The Road Towards P-less-tic:
The influence of narratives, factual messages and involvement
on persuasive outcomes and the level of online message
engagement concerning plastic waste separation
Master’s Thesis
Name student: Bart Tetteroo Student number: 10198423
Institution: University of Amsterdam
Graduate School of Communication Program: Persuasive Communication
Supervisor: Dr. S. Mollen
Date: 29-06-2018
Abstract
Plastics have become increasingly important in global daily lives. They are used in almost every possible way, but the consequences are devastating. Now, almost five billion kilos of plastic is floating somewhere in the ocean. How can such a tidal wave of plastics be stopped from swallowing the earth? The goal of the current study is to examine whether the use of narrative messages compared to factual messages can improve people’s attitude and motivation towards separating plastic waste, and what the role of level of involvement in the topic sustainability is. It is also examined if narrative messages could increase the level of message engagement someone has with online Facebook messages, and if this engagement would mediate the effect of
message type on attitude and intention. This is important, because there is a lack of research in concerning the influence online engagement behavior can have on persuasive outcomes. An online experiment among 234 Dutch participants was conducted to answer these questions. Participants were exposed to one of two online Facebook messages (narrative vs factual), ostensibly posted by the Ocean Cleanup. It was hypothesized that exposure to narrative messages would have a more positive effect on intended message engagement, attitude and intention than factual messages. Also, a higher involved person would have a more positive attitude and intention than a less involved person, and this higher level of involvement would moderate the effect of message type on message engagement, attitude and intention. At last, it was
predicted that the higher the level of engagement was, the more positive the attitude and intention would be. Contrary to expectations the type of message did not
influence the message engagement, attitude or intention. The level of involvement did not effectuate the attitude and intention, and there were no interaction effects between message type and involvement. Nevertheless, the level of intended message
engagement and the attitude and intention did correlate positively. No other significant effects were found. Future research should focus on the actual message engagement instead of the intended message engagement, and on the factor of identity, because that could be a relevant predictor of the level of message engagement.
Introduction
Plastic is a wonderful creation; it is strong, can be used in many different ways and, most of all, plastic is produced at very low costs. But one person’s gain is another person’s loss. Straws, plastic bags, plastic bottles, plastic wrappers for fruits and vegetables are common items that are used globally on a daily basis. More than 4.7 billion kilos of plastic are floating somewhere in the ocean (Jansen, 2014).
Unfortunately, many people are using plastic more and more every day, because of its low cost (Laville & Taylor, 2017). The widespread use of plastic has devastating consequences, not only for the environment, but also for human wellbeing, as plastic ends up in the food chain and could ultimately be consumed by human beings (Derraik, 2002).
That is why human behavior must be changed to achieve environmental sustainability for the earth (Oskamp, 2002). It is very important to increase awareness about plastic pollution, make consumers feel positive about solving this problem, and ultimately to motivate people to engage in pro-environmental behavior (Kollmuss & Agyeman, 2002). One way to reduce the impact of plastics on the environment is to recycle plastics, because recycling prevents plastic from ending up in nature, while at the same time the need for new resources decreases (Gourmelon, 2015). That is why it is important that consumers separate their plastic waste from the rest, so that plastic waste can be recycled. In Europe, the average recycling rate per country is only 39%, so a lot of progress could be made (European Environmental Agency, 2017). To achieve this behavioral change, it is necessary to create optimal persuasive message strategies. An often-used message type to achieve behavioral change is a narrative message. It is therefore interesting to examine what influence such a message type can have on persuasive outcomes.
A narrative message is a message structure wherein arguments are being brought to the audience through an indirect, story-like format, in which a character experiences the consequences of a certain behavior (Wentzel, Tomczak, & Herrmann, 2010). A message type that is opposite to a narrative is a factual message: a message structure wherein arguments are being brought to the audience through a direct, explicit and informational (factual) format, with no specific characters included (Wentzel et al., 2010). For instance, when an organization like a restaurant wants to address the care and hard work they perform to be able to serve products with the
highest quality, it could be done by telling it from the perspective of an employee and how s/he experiences a hard working day in the restaurant or by using no story-like format and just summing up the facts about how the restaurant tries to preserve the highest quality. Several researchers have studied the effectiveness of narrative vs factual messages, with divergent results. Some have found that the persuasive effects of a narrative are stronger than a factual frame (e.g. Mattila, 2000), while others claimed that these effects do not always occur (e.g. Zebregs, Van den Putte, Neijens, & De Graaf, 2015). It is therefore important to find some clarity concerning the use of these different types of messages.
