Investigating the Effectiveness of Message Framing and Regulatory Fit in Increasing Positive Animal Attitude and
Reducing Animal Products Consumption. A study for Animal Equality
Author: Carolina Bertolaso Tutor: Suzanne de Bakker
Abstract
The purpose of this study was to help Animal Equality to improve its Facebook communication strategy by testing the individual and combined effectiveness of Facebook posts involving animal cruelty, individualization of farmed animals, and promotion or prevention focused messages, in increasing positive attitude towards animals and decreasing actual consumption of animal products.
A two-phases experiment was conducted among 511 non-vegan females between 15 and 35 years old, distributed into four experimental conditions and a control group. Participants were exposed to a total of six Facebook posts and their attitude towards animals and animal products consumption were measured both one week before and one week after exposition.
The results show that messages involving moral shocks produce behavioral change, while messages involving individualization are not effective. Promotion focused messaged are more effective than prevention focused messages, being the combination of moral shocks and promotion focused messages the most effective strategy.
Dedication
To all the voiceless animals who are now suffering. May they be free. And to all those humans that care for them. May they be their voice.
Acknowledgements
First of all, I express my warm thanks to Sharon Nuñez Gough for believing in me and giving me this opportunity, and to all the Animal Equality team for their priceless help.
I would also like to thank my thesis supervisor, Suzanne the Bakker, for her criticism, her honesty and her constant encouragement.
A final, special thank to my family, that supported me during my entire life and especially during this very long year, and to my beloved boyfriend Andrea, which has always, always been there.
Thank you,
Facebook Communication for Animal Rights: Investigating the Effectiveness of Message Framing and Regulatory Fit in Increasing Positive Animal Attitude and
Reducing Animal Products Consumption.
Animal Equality is an animal rights organization whose main goals are to raise awareness about animal issues through campaigns, educational initiatives,
investigations, outreach, and legal advocacy, and change people’s attitudes and behaviors towards non human animals (http://www.animalequality.net/about-us). As every animal rights organization, Animal Equality aims at reaching benefits for a segment of the society that is not even considered part of it – the animals- which makes it deal with the difficult challenge of having to defy the dominant cultural belief that is based on an anthropocentric view of the world.
Anthropocentrism can be defined as the belief that human beings are the most significant species on the planet and that their moral status and value is higher than the one of other species (Merriam-Webster Dictionary). This assessment leads to a society that looks at reality through an exclusively human perspective, thus relegating other species in a second, inferior level. In a culture in which animal exploitation is socially accepted and normalized, Animal Equality represents an antagonistic minority that threatens the mainstream belief in a way that not only endangers an entire economic system based on the absolute dependence on non human animals, but also dares questioning individuals’ personal habits.
The difficulty of maintaining such a posture in a society that looks at animals as mere resources can be partially explained by some existing theories and by a quite wide range of research; in fact some studies have already been conducted in order to better understand the psychological mechanisms underlying the human refusal to
acknowledge that animals have rights and the consequent complications that this mechanisms have on animal rights organizations.
For example the Minority Influence Theory (Moscovici, 1976) suggests that it can be extremely difficult for minority groups and their messages to be accepted, and animal rights organizations like Animal Equality constitute a sector that not only cannot count on support from institutions but also has to deal with the strong resistance that individuals tend to have toward their messages. In fact, one of Animal Equality’s principal goals is the one of persuading people to quit animal products and adopt a vegan diet; such a goal is not an easy goal to achieve, since it directly concerns one of the most relevant aspects of people’s life: the eating behavior.
People’s denial to recognize animals’ rights in order to perpetuate their eating behaviors has been demonstrated in previous research. For example, a study by Bastian, Loughnan, Haslam, & Radke (2011) found that people ascribe less mental capacities to animals that are meant for human consumption than to those that are considered
companion animals; furthermore, the authors discovered that when meat eaters are made aware of the link between animal products and animal suffering, they become motivated to deny mental capacities to farmed animals. The most interesting discovery of the research was that when participants were expecting to consume meat in the short time, their mind denial increased. This mind denial was found to reduce the negative affect of cognitive dissonance.
According to the Cognitive Dissonance Theory (Festinger, 1957), human beings have an inner drive to maintain their attitudes and beliefs aligned in order to avoid disharmony; when the cognitions that people hold about themselves and the world are not congruent, a discrepancy is evoked. This results in a state of tension known as cognitive dissonance. Because of the unpleasant experience that this status of tension implies, people are motivated to reduce it by means of cognitive strategies aimed to
rebuild consonance. This means that they focus on supportive beliefs and outweigh the dissonant beliefs or behaviors; reduce the importance of the conflicting beliefs; or change the conflicting beliefs so that it is consistent with other beliefs or behaviors.
Literature from the philosophy of animal rights (Bruers, 2014; Regan, 2005) has put a lot of emphasis in recognizing the resistance mechanisms that animal rights organizations’ messages activate among people. As examples of ways in which people react to these messages by focusing on supportive beliefs, Regan (2005) lists “animals are not humans”, “animals do not understand what rights are”, “animals do not respect our rights”, “animals do not respect other animals’ rights”, “animals are not conscious beings” and “animals do not own a soul”.
According to Regan, an example of how people try to avoid dissonance by reducing the importance of the conflictive belief is “the idea of animal rights is absurd”, while “vegetables suffer as well” is the most frequent example of the strategy focused on changing the conflicting belief in order to perpetuate the conflicting behavior.
A study by Rothgerber (2014) identified a number of strategies that omnivores adopt to reduce this dissonance including avoidance, dissociation, perceived behavioral change, denial of animal pain, denial of animal mind, pro-meat justifications, reduction of perceived choice, and actual behavioral change.
An interesting study by Bilewicz, Imhoff, & Drogosz (2010) focused on the concept of dehumanization and found that omnivores deny farmed animals to have human characteristics, like uniqueness and ability to feel secondary emotions, as a strategy for moral disengagement. According to the Theory of Moral Disengagement (Bandura, 1996; 1999) people like to have an image of themselves as ethical and moral, so that in situations in which they behave unethically or unjustly, they adopt the
behavior. In this case, denying farmed animals characteristics that would produce empathetic emotions, works as a mechanism that enables taking moral distance from them, thus allowing avoidance of feelings of guilt.
The Theory of Psychological Reactance (Brahm, 1996) suggests that persuasive messages constitute a potential threat to personal freedom and therefore are strongly liable to be rejected. It is not difficult to realize that a message carrying the idea that animals are not resources made for human consumption represents a serious threat in that its purpose is to subvert the anthropocentric status quo, which constitutes a menace for people’s perceived freedom to act as if animals had not rights and therefore use them for human purposes.
Given all the above mentioned information, it is clear that human beings are hardly persuadable and that modifying their behaviors becomes especially difficult when the change implies ingrained habits such as eating habits. Animal rights
organizations are therefore in an urgent need for research in message strategies in order to identify what could allow them to better overcome the barriers that their advocacy triggers in people’s attitudes and behaviors. As a result of this, animal rights
organizations like Animal Equality could be more effective in their goal of saving as many animals as possible.
