at all to contribute.
In the future, it would be interesting to dive deeper into the difference between textual and visual emotions, in order to explore how these differ in effect. It would be interesting to repeat this study with visual emotions instead of textual and see whether various types and intensities of emotion have a different effect in a humanitarian context. Moreover, it would add to the literature to expand on the effect of textual emotions. Most likely, this kind of research would require a different methodological approach in order to show actual effect. A more qualitative approach from a psychological or neuro-scientific point of view might be more suitable to discover whether emotions can be conveyed through text. Finally, this research showed that about 44% of the variances in contribution intention can be explained by people’s levels of engagement and compassion fatigue. It would be interesting to look at other variables that influence people’s intention to donate or volunteer, such as social
pressure, religious considerations or any other factors that were not considered in this study.
Practical recommendations
These results could also be very useful from a practical point of view for practitioners in the field of humanitarian communication. As mentioned in the introduction, there are several ongoing trends in the not-for-profit sector that are disadvantageous for charitable
organisations. Nowadays, there are so many charities competing for attention and donations in time, money or material items, that only the best humanitarian campaigns get optimal results. However, the question remains: when is such a campaign convincing enough to generate these contributions? To answer this question, there are two main recommendations for charities’ advertising strategies.
First of all, advertisements are most successful when they make people feel engaged with the cause; it should make it easy for people to relate to the victim or the cause, it should generate feelings of sympathy, pity, compassion or outrage towards the cause and it should give people the tools to contribute. One way to do this is using emotions, because according to the theory of emotional contagion, people copy an emotion displayed in an advertisement. In other words, if one wants to make their audience feel outraged, the emotional cues in the advertisement should express outrage or anger.
Second of all, advertisements are most successful if they avoid triggering people’s feelings of compassion fatigue. One person might display higher levels of compassion fatigue than someone else, but in the end, the majority of people experience the bystander effect or the boomerang effect at one point in time. The bystander effect is triggered when people get the idea that they have seen the same or similar images many times before. Therefore, it is important to refrain from stereotypical images, text and calls to action and to aim new and innovative advertisements with an element of surprise. The boomerang effect is triggered when people get the feeling that the charity is trying to manipulate them into feeling guilty and donating to the cause. Therefore, it is crucial that the advertisement is sincere and does not blame the audience for the situation. In conclusion, an advertisement is most likely to be successful when it triggers engagement and sympathy through the use of emotions, gives people concrete tools to contribute, is innovative and original and does not put the blame on them or the Western world.
References
Ansolabehere, S., Iyengar, S., Simon, A. & Valentino, N. (1994). Does attack advertising demobilize the electorate? The American Political Science Review, 88(4), 829-838. Baberini, M., Coleman, C. L., Slovic, P., & Västfjäll, D. (2015). Examining the effects of
photographic attributes on sympathy, emotions, and donation behavior. Visual
Belastingdienst. (2019). Wat is een ANBI? Retrieved from https://www.belastingdienst.nl/ wps/wcm/connect/bldcontentnl/belastingdienst/zakelijk/bijzondere_regelingen/goede_ doelen/algemeen_nut_beogende_instellingen/wat_is_een_anbi
Bennett, R. (2015). Individual characteristics and the arousal of mixed emotions: Consequences for the effectiveness of charity fundraising
advertisements. International Journal of Nonprofit and Voluntary Sector
Marketing, 20(2), 188-209.
Brandtzaeg, P.B. & Haugstveit, I.M. (2014). Facebook likes: a study of liking practices for humanitarian causes. International Journal Web Based Communities, 10(3), 258-280. Brown, S.P., Homer, P.M. & Inman, J.J. (1998). A meta-analysis of relationships between ad-
evoked feelings and advertising responses. Journal of Marketing Research, 35(1), 114-126.
Burt, C. D., & Strongman, K. (2005). Use of images in charity advertising: Improving donations and compliance rates. International Journal of Organisational
Behaviour, 8(8), 571-580.
Cheshin, A., Rafaeli, A., & Bos, N. (2011). Anger and happiness in virtual teams: Emotional influences of text and behavior on others’ affect in the absence of non-verbal cues.
Organizational behavior and human decision processes, 116(1), 2-16
Cotte, J. & Ritchie, R. (2005). Advertisers’ Theories of consumers: Why use negative emotions to sell? Advances in Consumer Research, 32, 24-31.