When studying different types of messages, a factor that is often studied in combination with exposure to different message types is the boundary effect of one’s personal perceived relevance of the topic environmental sustainability based on inherent needs, values, and interests, or involvement (Zaichkowsky, 1985). While some have found that people that have a high level of involvement will generally be stronger influenced by factual messages, and people that have a low level of
involvement will generally be stronger influenced by narrative messages (Braverman, 2008), others have found no difference between these two. With the current research, the author tried to fill (parts of) these gaps so that the scientific knowledge concerning these topics will be more consistent.
More and more of the campaigns to promote sustainability take place on the Internet. Because people spend more time on online rather than offline media nowadays, it is important to study the effects of persuasive messages in an online context (Wirtz, Schilke, & Ullrich, 2010). Online organizations are trying to acquire as much engagement with their messages as possible. People can engage in for
instance Facebook posts in forms of liking, sharing or commenting on those messages (Kabadayi & Price, 2014). People tend to engage with messages, because they believe that those messages contain information that is important to them (Fulgoni, 2015).
The type of message that people receive could influence the level of message engagement. However, current scientific research lacks knowledge about the
influence a narrative (vs factual) structure and the level of involvement could have on message engagement, and what effects this engagement could have on persuasive outcomes. The storytelling factor in narratives could be a key factor to increase the level of message engagement (Yuki, 2015). The role that message frames can play in the level of intended message engagement is hardly examined and is therefore a fairly
new side of research in communication science, which makes it an important topic to study.
Nowadays, people spend hours and hours on social media, liking, sharing and commenting on dozens of ‘posts’ (Wirtz et al., 2010), it is however unclear whether this online engagement can also cause more positive feelings towards the advocated behavior and a higher intention to engage in such behavior among people. Or will they have the feeling that liking, commenting on or sharing a message is satisfying enough to think that a person ‘has already done his/her part’ and will not perform in the target behavior anymore? By studying this, the author tried to fill this gap in scientific knowledge concerning the effect that different message types can have on the level of intended message engagement, and what effect it could have on
persuasive outcomes. Therefore, the following research question has been drafted:
RQ: Does an online narrative message or an online factual message have a more positive effect on attitude towards separating plastic waste and intention to separate plastic waste directly and through message engagement? Furthermore, does
involvement with the topic sustainability moderate the effect of the different message frames on message engagement, attitude and intention?
Theoretical Framework
Message type
When it comes to the environment, it is of utmost importance to frame messages in a way that proves the necessity of action to preserve and protect the environment (Lakoff, 2010). One of many ways to frame a message is to provide the information of the message in a direct, explicit and informational (factual) way. Alternatively, the information in a message can be framed as a narrative, wherein arguments are brought to the audience in an indirect story-like format (Wentzel, et al., 2010). Braddock and Dillard (2016) define narratives as story-like messages that contain one or more events, wherein one or more characters perform purposeful actions that are linked to a positive outcome. So, narratives tend to follow a chronological timeline and indicate causality between a behavior and an outcome (Escalas, 1998), and involve one or more characters (Adaval & Wyer, 1998). Factual messages, on the other hand, do not have such a story-like format and do not contain
‘actors’ that perform certain actions. However, the information presented in a factual message could be the same as in a narrative message, the difference lies in the format that it is presented in (Wentzel et al., 2010).
By experiencing a message in a story-like format, people can create a link between different events and content in a story like the behavior and the consequences of that behavior, as they would do in real life, which makes it easier to process the given information (Wentzel et al., 2010). When an actor in a story-format for instance shows that separating plastic waste is good for the environment by separating his/her own plastic waste, it is much easier to process than when the same information is presented without actor and by summing up the facts. Also, people tend to think less critically about information given in narrative form then when given in a more factual or statistical form, because narratives give the impression of just being an often fun story, rather than containing a persuasive message (Green & Brock, 2000). Moyer-Gusée (2008) refers to this as reactance; people have a strong need to choose what they want for themselves, instead of others deciding it for them. When someone else tries to overtly persuade them to do something they did not choose themselves, their resistance towards that person or message increases. By bringing the message through a narrative format, this feeling of reactance can be reduced.