The animal rights movement is currently witnessing the professionalization of animal rights organizations, with the rising of institutions as The Humane Lab and the Humane Research Council, research projects that perform direct testing in order to improve vegan advocacy efforts. Also, the Animal Rights Leadership Council, an organization that provides highly professional education and training to young excellent activists, is seeking to lead the movement into the next decades.
Animal Equality has recently been indicated by Animal Charity Evaluators as one of the three most effective international animal rights organizations
(http://www.animalcharityevaluators.org/recommendations/top-charities/). Present in the United Kingdom, Germany, Italy, Spain, Venezuela, India and the United States, Animal Equality performed several successful campaigns and counts with more than 2.200.000 followers on their Facebook pages (January 2015), positioning itself as the most popular Europe - based animal rights organization
(http://www.animalcharityevaluators.org/research/organizations/animal-equality-review/). Following the recently diffused tendency, Animal Equality is now willing to continue improving its effectiveness and enhance its communication strategy from relying on assumptions to being based on a solid, rigorous scientific method.
Therefore, the goal of this thesis will be to extend the knowledge regarding effective communication for animal rights, in order to provide Animal Equality and other animal rights organizations with practical advices about how to persuade their publics to adopt a vegetarian or vegan diet.
Message framing
When looking for message effectiveness, one of the most relevant issues to address is the one of choosing the right frame (Cornelissen, 2011). Message framing can be defined as the way in which a message is presented (McQuail, 2010). Different modes to present a message can lead to different responses, depending on the language used to describe an issue or the visual context which is given to the message. Framing regards the selection and highlighting of specific dimensions of a problem and is used by organizations to provide their publics with interpretative shortcuts and thus
increasing the effectiveness of the persuasive message (McQuail, 2010). One of the most well studied types of message framing is the one of the regulatory Focus. Regulatory focus can be defined as a way in which someone moves
towards pleasure but tries to avoid pain. (Higgins, 1997). According to Regulatory Focus Theory (Higgins, 1997), two self-regulatory orientations exist: prevention and promotion. The prevention focus highlights security, responsibility, and safety necessities, frames goals as obligations, and carries a strategic concern with approaching non-losses (absence of negatives) and avoiding losses (presence of
negatives). The promotion focus gives emphasis to expectations, accomplishments, and development needs; it depicts goals as ideals, and carries a strategic concern with approaching gains (presence of positives) and avoiding non-gains (absence of positives). To better understand the Regulatory Focus Theory, it is necessary to mention the Regulatory Fit Theory (Higgins, Cesario, Hagiwara, Spiegel, & Pittman, 2010). According to this theory, when there is a match between personal orientation towards a goal and the way that is used for moving towards that goal, then a state of regulatory fit is generated that produces two positive outcomes on the individual: a feeling of rightness regarding the goal accomplishment, and increased engagement with the goal pursuit (Cesario & Higgins, 2008; Higgins, 2005).
The first to demonstrate that regulatory fit theory could increase the
effectiveness of a persuasive appeal were Cesario, Grant, and Higgins (2004) and Lee and Aaker (2004). In their studies, they framed the arguments of a persuasive message in a way that fit or did not fit the chronical orientation of the recipient of a message, and demonstrated that messages that fit had a more persuasive power.
It was initially assumed that in order to produce this fit it was necessary to measure individuals’ chronical regulatory focus and then adapt the message to the result; on the other hand, recent research has dismantled this idea by showing that it was possible to induce regulatory fit on an individual without previously knowing his/her regulatory focus. The way to do this is by manipulating the orientation by directly framing the topic of the message in terms of prevention or promotion concerns.
Concretely, the authors primed either a promotion or a prevention focus by framing the tagline of an ad in terms of presence of positive outcomes and absence of negative outcomes, and then delivered the following message in a way that did fit.
The functioning of Regulatory Focus has been widely investigated as it represents one of the most used framing theory in persuasive communication. A wide range of literature that is focused on the operating of this construct, both for commercial
purposes (Kang & Kim, 2014; Seunghyun, 2014; Smith, Gleim, Robinson, Kettinger, & Park, 2014;) and for health objectives (Gomez, Borges, & Pechmann, 2013; Johnsen, Vambheim, Wynn, & Wangberg, 2014; Pula, Parks, & Ross, 2014) already exists. However, its effectiveness in promoting altruistic behaviors has not been tested yet.
Animal rights strategies. One of the most salient strategies used by animal rights organizations in their communications is the one of rising emotional appeals to empathy and compassion (Bertolaso, 2014). There are two ways in which Animal Equality triggers these emotions: the first one is the use of moral shocks, stimuli that raises a sense of outrage which in turn leads individuals to react in response to them (Jasper & Poulsen, 1995). According to the authors, all social movements aim at
producing moral claims and shocks using verbal or visual images. In this sense, Animal Equality does not differ from other social movements in that it strongly relies on moral shocks; in fact, one of the most salient aspects of its outreach is the presentation of undercover investigations, graphic imageries, videos and narrations showing the brutal treatment of animals in factory farms and slaughterhouses.
The second strategy that animal rights organizations use to raise compassion and empathy is the use of narratives that dismantle the deindividuation of animals, namely the process through which people think about farmed animals only in terms of their group identity, and as being all the same and having all the same characteristics (Joy, 2010). Most animal rights organizations use this strategy and Animal Equality is
not an exception. In their Facebook page it is possible to observe several examples of this, like posts in which they describe farm animals in terms of their ability to feel and have emotions. For convenience, I will refer to this concept as “individualization”.
Both moral shocks and individualization of farm animals are widely used by Animal Equality and by many other animal rights organizations in their Facebook communications, and it is my opinion that discovering which one actually works better would enable the development of a more effective social media communication
strategy.
The importance of social media
During the last decade, the rise of social media has provided animal rights organizations with an extremely powerful tool for spreading their messages to a wide range of publics; moreover, new strategies are in process of being developed in order to become as effective as possible. Nowadays, Facebook is acquiring more and more importance in that it provides a window through which the campaigns of organizations like Animal Equality can go from local to broad and therefore allows a wider reaching and more visibility. Furthermore, Facebook provides Animal Equality with its own platform, which unbinds it from the constraints derived from gatekeeping processes, allows transparency and gives room for independent publications. Considering that a big part of Animal Equality’s advocacy work takes place on their Facebook page, it is reasonable to conclude that this social medium is a remarkable asset for the
organization and deserves to be investigated.
For all these reasons, the aim of this research will be to gain deeper insight into the specific message frames via social media that are effective in changing people’s attitude and behavior toward animals. The main research question is whether the individual effects or combined effects of online message frames involving moral shocks, individualization, promotion, and prevention are effective in favorably
changing the attitude towards animals and the actual consumption of animal products among young female Facebook users. The answer to this question should offer animal rights organizations knowledge on how to be more effective in achieving their
advocacy goals of persuading followers to quit animal products. With this purpose, two different lines of research will be analyzed in the literature review: specific animal rights advocacy strategies and findings from the animal rights literature, and message strategies and findings from the field of persuasive communication.