Daignault, P., Soroka, S. & Giasson, T. (2013). The perception of political advertising during an election campaign: A measure of cognitive and emotional effects. Canadian
Journal of Communication, 38(2), 167-186.
DeSteno, D., Petty, R. E., Rucker, D. D., Wegener, D. T., & Braverman, J. (2004). Discrete emotions and persuasion: the role of emotion-induced expectancies. Journal of
Dobele, A., Lindgreen, A., Beverland, M., Vanhamme, J. & Van Wijk, R. (2007). Why pass on viral messages? Because they connect emotionally. Business Horizons, 50, 291- 304.
Eckler, P. & Bolls, P. (2011). Spreading the virus. Journal of Interactive Advertising, 11(2), 1- 11. DOI: 10.1080/15252019.2011.10722180
Fehrenbach, H. & Rodogno, D. (2015). “A horrific photo of a drowned Syrian child”: Humanitarian photography and NGO media strategies in historical perspective.
International Review of the Red Cross, 97(900), 1121-1155.
Ford, J. B., & Merchant, A. (2010). Nostalgia drives donations: The power of charitable appeals based on emotions and intentions. Journal of Advertising Research, 50(4), 450-459
Hancock, J. T., Landrigan, C., & Silver, C. (2007). Expressing emotion in text-based communication. In Proceedings of the SIGCHI conference on Human factors in
computing systems (929-932). ACM.
Hatfield, E., Cacioppo, J. T., & Rapson, R. L. (1992). Primitive emotional contagion. Review of personality and social psychology, 14, 151-177.
Hayes, A. (2013). Introduction to mediation, moderation and conditional process analysis: A
Regression based approach (2nd ed.). New York, NY: The Guilford Press.
Hibbert, S., Smith, A., Davies, A., & Ireland, F. (2007). Guilt appeals: Persuasion knowledge and charitable giving. Psychology & Marketing, 24(8), 723-742.
Höijer, B. (2004). The discourse of global compassion: The audience and media reporting of human suffering. Media, Culture & Society, 26(4), 513-531.
Joffe, H. (2008). The power of visual material: Persuasion, emotion and identification.
Diogenes, 217, 84-93.
Jones, K. S. (2006). Giving and volunteering as distinct forms of civic engagement: The role of community integration and personal resources in formal helping. Nonprofit and
Landreville, K.D. & LaMarre, H.L. (2011). Working through political entertainment: How negative emotion and narrative engagement encourage political discussion intent in young Americans. Communication Quarterly, 59(2), 200-220.
Lau, R. R. & Pomper, G. M. (2001). Effects of negative campaigning on turnout in U.S. Senate elections, 1988-1998. The Journal of Politics, 63(3), 804-819.
Lwin, M., & Phau, I. (2014). An exploratory study of existential guilt appeals in charitable advertisements. Journal of Marketing Management, 30(13-14), 1467-1485.
Marmor-Lavie, G. & Weimann, G. (2005). Measuring emotional appeals in Israeli election campaigns. International Journal of Public Opinion, 18(3), 318-339.
Merchant, A., Ford, J. B., & Sargeant, A. (2010). Charitable organizations' storytelling influence on donors' emotions and intentions. Journal of Business Research, 63(7), 754-762.
Nederland steeds ruimhartiger voor goede doel. (2015, December 18). Centraal Bureau voor
de Statistiek. Retrieved from https://www.cbs.nl/nl-nl/nieuws/2015/51/nederland-
steeds-ruimhartiger-voor-goede-doel.
Ortony, A. (1987). Is guilt an emotion? Cognition and Emotion, 1(3), 283-298.
Pama, G. (2017, April 20). Nederlanders geven steeds minder uit aan goede doelen. NRC. Retrieved from https://www.nrc.nl/nieuws/2017/04/20/nederlands-geven-minder-uit- aan-goede-doelen-8274772-a1555259
Sciulli, L., Bhagat, P. S., & Bebko, C. P. (2012). Eye tracking analysis: Engagement levels and donor tendencies using print advertisements with emotional appeals. Innovative
Marketing, 8(2), 91-98.
Small, D. A., & Verrochi, N. M. (2009). The face of need: Facial emotion expression on charity advertisements. Journal of Marketing Research, 46(6), 777-787
Valentino, N. A., Brader, T., Groenendyk, E. W., Gregorowicz, K., & Hutchings, V. L. (2011). Election night’s alright for fighting: The role of emotions in political participation. The
Vinkenvleugel, M. (2017, May 3). Goede doelen, pak je communicatie anders aan. Non-
Profitmarketing. Retrieved from https://www.marketingfacts.nl/berichten/goede-doelen -pak-je-communicatie-anders-aan.