RQ: What are the individual and combined effects of online message frames involving moral shocks, individualization, promotion, and prevention on the attitude towards animals and the actual consumption of animal products?
Theoretical Framework
As mentioned above, the aim of this thesis is to help Animal Equality and other animal rights organizations to improve their strategic social media communication. According to Cornelissen (2011), a good communication strategy involves seven steps: defining the strategic intent; defining the communication objectives; identifying the target audiences; identifying themed messages; developing message styles; developing a media strategy, and preparing the budget. The focus of this paper will be on the fourth and fifth steps, in that the identification of the right messages is what Animal Equality needs more at the moment; nevertheless, a brief overview of what has already been done in the animal rights movement with respect to the identification of the strategic intent and the target audience will be provided here since it will help to understand the decisions underlying the method of this study.
As for the strategic intent, animal rights organizations share the one of preventing animal suffering. In this direction, the messages that animal rights organizations use on their social media are principally focused on achieving legal
changes for the animals, creating a mindset of helping animals, making people question the relationship with animals, showing the truth of animal exploitation and positioning animals as having rights (Bertolaso, 2014; Freeman, 2010). A study by Swan and John (2003) examined the contributions to the animal rights issue on internet sites and found that pro animal rights sources use two principal argumentative strategies. The first strategy depicts animal use as a moral problem by attributing rights to non human animals trough discourses of suffering, oppression, and depravity, while the second strategy defines animal rights as something mutually reinforcing human well-being by presenting animal exploitation as unnecessary and even dangerous for human health. Animal Equality’s main strategic intent fits with the first strategy, being defined by the organization itself as “to achieve a long-term social change, where all animals are treated with compassion and respect” (Animal Equality (n. d.). In Facebook [English page]. Retrieved November 10, 2014 from https://the
www.facebook.com/AnimalEquality?fref=ts).
The step of identifying the target audience is extremely relevant for this research. In fact, in order to study which are the best message framing strategies, the identification of the right public is vital. In this sense, an ample range of research has already been conducted in order to investigate the predictors of vegetarian and vegan eating. Several variables have been found to have an impact on dietary change among which gender, living with an animal, anthropomorphic beliefs, cognitive and affective dimensions of empathy towards animals (Apostol, Rebega, & Miclea, 2013), female gender (Beardsworth & Bryman, 1999; Gale, Deary, Schoon, & Batty, 2007; Perry, McGuire, Dianne Neumark, & Story, 2001; Santos & Booth, 1996), altruism (Kalof, Dietz, Stern, & Guagnano, 1999), perceived animal intelligence and appearance (Ruby & Heine, 2012), high IQ in childhood, high social class and high academic or
orientation and weak authoritarism (Allen, 2000). In his compilation of studies regarding vegetarians and their habits, Cooney (2013) concludes that the target
audience of animal rights organizations are females under thirty years of age, and with a high education level. Animal Equality is already applying this knowledge in its campaigns, and is currently in need of a deeper investigation of the most effective ways to influence this specific segment of the population.
The identification of the themed messages and the message style is another important step of a communication strategy, and the principal focus of this thesis. Since the aim of this study is to expand the knowledge about message effectiveness, a look into what has already been discovered will serve as a basis upon which to develop this research.
The first important study to mention regarding animal rights advocacy is from Cherry (2010). This study identified two principal strategies directed at shifting the symbolic boundaries that exist between humans and other animals as well as between companion animals and farmed animals: boundary blurring and boundary crossing. Boundary blurring strategies are in turn divided into two categories: in the first
category there are the focusing strategies; these strategies stress the biological evidence that humans are animals as well and that there is no difference between animals that are culturally loved and respected and animals that are culturally seen as mere resources; the second category is composed by the universalizing strategies, which place non human animals beside humans as victims of violence and compare the animal rights movement to the movement for human rights. Cherry emphasizes the relevance of universalized victimization, which consists into linking oppressions and abuses toward non human animals to those toward humans; this concept is described by the term “speciesism”, which is defined as "a prejudice or attitude of bias in favor of the interests of members of one's own species and against those of members of other
species" (Singer, 1990, p. 6). Boundary crossing strategies implicate physically
crossing the human-animal boundary through graphic descriptions of a human body as body of an animal, or by visually situating companion animals in circumstances that are normally endured by farmed animals in slaughterhouses.
Another study by Freeman (2010) examined how the five major U.S animal rights organizations frame issues and solutions in their campaigns directed to persuade people to modify their food habits. The results showed that the most implemented frames are “commodification”, “cruelty and suffering”, “harms to humans and the environment” and “needless killing”. The organizations involved in the study requested consumers to reduce animal products, adopt a vegetarian diet, or help to achieve some animal welfare reforms. The study revealed that only a little percentage of the
communication was focused on supporting veganism and animal rights, being the most of the messages framed using more mildly positioned arguments such as arguments related to animal welfare.
In the case of Animal Equality, the persuasive attempt of the organization that emerges from the observation of their Facebook page is mainly directed towards two goals: on one hand, to attract and maintain followers through cute images and videos featuring companion animals or rescued farmed animals; on the other hand, to show the brutality to which farmed animals are exposed all over the world through their
investigations, that are given high importance on the page.
What can be concluded is that Animal Equality is clearly attempting to move their publics to a more compassionate attitude towards animals using a mix of the two previously described strategies: moral shocks and individualization.
Moral shocks and individualization - which is more effective?
An ample amount of literature has focused on the relevance of moral shocks for animal rights purposes, but the often contradictory results make it difficult to confirm their actual efficacy.
On one hand, several studies came to the conclusion that moral shocks are the more effective strategy when it comes to changing people’s attitudes toward animals and persuading people to reduce animal products in their diets. A study by Herzog (2009) examined the relationships that exists between the moral emotion of disgust and animal rights activism, attitudes toward animal welfare, and meat consumption. What they found is that using moral shocks in animal rights advocacy was a successful strategy in that it triggered powerful emotional reactions and feelings of disgust which in turn were able to motivate activism. In another study (Nabi, 2009), participants that were exposed to extremely graphic videos featuring images of animal experimentation proved to be willing to change their attitude as a consequences of experiencing strong feelings of disgust.
Effectiveness of moral shocks was also confirmed by McDonald (2000), according to which providing information regarding cruelty towards animals represented an important catalytic event that triggered willingness to have more information and eventually become a vegan.
Another confirmation of the usefulness of moral shocks was found in a study of the Human research Council (Human Research Council, 2012). In that study, the effectiveness of four vegan/vegetarian outreach videos was tested among a sample of randomly selected respondents between fifteen and twenty three years old. The results showed that the graphic Farm to Fridge was more effective in promoting interest in dietary change than videos focused on the environment, health, or happy farm animals, thus demonstrating the higher effectiveness of cruelty framing.
In another occasion (Humane League Labs, 2014) a large-scale study was carried out with the purpose of determining what elements make a pro-vegetarian booklet more likely to inspire young people to reduce animal products intake. Among other variables, the study investigated the effectiveness of focusing on cruelty versus health benefits, and the results suggest that animal rights organizations should design leaflets focused on cruelty.
Recently, another experiment (Humane League Labs, 2015) was conducted in order to investigate which photos of farm animals are most likely to persuade young people to eat vegan, and the results showed that the most effective photos were the ones depicting sick or injured animals, dead animals, animals being killed, and animals maintained in tight confinement.
The results of all these studies suggest that moral shocks can be a very powerful tool for animal rights organizations in order to persuade people to switch their dietary habits and quit animal products. Nevertheless, other studies contradict this assumption. For example Mika (2006) conducted a study in which the effectiveness of PETA’s advertisements on high school students was measured, finding that advertisements that were shocking or threatening students’ deeply detained cultural beliefs received a strong negative reaction and were considered as not trustworthy. The results of the study suggested that moral shocks produced a strongly negative reaction and did not have any persuasive power. Furthermore, the results showed that shocking
advertisements advocating for the adoption of a vegetarian diet as a way to help animals invoked the least interest among students.
In another study (Humane League labs, 2013) the Humane League ran a series of online advertisements to discover which video was more likely to inspire young women to think about changing their diet. Two different videos were shown through Facebook advertising. The first video was “What Came Before”, a video that narrates
the life of a pig called Nikki, a farmed animal that was rescued from the industry. The video focuses on the individuality and uniqueness of Nikki, showing her intelligence and letting the public empathize with her, while simultaneously showing the story of what came before her rescue, when she was living in a factory farm. Farm To fridge is a documentary that reveals the truth of what happens in the US’ largest industrial farms, hatcheries, and slaughter plants and shows the treatment to which farmed animals are subjected on a daily basis. The videos were shown to a targeted sample of 83.000 young American women, tracking the percentage of participants that ordered a vegetarian starter guide after the exposition to the video. Results showed that
participants exposed to What Came Before were 70% more likely to order a vegetarian starter guide than those who saw Farm To Fridge, indicating that focusing on the story of a specific individual was more effective in promoting behavioral change intention that showing the cruel living conditions of farmed animals.
What can be inferred from the totality of these studies is that even if the moral shocks seem to be a powerful strategy when it comes to persuade people to reduce animal products, in some occasions they have also been found to be less effective than the use of individualization and can even produce resistance to change.
Considering this knowledge, it is clear that so far no consensus has been produced about which approach is consistently more effective. For this reason, the following sub questions are formulated:
(A) Does the use of messages involving moral shocks, regardless of prevention or promotion focus, produce a more favorable attitude towards animals and a reduction of animal products consumption as compared to a control condition involving no message exposure?
(B) Does the use of messages involving individualization, regardless of prevention or promotion focus, produce a more favorable attitude towards
animals and a reduction of animal products consumption as compared to a control condition involving no message exposure?
(C) Does the use of messages involving moral shocks, regardless of prevention or promotion focus, produce a more favorable attitude towards animals and a larger reduction of animal products consumption as compared to the use of messages involving individualization?
Promotion focus versus prevention focus - what works better?
As previously discussed, the Regulatory Focus Theory has been widely
investigated as a moderator of health and commercial motivation. Nevertheless, even if to a remarkably minor extent, some research already exists about the functioning of this framing strategy in areas that are relevant for this study, namely areas that involve behaviors which are not focused on one’s own advantage. For example, Zaal, Laar, Stahl, Ellemers, and Derks, (2012) studied the role of regulatory focus in influencing young women’s commitment to collective action directed at achieving social change related to discriminative treatment of women at work in terms of salary and job
opportunities. Participants were either primed with the prevention or promotion focus, and the results showed that the perceived likelihood of collective action to be
successful was able to moderate the promotion focus primed participants’ willingness to participate in it. This study seems to suggest that a regulatory fit can trigger
willingness to engage altruistic behaviors. Nevertheless, whether a promotion focus or a prevention focus is generally more effective, still remains uncertain. For this reason, I formulate the following sub questions:
(D) Does the use of promotion-focused messages (regardless of message frame) produce a more favorable attitude towards animals and a larger reduction of animal products consumption as compared to the use of prevention-focused messages?
(E) Does the use of promotion-focused messages, regardless of moral shocks or individualization, produce a more favorable attitude towards animals and a reduction of animal products consumption as compared to a control condition involving no message exposure?
(F) Does the use of prevention-focused messages, regardless of moral shocks or individualization, produce a more favorable attitude towards animals and a reduction of animal products consumption as compared to a control condition involving no message exposure?
Eventually it has to be considered that animal rights organizations like Animal Equality often use a message (either involving individualization or moral shocks) accompanied by a prevention or promotion focused appeal to quit animal products in their Facebook posts. For this reason, it is my wish to better investigate these
combinations in order to have a better comprehension of their actual effectiveness. (G) Does the use of messages involving a combination of moral shocks and prevention focus, or a combination of individualization and promotion focus, produce the most favorable attitude towards animals and the largest reduction of animal consumption?
Method
Experimental design
For the purpose of this study, a 2 X 2 (frame: shocks vs individualization X regulatory focus: promotion goal vs prevention goal) design was implemented. The study was conducted in two phases.
Manipulation. The independent variables moral shocks and individualization were manipulated by presenting Facebook posts respectively showing the horrifying reality that farmed animals endure in factory farms and the special characteristics of animals presented as unique individuals. Each of these six Facebook post was dedicated
to a specific animal product, namely: beef, pork, chicken, fish, eggs and dairy. The Facebook posts were taken from the English Facebook page of Animal Equality and in some occasion were slightly modified in order to ensure a more strong and
homogeneous manipulation of the independent variables.
Moral shocks. Participants in the moral shocks condition were exposed to the following posts: the first post was titled “sows; mothers in loneliness and despair”. This post featured an image of a visibly distressed sow enclosed into a gestation crate. The accompanying text was “COULD YOU IMAGINE IF… You could not turn around for weeks at a time. ‘Sows’ used for breeding are repeatedly forcibly impregnated
throughout their lives, often severely confined, and then separated from their babies soon after giving birth. The lives of these pigs, and their capacity to reproduce, are seen as no more than a way of creating more units of production. Mothers unable to give birth to re required number of piglets are sent to the slaughterhouse. Learn more at https://www.youtube.com/watch?v=TUxg93ghBVA”.
The second post was titled “An investigation reveals the suffering of tuna” and presented a two minutes video featuring the brutal images from an Animal Equality’s investigation during the traditional tuna slaughter in Italy. The video was accompanied by a text saying “victims of commercial fishing nets tend to die of suffocation, crushed under the weight of the other fish or frozen in the boat’s cooling chambers. Others go through agonizing decompression when they are hauled to the surface and their eyes and internal organs literally explode. Fish hooked on lines or anglers endure having their sensitive mouth parts ripped open by sharp hooks before being clubbed or
suffocated. On fish farms many thousands of fishes are crammed together and forced to swim in circles. Parasites and infections run rampant in fish farms, and can spread to wild fish. Learn more at www.TheKillingOfTuna.org.
The third Facebook post consisted in the text “Do you know how pigs live in factory farms?” accompanied by a thirty seconds video showing the life of a pig from the birth in a gestation crate to the death in a slaughterhouse.