Warren, A.M., Sulaiman, A. & Jaafar, N.I. (2014). Facebook: The enabler of online civic engagement for activists. Computers in Human Behavior, 32, 284-289.
Weber, C. (2013). Emotions, campaigns, and political participation. Political Research
Appendix A – stimulus material
Anger high (n = 40)
Anger medium (n = 34)
Anger low (n = 41)
Sadness high (n = 46)
Sadness medium (n = 37)
Sadness low (n = 44)
Advertising texts with emotional cues
➢ Sadness – high intensity – 20 cues
Mauro and his mother are suffering from the devastating conditions in Africa. They feel helpless and miserable every day. Help us end this tragic agony!
Why are millions of children suffering?
In Sub-Saharan Africa, children are suffering from conditions and diseases that could be prevented by the proper measures. Undernutrition, unsafe water, poor hygiene practices and inadequate sanitation are tragedies that significantly hurt children.
Children like Mauro find themselves in hopeless situations every day. The truth is that they are deprived from a safe and healthy environment and are longing for a better world. It is devastating and tragic that little children end up in such sorrowful situations. This tragedy needs to stop now!
It is time that we start to help the children who are lost and helpless in this misery. Check our website to find out what you can do to help.
➢ Sadness – medium intensity – 11 cues
Mauro and his mother are suffering from the problematic conditions in Africa. They feel helpless every day. Help us end this agony!
Why are millions of children suffering?
In Sub-Saharan Africa, children are suffering from conditions and diseases that could be prevented by the proper measures. Undernutrition, unsafe water, poor hygiene practices and inadequate sanitation are problems that significantly hurt children.
Children like Mauro find themselves in hopeless situations every day. The truth is that they deserve a safe and healthy environment and are longing for a better world. It is tragic that little children end up in such sorrowful situations. This needs to stop now!
It is time that we start to help the children in this misery. Check our website to find out what you can do to help.
➢ Sadness – low intensity – 1 cue
Mauro and his mother encounter problematic conditions in Africa every day. Help us help Mauro!
Why are millions of children suffering?
In Sub-Saharan Africa, children endure conditions and diseases that could be prevented by the proper measures. Undernutrition, unsafe water, poor hygiene practices and inadequate sanitation are issues that significantly contribute to child mortality.
Children like Mauro find themselves in unfortunate situations every day. The truth is that they deserve a safe, healthy environment and a better life. Little children should not be in such situations, so we need to stop it now!
It is time that we start to help the children! Check our website to find out what you can do to help.
➢ Anger – high intensity – 20 cues
Mauro and his mother are victims of the dreadful conditions in Africa. They face threats and dangers every day. Help us fight this cruelty!
Why are millions of children dying?
In Sub-Saharan Africa, children are dying from conditions and diseases that could be
prevented by the proper measures. Undernutrition, unsafe water, poor hygiene practices and inadequate sanitation are dangers that significantly contribute to child mortality.
Children like Mauro find themselves in terrifying situations every day. The truth is that they are victims, held hostage in a life that they don’t deserve. It is outrageous and frustrating that little children have to fight for their lives against the evil in this world. This cruelty needs to stop now!
It is time that we start to defend the children who are victims of these threatening conditions. Check our website to find out what you can do to help.
➢ Anger – medium intensity – 11 cues
Mauro and his mother are victims of the problematic conditions in Africa. They face dangers every day. Help us fight for Mauro!
Why are millions of children dying?
In Sub-Saharan Africa, children are dying from conditions and diseases that could be
prevented by the proper measures. Undernutrition, unsafe water, poor hygiene practices and inadequate sanitation are dangers that significantly contribute to child mortality.
Children like Mauro find themselves in unhealthy situations every day. The truth is that they are victims and they don’t deserve this life. It is frustrating that little children have to fight for their lives. This needs to stop now!
It is time that we start to defend the children who are victims of these conditions. Check our website to find out what you can do to help.
➢ Anger – low intensity – 1 cue
Mauro and his mother encounter problematic conditions in Africa every day! Help us help Mauro!
Why are millions of children in Africa dying?