In the fourth Facebook post, entitled “the truth behind the egg industry: baby chicks ground up alive”, a five seconds video was inserted showing the process through which newborn male chicks are ground up alive in the egg industry. The video was preceded by the text “All egg production, whether “free-range” or “battery” involves the death of male chicks. Commercial egg-laying hens start out their lives in hatcheries, where hatched in artificial incubators they will never know the care of comfort of a mother hen. Once hatched, the day-old chicks are separated by “sexers” into females and males. Female chicks go on to become units of production – “layer hens”- while male chicks, as they neither ley eggs or produce meat, they are killed are just a day old. Either gassed, suffocated or minced alive.
The fifth Facebook post was titled “inside slaughterhouses” and featured a seventy seconds video showing images from investigations of Animal Equality in several European slaughterhouses. The video was complemented with the text “Slaughterhouses are hell on Earth. We have investigated several slaughterhouses around Europe. Check out the shocking images we have filmed
https://youtube.com/watch?v=FdYfzP3s3ro.
The sixth and last Facebook post contained a one minute video revealing the process through which in the dairy industry newborn calves are systematically taken away from their mothers. The accompanying text was “To produce milk, cows are forcibly impregnated every year. Her calf is removed within a few days so that the milk meant for him or her can be drunk by us. Just like us, cows have a very strong maternal instinct to nurture and protect their calf and suffer immensely when they are taken away from her. Cows suffer grief and they may bellow and search for their calf for
days. Male calves are of no benefit to the dairy industry, and most are killed and turned into veal at just a few months old. Consuming dairy products directly supports the veal industry and the death of these calves. Watch this short, non-graphic video and learn where dairy products come from hhttp://www.youtube.com/watch?v=Q4QNi3WZdQ.
In order to ensure a more attentive exposition to the Facebook posts, trivial questions like “would you open the link?”(yes; no) and “how much of the video did you actually see?” (measured with a 5 points Likert scale: 1= I did not open it; 5= I watched it entirely) where asked after each post.
Individualization. As for the Individualization condition, the first post “Milkshake, a rescued cow with identity issues”, showed a ninety seconds video starring a rescued cow that believes she is a dog. The video was accompanied by the text “Meet Milkshake, a rescued cow with identity issues. Who said that cows are different from cats and dogs? Learn how Milkshake’s end could have been http://vimeo.com/4270529“.
The second post was entitled “Leon Trozky, a tiny pig with a strength that will move you to tears” featured a 2 minutes video of the story of Leon Trozky, a pig in wheelchair. The accompanying text said “Meet Leon Trozky, a super cute piggy in wheelchair. When his injured mother fell on him, he suffered a broken and dislocated hind femur, along with an injured hock and ligament damage on the other side. But no one told the plucky little Leon that these obstacles of his world should hold him back. His life was precious, and thankfully this was recognized by a kind hearted soul who sought a better outcome for Leon. Learn how Leon Trozky’s life could have been http://www.youtube.com/watch?v=LVmloBvnHvY. Learn how Leon Trozky’s end could have been http://vimeo.com/1321076”.
In the third post, entitled “Javier loves to be dried with and hair dryer” the fifteen seconds video of a rescued chicken enjoying to be dried by his caregiver was
shown, accompanied by the text “We just love this!! Javier is a chicken rescued from the meat industry and living nor in Wings of Heart Sanctuary. As you can see, he has this peculiar habit: he loves to feel the warm air of a hair dryer on his body. He really enjoys it!! Did you know? Chickens like Javier have individual personalities and are intelligent creatures that live in social hierarchies, make friends with chickens they know and avoid the ones they do not know. “Share” if you think that chickens are friends, not food!”
The fourth post was named “Lucas, a happy rescued calf” and showed the image of a cute baby calf being fed by his caregiver with the text “Lucas is a calf rescued from the dairy industry. He is now a healthy, happy calf who lives in a
sanctuary, loves being pet, running with his friends and sleeping in a comfortable bed. Before being rescued, Lucas used to live unchained in a small box, alone, and was found crying, desperately calling his mother. In fact, to produce milk, Lucas’ mom was forcibly impregnated and he was removed within a few days so that the milk meant for him could be drunk by us humans. Learn more about the dairy industry at
http://www.animalequality.net/food/milk”.
In the post number five the text “Aren’t these bay adorable? As every baby, they look for the warm protection of their mother. And they don’t seem to care about the fact that their mother is a dog. Did you know that all egg production, whether “free range” or battery involves the death of male chicks? Learn more at
http://www.animalequality.net/food/eggs” accompanied a thirty seconds video featuring a Golden Retriever behaving as a mother for a litter of baby chicks entitled “adorable baby chicks have a dog for mom”.
The sixth and last post, titled “fish are friends, not food” featured a one minute video of a fish interacting with a man accompanied by the text “You can’t miss this video of a fish that loves being petted! Did you know? Contrary to what many people
think, fishes are individuals with the capacity to experience pleasure and pain, as confirmed by numerous scientific studies. If you don’t believe it, look how this super sociable fish loves cuddles!”
Manipulation check. In all the four surveys questions were included in order to check for the success of the manipulations of both independent variables moral shocks and individualization. The question “to which extent does this post lead you to feel that you want to do something about it?” for the moral shocks condition and the question “To which extent does this post lead you to think that cows/chickens/pigs/fish are not all the same but are instead individuals with their own traits, preferences and
personalities?” measured with a 5 points Likert scale (1= not at all; 5 = very much) were asked after each Facebook post.
Dependent variables: promotion – prevention. The independent variables “promotion focus” and “prevention focus” were manipulated by including a text after the section containing the six Facebook posts.
Promotion focus: In the moral shocks condition the promotion focus was manipulated by priming it in the title “Each vegan saves 200 animals per year” and creating the regulatory fit by delivering the following message:
” Dear participant,
We can imagine how it makes you feel now to think of all the atrocities you have just witnessed. For this reason, we want you to know that there are very good news: we can change the life of all these animals in our everyday life, and it is extremely easy! We just have to choose a different menu.
Did you know that only by choosing plant based alternatives to animal products, it is possible to spare nearly 200 animals per year? For each person that will stop
consuming animal products, 200 animals will be spared from being taken away from their mothers, from spending a life in misery and despair confined to filthy cages or
tanks, and from dying in slaughterhouses with their throats cut or for suffocation when pulled out from the deep sea. For each new person that adopts a vegan diet, all these animals will be spared. Don't you think it is worth trying?”
As for the Individualization condition, the promotion focus was induced by priming it in the title “Each vegan saves 200 animals per year” and creating the regulatory fit with the following message:
“Dear participant,
You have just seen that farmed animals are intelligent, beautiful and sensitive creatures that, just like cats and dogs, have their own personalities, enjoy relationships, and are completely capable of experiencing feelings and emotions.
We would like you to know that only in the United States 80.000 millions of individuals like Leon Trozky, Lucas and Milkshake live the most painful and miserable lives you can imagine in factory farms and are killed every year for human consumption. We want you to know that we can change the life of all these animals in our everyday life, and it is extremely easy! We just have to change a different menu.
Did you know that only by choosing plant based alternatives to animal products, it is possible to spare nearly 200 animals per year?
For each person that will stop consuming animal products, 200 animals will not be taken away from their mothers, will not spend a life in misery and despair confined to filthy cages or tanks, and will not die in slaughterhouses with their throats cut or for suffocation when pulled out from the deep sea.
If people change their dietary habits, all these animals will be spared. Don't you think it is worth trying?”
Prevention focus: In the moral shocks condition, the prevention focus was manipulated by priming it in the title “Each vegan prevents 270 animals per year from being killed” and creating the regulatory fit with the following message:
“Dear participant,
We can imagine how it makes you feel now to think of all the atrocities you have just witnessed. For this reason, we want you to know that there are very good news: we can change the life of all these animals in our everyday life, and it is extremely easy! We just have to change a different menu.
Did you know that an average American consumes around 270 animals per year? For each person that will stop consuming animal products, 270 animals will not be taken away from their mothers, will not spend a life in misery and despair confined to filthy cages or tanks, and will not die in slaughterhouses with their throats cut or for suffocation when pulled out from the deep sea. For each new person that adopts a vegan diet, all these animals will not die. Don't you think it is worth trying?”
In the Individualization condition, the prevention focus was manipulated by priming it in the title: Each vegan prevents 270 animals per year from being killed and creating the regulatory fit by delivering the following message:
“Dear participant,
You have just seen that farmed animals are intelligent, beautiful and sensitive creatures that, just like cats and dogs, have their own personalities, enjoy relationships, and are completely capable of experiencing feelings and emotions.
We would like you to know that only in the United States 80.000 millions of individuals like Milkshake, Javier, Lucas and Leon endure the most painful and miserable lives you can imagine in factory farms and are killed every year for human consumption.
We want you to know that we can change the life of all these animals in our everyday life, and that it is extremely easy! We just have to change a different menu.
Did you know that an average American consumes around 270 animals per year? For each person that will stop consuming animal products, 270 animals will not be taken away from their mothers, will not spend a life in misery and despair confined to filthy
cages or tanks, and will not die in slaughterhouses with their throats cut or for suffocation when pulled out from the deep sea. For each new person that adopts a vegan diet, all these animals will not die. Don't you think it is worth trying?”
Manipulation check. To check for the success of the manipulation an item was added to the survey immediately after the above described messages asking respectively “To which extent did the previous message make you feel that going vegan can actually spare animals' lives?” for the promotion focus condition and “To which extent did the previous message make you feel that going vegan can actually prevent the negative consequences of a non vegan diet on the animals? For the prevention focus condition. Both items were measured with a 5 points Likert scale (1 = not at all; 5= very much).
Structure of the surveys. All the five surveys of the first phase had an initial common section in which participants were asked to indicate their biological gender (male; female), chronological age (< 15; 15-35; >35) and diet (omnivore; pescatarian; vegetarian; vegan; other). For the purpose of the study only non-vegan females between 15 and 35 years were considered; the rest of participants were automatically directed to a dedicated page in which a text thanked them for their willingness to participate and explained to them the reason why their answers could not be taken into account. All the five surveys continued with other sections aimed at measuring the two dependent variables measured in this study: Attitudes toward farmed animals and actual animal products consumption.
The second section consisted of five questions from the Animal Attitude Scale (Herzog, Betchart, & Pittman, 1991). In its original version, the scale consisted of 29 items. Since factor analysis indicated that all of them converged into one factor, only some of the items were used for this study, accordingly with its specific aim of
analyzing attitudes towards farmed animals. The following five items were used: 1 -“I think people who object to raising animals for meat are too sentimental”; 2 - I think it
is perfectly acceptable for cattle and hogs to be raised for human consumption”; 3 -“Basically, humans have the right to use animals as we see fit”; 4 - Breeding animals for their skin is a legitimate use of animals”; 5 - The production of inexpensive meat, eggs, and dairy products justifies maintaining animals under crowded conditions”. In order to adjust them for the specific purposes of the present study, the items 2 and 4 were slightly modified into respectively “I think it is perfectly acceptable for cows, pigs and chickens to be raised for human consumption” and “Breeding animals for their meat is a legitimate use of animals”. Animal attitude is originally measured with a 5 points Likert scale where a high score indicates pro animal welfare attitudes (5=
Strongly Agree; 1= Strongly Disagree). Since the five items selected for this study were reverse- worded, the Likert Scale in the survey was also reversed in such a way that 1= Strongly Agree and 5= Strongly Disagree.
The five surveys proceeded then with a third common section: a table in which participants were asked to indicate their animal products consumption by indicating the number of times (0: 1-5; 6-10; 11-15; 16-20) they consumed each food (poultry; fish; beef; pork; eggs; dairy) during the past week. In order to help participants to be more accurate in their answers, the text of the question asked to please consider all the forms in which each food was consumed.
Before presenting the same above described final section, the four experimental conditions surveys presented another section that included the manipulations. Both the surveys of the moral shocks condition and the individualization condition presented the previously described 6 Facebook posts.
Timing of the study. On Friday November 21 Animal Equality sent 400.000 newsletters and published a post in all their Facebook pages asking to participate to the study with the following message:
“Do you want to help us be more efficient in our work on behalf of animals? If you are a woman between the age of 15 and 35 and you have a good comprehension of English, you can participate in our survey today.
Animal Equality is constantly striving to improve the quality of our work for animals. For this reason, we are working on a study that will allow us be more effective in what we do. If you would like to help us to save more lives, participate today!
The survey consists of a questionnaire of 13 minutes (open from 21 until 23 November) and a second questionnaire of 2 minutes (open from 30 until November 1.
Want to help? Click here and get started now! ---> Http://goo.gl/jwhnX8”.
The possibility to win one randomly assigned prize of 100€ in case of completing both phases was announced in the first page of the survey in order to motivate the participation to the entire study.
The link randomly led each participant either to one of the four experimental conditions surveys or to the control group survey of the first phase.
Once submitted the completed first survey each participant received a message saying that he/she would receive the second part in a 10 days. On 29 November the second survey was sent to all the mails of valid participants (non vegan females between 15 and 35 years old), while the rest was directly excluded. The survey of the second phase consisted in two sections exactly identical to the ones of the first phase: participants were asked to complete the Animal Attitude Scale and to fill the table regarding the animal products consumed in the past seven days.
Data processing. In order to prepare the datasheet, I took the following steps: first, I transformed the AAS response categories (ordinal data) into quantitative
(interval) data (‘totally agree’ = 1, ‘agree’= 2, ‘not agree/not disagree’ = 3, ‘disagree’ = 4, and ‘totally disagree’ = 5) for each of the 5 items. Subsequently, I summed the 5 item scores and calculated the average, resulting in a single total AAS score (range:
1-5) for each participant at T0 and T1. Since only 5 items of the original scale were used, the Cronbach’s α was calculated and the new 5-item AAS scale proved to be reliable both at T0 (Cronbach’s α = .88) and T1 (Cronbach’s α = .88).