In Sub-Saharan Africa, children endure conditions and diseases that could be prevented by the proper measures. Undernutrition, unsafe water, poor hygiene practices and inadequate sanitation are issues that significantly contribute to child mortality.
Children like Mauro find themselves in unhealthy situations every day. The truth is that they are stuck in a life they don’t deserve. Little children struggle for their lives every day and this needs to stop now!
It is time that we start to help the children in these conditions. Check our website to find out what you can do to help.
Appendix B – survey experiment
After seeing the humanitarian advertisement, participants were taken to the survey experiment. This appendix includes all questions asked to the participants.
Manipulation check
Q1: Please, describe very briefly what you just saw?
Guilt
Participants were asked to indicate to what extent they agreed with the following statements on a five-point Likert scale:
Q2: I feel guilty about the situation.
Q3: I feel bad for the people who are in this situation
Engagement
Participants were asked to indicate to what extent they agreed with the following statements on a five-point Likert scale:
Q4: I feel emotionally attached to the cause.
Q5: I feel like a part of it when I participate (e.g. donate or volunteer). Q6: I feel a sense of belonging when I participate (e.g. donate or volunteer). Q7: I feel like I can actually make a difference.
Q8: I am determined to help develop solutions to the cause. Q9: I feel like I have control over the situation.
Q10: I am confident about my ability to improve the situation.
Contribution
Participants were asked to indicate how likely they were to do a number of things on a five-point Likert scale. Please indicate how likely you are to…
Q12: … donate a certain amount of money once? Q13: … donate money during an organized collection? Q14: … send a text to raise money?
Q15: … become a regular contributor to the organization or charity? Q16: … donate material items, such as clothes or toys?
Q17: … become a member of the organization for a small, periodical fee? Q18: … organize a collection for charity?
Q19: … support an organization or initiative by volunteering sporadically? Q20: … volunteer on a regular basis?
Q21: … start your own initiative to help the cause?
Compassion fatigue
Participants were asked to indicate to what extent they agreed with the following statements on a five-point Likert scale:
Q22: There is not much I can do to help the cause.
Q23: Whatever I do to help the situation, it will not make a big difference. Q24: It is not my job to try to change something about the situation. Q25: Thinking about the cause makes me feel powerless.
Q26: I don’t feel responsible when asked for help. Q27: I see similar ads to the one in this study every day.
Q28: Charities use these images to manipulate me into donating.
Q29: It would be better if organisations make use of positive images to illustrate the cause. Q30: The images are used to make me feel guilty.
Q32: Whenever I see such images, I feel: a. Indifferent b. Distant c. Neutral d. Compassionate e. Sympathetic f. Indignant (=verontwaardigd). Demographic questions What is your gender? What is your age? What is your nationality?
What is your current state of employment?
Appendix C – Pre-test Results
Each advertisement was rated for Anger and Sadness intensity by one participant:
Anger rate Sadness rate Highest emotion Difference
Anger high 10 8 Anger 2
Anger medium 4 8 Sadness 4
Anger low 3 2 Anger 1
Sadness high 1 10 Sadness 9
Sadness medium
4 8 Sadness 4
Sadness low 5 3 Anger 2
Furthermore, two participants rated all ads for anger and sadness:
A-L A-M A-H S-L S-M S-H
A S A S A S A S A S A S
Part 1. 2 8 2 8 2 8 2 8 2 8 7 9
Part 2. 4 3 5 5 7 7 6 6 2 2 2 3
All participants were also asked to extensively describe their feelings when looking at the advertisement. The most relevant results are stated below:
Anger high
- When I read the text, I wonder what we can do to help. - Strongest feeling comes from this ad.
- Feelings of guilt, not doing enough to help - Somewhat compassionate feelings
Anger medium
- Very subtle difference in meaning with anger high - Sad to see the difficult circumstances
- Feelings of sadness for suffering
- Also mentions high levels of compassion fatigue
Anger low
- Very subtle difference in meaning with anger medium - Little sympathy
- Feeling sad and mad, but what can we do about it?
Sadness high
- Feelings of guilt, feeling sorrow, unfairness
- Maybe look at the website to see what can be done - Feeling of sadness, looking to be able to do something
Sadness medium
- Feeling that something has to be done is less strong than sadness high - Still feel sorry, less urge to do something about it
- Sorrow for Mauro, average intensity
Sadness low
- No intention at all to look at the website