As for the APC data, I also transformed the ordinal data into quantitative data by recoding the response categories to single numbers (‘0 times APC per week’ = 1, ‘1-5 times APC per week’ = 2, ‘6-10 times APC per week’ = 3, ’11-1‘1-5 times APC per week’ = 4, ‘16-20 times APC per week’ = 5) for each of the 6 questions. Next, the 6 individual scores were summed and averaged, resulting in a single total APC score (range: 1-5) for each participant at T0 and T1.
Finally, I labeled the different experimental conditions (control = 0, moral shocks + promotion = 1, moral shocks + prevention = 2, individualization + promotion = 3, individualization + prevention = 4).
I transformed the value of the variable “condition” so that control → 0; moral shocks + promotion → 1; moral shocks + prevention → 2; individualization +
promotion → 3; individualization + prevention → 4.
Statistical analysis. The research questions were addressed using seven separate mixed-model ANOVAs including experimental condition as a between-subjects factor (2 levels: experimental condition/treatment and control condition) and time of measurement as a within-subjects factor (2 levels: T0 = baseline/pretreatment and T1 = one week later/post-treatment), separately for AAC and APC scores.
Significant interactions between Time and Condition, if present, were followed up by tests of simple main effects using one-factor ANOVAs (with either Time or Condition included as a factor). For all analyses, the statistical level of significance was set at .05. The effect size estimate, partial-eta squared (p), is also reported to provide
information about the strength of the experimental effects, that is, the size of the observed differences in the AAS and APC scores between the different conditions.
Results
Table 1 (see appendix) presents the mean total scores on the AAS and APC measured before (T0) and after (T1) exposure to the experimental treatments.
Question A. To test the effect associated with exposure to moral
shocks, regardless of promotion or prevention, two one-way repeated-measures ANOVAs were conducted, separately on the AAS and APC scores, in which the participants in Condition 1 and Condition 2 were combined (n = 214) and compared to the control group (n = 107). The ANOVA on the AAS scores showed no significant main effects of Time, F (1, 31) = 3.47, p = .063, p = .011, or Condition (Condition 1 + Condition 2), F (1, 31) = 0.49, p = .485, p = .011; the interaction was also not significant, F (1, 31) = 0.39, p = .531, p = .001. These results indicate that no significant differences existed, averaged across T0 and T1, between the condition involving moral shocks and the control condition. Neither existed differences for both conditions in the AAS scores between T0 and T1 (Table 1).
The ANOVA on the APC data, however, showed a significant main effect of Time, F(1, 31) = 24.38, p < .001, p = .071, indicating that the APC score observed at T1, averaged across the two conditions, was significantly lower than the APC score measured at T0 (M = 1.53, SD = 0.36 versus M = 1.63, SD = 0.43) (Table 1). No effect of Condition (Condition 1 + Condition 2) was found, F (1, 31) = 2.36, p = .125, p = .007, but the interaction between Time and Condition proved to be significant, F (1, 31) = 7.71, p = .006, p = .024. The interaction effect signifies that the decrease in the APC score from T0 to T1 was different for the experimental condition as compared to the control condition (Figure 1).
To examine the interaction in greater detail, tests of simple main effects of Time were performed, using a repeated-measures ANOVA separately for each condition. The
results revealed that the main effect of Time was not significant for the control condition, F (1, 10) = 1.99, p = .162, p = .018, but was significant for the
experimental condition (T0: M = 1.61, SD = 0.43 versus T1: M = 1.5, SD = 0.35; F (1, 21) = 42.17, p < .001, p = .165). Accordingly, whereas the control condition
exhibited no systematic change from T0 to T1, the total APC score showed a significant decline over time after being exposed to the experimental condition (Condition 1 + Condition 2) involving moral shocks (Figure 2).
Question B. To test the effect associated with exposure to
individualization, regardless of promotion or prevention, the same procedure as above was followed: two mixed - model ANOVAs were conducted, separately on the AAS and APC scores, in which the participants in Condition 3 and Condition 4 were combined (n = 190) and compared to the control group (n = 107). The ANOVA conducted on the AAS showed no significant main effects of Time, F (1, 29) = 2.09, p = .150, p = .007, or Condition (Condition 3 + Condition 4), F (1, 29) = 0.68, p = .412,
p = .002; the interaction was also not significant, F (1, 29) = 0.03, p = .873, p < .001. These results suggest that the AAS score did not change systematically from T0 to T1, neither for the control condition nor the individualization condition (Condition 3 + Condition 4) (Table 1 and Figure 1).
The ANOVA conducted on the APC score, however, showed a significant main effect of Time, F(1, 29) = 13.28, p < .001, p = .043, indicating that the APC score, averaged across the two conditions, observed at T1 was significantly lower than at T0 (M =1.55, SD = 0.35 versus M = 1.6, SD = 0.4) (Table 1). No significant effect of Condition (Condition 3 + Condition 4) was found, F (1, 29) = 2.81, p = .095, p = .009; the interaction between Time and Condition was also not significant, F (1, 29) = 0.68, p = .178, p = .006. These findings demonstrate that the APC score was
significantly lower at T1 than at T0, but these differences did not vary between the control condition and the individualization condition (Condition 3 + Condition 4) (Figure 2).
Question C. To compare the effect of messages involving moral shocks to the effect of messages involving individualization, regardless of message focus (prevention or promotion), two mixed-model ANOVAs were carried out. For these ANOVAs, the participants in Condition 1 and Condition 2 were combined to form the condition involving moral shocks (n = 214) and compared to the participants in Condition 3 and Condition 4 involving individualization (n = 190). For the AAS score, a significant Time main effect was detected, F (1, 40) = 6.77, p = .010, p = .017, reflecting that this score, across the two conditions, was significantly higher at T1 than at T0 (M = 4.34, SD = 0.86 versus M = 4.24, SD = 0.94) (Table 1). However, the interaction between Time and Condition did not approach significance, F (1, 40) = 0.37, p = .542,
p = .001, which indicates that the increase in AAS score from T0 to T1 did not differ between the condition involving moral shocks and the condition involving
individualization. No significant differences across T0 and T1 were present in the AAS score between these two conditions, F (1, 40) = 0.02, p = .891, p < .001.
The ANOVA on the APC score also yielded a main effect of Time, F (1, 40) = 57.07, p < .001, p = .124, reflecting a lower score at T1 that at T0 (M = 1.5, SD = 0.35 versus M = 1.6, SD = 0.41) (Table 1). No significant main effect of condition was observed, F (1, 40) = 0.01, p = .920, p < .001. The interaction between Time and condition, however, approached, though did not exceed, the level of significance, F(1, 40) = 3.35, p = .068, p = .008, which reflects that the decrease in the APC score tended to be larger following the condition involving moral shocks than the condition involving individualization (Table 1 and Figure 2).
Question D. To compare the effect of promotion-focused messages to the effect of prevention-focused messages, regardless of moral shocks or individualization, two mixed-model ANOVAs were performed, in which the participants in Condition 1 and Condition 3 were combined to form the condition involving promotion-focused messages (n = 198) and compared to the participants in Condition 2 and Condition 4 involving prevention-focused messages (n = 206). The results showed a significant Time main effect, F (1, 40) = 3.35, p = .009, p = .017, on the AAS score, while the main effect of condition, F (1, 40) = 1.05, p = .307, p = .003, and the two-way interaction were not significant, F (1, 40) < 0.01, p = .957, p < .001. This result means that the AAS score exhibited an overall increase from T0 to T1 (M = 4.24, SD = 0.94 versus M = 4.34, SD = 0.86), yet this increase did not differ between the condition involving promotion-focused messages (Condition 1 + Condition 3) and the condition involving prevention-focused messages (Condition 2 + Condition 4) (Table 1).
However, for the APC score, a significant Time main effect, F (1, 40) = 60.39, p < .001, p = .131, was found, along with a significant interaction, F (1, 40) = 7, 66, p = .006, p = .019, and a non-significant condition main effect, F (1, 40) = 0.18, p = .669,
p < .001. Follow-up tests demonstrated that the Time effect was significant for the condition involving promotion-focused messages (Condition 1 + Condition 3), F (1, 19) = 40.33, p < .001, p < .170, and the condition involving prevention-focused messages (Condition 2 + Condition 4), F (1, 20) = 19.25, p < .001, p < .086. These findings reflect that the APC score decreased from T0 to T1 for each of the two conditions, but -as evidenced by the significant two-way interaction- the decline in the APC score was systematically larger following promotion-focused messages
(Condition 1 + Condition 3) (T0: M = 1.61, SD = 0.45 versus T1: M = 1.48, SD = 0.35) than following prevention-focused messages (Condition 2 + Condition 4) (T0: M = 1.6,
SD = 0.38 versus T1: M = 1.53, SD = 0.33). Indeed, as reflected by the effects size estimate (p), the exposure to promotion-focused messages accounted for
approximately 17% of the time-related variance in the APC scores, whereas this value was approximately 8.6% following prevention-focused messages
Question E. To test the effect associated with exposure to promotion-focused messages, regardless of moral shocks or individualization, two mixed – model
ANOVAs were conducted, separately on the AAS and APC scores, in which the participants in Condition 1 and Condition 3 were combined (n = 198) and compared to the control group (n = 107). The ANOVA conducted on the AAS score showed no significant main effects of Time, F (1, 30) = 3.16, p = .076, p = .010, or Condition (Condition 1 + Condition 3), F (1, 30) = 1.46, p = .228, p = .005; the interaction was also not significant, F (1, 30) = 0.18, p = .674, p = .001. Accordingly, no significant differences existed in the AAS scores between time of measurements and/or
conditions.
On the other hand, the ANOVA conducted on the APC score showed a
significant main effect of Time, F(1, 303) = 24.82, p < .001, p = .076, indicating that the APC score observed at T0, averaged across the two conditions, was significantly lower at T1 (M = 1.51 , SD = 0.36 versus M = 1.61, SD = 0.43) (Table 1 and Figure 2). No main effect of Condition (Condition 1 + Condition 3) was found, F (1, 30) = 3.03, p = .083, p = .010, but the interaction between Time and Condition was found to be significant, F (1, 30) = 8.89, p = .003, p = .003, indicating that the decrease in the APC score from T0 to T1 was different for the experimental condition as compared to the control condition (see Figure 1).
To examine the interaction in greater detail, tests of simple main effects of Time were performed, using a repeated-measures ANOVA separately for each condition. The
results revealed that the main effect of Time was not significant for the control
condition, F (1, 10) = 1.99, p = .162, p = .018, as reported in relation to Hypothesis 1, but was significant for the experimental condition (T0: M = 1.61, SD = 0.43 versus T1: M = 1.46, SD = 0.35; F (1, 19) = 40.33, p < .001, p = 170). Correspondingly, whereas the APC score did not differ as a function of time of measurement in the control
condition, the score was markedly decreased in the experimental condition (Condition 1 + Condition 3) involving exposure to promotion-focused messages (Table 1 and Figure 2).
Question F. To test the effect associated with exposure to prevention-focused messages, regardless of moral shocks or individualization, the same procedure was used in which the participants in Condition 2 and Condition 4 were combined (n = 206) and compared to the control group (n = 107). The ANOVA conducted on the AAS score showed a not significant main effects of Time, F (1, 31) = 2.52, p = .114, p = .008, nor Condition (Condition 2 + Condition 4), F (1, 31) = 0.12, p = .731, p < .001; the interaction was also not significant, F (1, 31) = 0.17, p = .685, p = .001. Thus, again, the AAS score did not differ between time of measurement and/or condition.
Nevertheless, the ANOVA conducted on the APC score, showed a significant main effect of Time, F (1, 31) = 13.50, p < .001, p = .042 indicating that the APC score, averaged across the two conditions, was significantly lower at T1 than T0 (M = 1.55, SD = 0.35 versus M = 1.61, SD = 0.4) (Table 1 and Figure 2). No significant main effect of Condition (Condition 2 + Condition 4) was found, F (1, 31) = 2.14, p = .144, p = .007, and the interaction between Time and Condition was also not
significant, F (1, 31) = 1.28, p = .259, p = .004. These results indicate that the APC score decreased from T0 to T1, but that the decrease did not differ between the control
condition and the experimental condition (Condition 2 + Condition 4) involving exposure to prevention-focused messages (Table 1 and Figure 2).
Question G. To evaluate the question if messages involving a combination of moral shocks and prevention (i.e., Condition 2), or involving a combination of individualization and promotion (i.e., Condition 3), are associated with the greatest changes in AAS and APC scores from T0 to T1, two mixed-model ANOVAs were performed, in which these two conditions were compared to each other, as well as to the condition involving moral shocks and promotion (i.e., Condition 1), the
experimental condition that seemed to be most effective, at least as reflected by the greatest change in the APC score over time (Table 1 and Figure 2). The ANOVA on the AAS score yielded a significant Time main effect, F (1, 30) = 4.02, p = .046, p = .013, but the main effect of condition, F (2, 30) = 1.49, p = .227, p = .010, and the two-way interaction were not significant, F (2, 30) = 0.99, p = .374, p = .006. Therefore, although the AAS score, averaged across the three conditions, was
significantly higher at T1 than at T0 (T0: M = 4.34, SD = 0.88 versus T1: M = 4.24, SD = 0.96, n = 306), the time-related increase did not differ significantly between the three individual conditions.
In contrast, the ANOVA on the APC score produced a significant Time main effect, F (1, 30) = 57.88, p < .001, p = .160, as well as a significant two-way
interaction, F (2, 30) = 3.14, p = .045, p = .020; the main effect of condition was not significant, F (2, 30) = 0.19, p = .828, p = .001. Pairwise comparisons revealed that the difference between the APC scores at T0 and T1 observed in Condition 1 was significantly greater than the difference observed in Condition 2, Time-by-Condition: F(1, 21) = 5.57, p = .019, p = .026, but was not significantly different from that seen in Condition 3, Time-by-Condition: F(1, 19) = 2.31, p = .130, p = .012; Condition 2