Because the message is brought in a story-like format, a narrative can evoke certain emotional responses (Escalas, 2004; Kopfman, Smith, Ah Yun & Hodges, 1998; Mazzocco, Green, Sasota, & Jones, 2010; McQueen, Kreuter, Kalesan, & Alcaraz, 2011; Mattila, 2000). On the contrary, it would be more unlikely that a factual message would evoke such emotional responses, because factual messages are read and processed in an analytical way (Green & Brock, 2000, Wentzel et al., 2010). Narratives have shown to positively affect consumer attitudes and behavioral intentions and other persuasive outcomes in general compared to statistical messages, according to a meta-analytical study by Braddock and Dillard (2016). Murphy, Frank, Chatterjee and Baezconde‐Garbanati (2013) and the meta-analysis of van Laer, de Ruyter, Visconti and Wetzels (2013) found that a narrative message could positively influence attitudes because of the emotional responses that a narrative could evoke. According to De Graaf, Sanders and Hoeken (2016), Mattila (2000) and Adaval and Wyer (1998), a narrative can cause a more positive attitude and stronger behavioral intentions than a factual message.
Other meta-analytical research, on the other hand, has concluded that factual messages could have a more positive effect on attitudes, while narrative messages have a more positive effect on intentions (Zebregs, et al., 2015). However, the differences found in this meta-analysis were marginal for both attitude and intention. An empirical study of Zebregs, Van den Putte, De Graaf, Lammers and Neijens (2015) that was not included in prior meta-analyses even found that no differences in effect could be measured for attitude between factual and narrative texts, while none of the two types of messages had an effect on intention. But all in all, it seems that narrative messages can evoke more positive effects in terms of attitude and intention than factual messages. It is therefore predicted that a narrative message has a more positive effect on attitude and intention than a factual message:
H1a: A narrative message has a more positive effect on a person’s attitude towards separating plastic waste and the intention to separate plastic waste than a factual message.
With the persistent digitalization of our daily lives, it is very important to understand how online behavior influences persuasive outcomes, so that organizations can enable people to change their behavior in a pro-environmental way. It is very important to reach the largest audience possible, so that more people can change their behavior (Beverland, Dobele, & Farrelly, 2015). One of the main tools for reaching such an audience is through intended message engagement; the liking, sharing or commenting on an online message (Kabadayi and Price, 2014). The more people like, share or comment on a message, the more befriended Facebook users of those people can also see the message, and the more may also engage in it (Beverland et al., 2015).
But what are the reasons someone would engage in an online message? Yuki (2015) found that the most significant predictor of sharing a message online was involving a storytelling factor in the message. People could be attracted to a storyline, or think the central idea of a message is important, and want other people to see it too. In the text above, it is already explained that narratives could evoke emotional
responses (e.g. Escalas, 2004). According to several studies, emotions can cause high levels of message engagement among consumers (Escalas, 1998; Sashi, 2012; Eckler & Bolls, 2011). Also, other studies have found that online Facebook posts that trigger emotional responses cause the highest levels of message engagement (Dobele et al., 2007; Eckler & Bolls, 2011). Therefore, it is likely to expect that a narrative message
will cause a higher level of engagement among consumers than a factual message. The following hypothesis is conducted:
H1b: A narrative message will have a stronger effect on intended message engagement than a factual message.
Topic involvement
The level of involvement someone has with a certain topic is also a factor that has to be kept in mind when studying different types of message frames. The level of involvement is a person's perceived relevance of a topic based on inherent needs, values, and interests (Zaichkowsky, 1985). Rothschild (1984) describes
environmental involvement as the degree of personal relevance and importance associated with the environment. Some people might be more interested in certain topics than others, because they have more affiliation with it or they have to deal with that topic more than the other. People that tend to be more surrounded by nature may be more involved with the preservation of the environment than someone that for instance, has lived his/her whole life in the city (Rothschild, 1984). Vermeir and Verbeke (2006) studied that a high level of involvement towards environmental sustainability increased the attitude towards buying green products and the intention to buy such products in comparison with a low level of involvement. Thus, the following hypothesis was conducted:
H2: People with a high level of involvement with the topic sustainability will have a more positive attitude towards separating plastic waste and the intention to separate plastic waste than people with a low level of involvement.
Message type and topic involvement
While a narrative message has a stronger effect on consumers’ attitudes and intentions (e.g. Murphy et al., 2013; van Laer et al., 2013), and a higher level of involvement with the topic sustainability could also result in more positive attitudes towards behaving in a sustainable way and a higher intention to act in a sustainable way (e.g. Vermeir & Verbeke, 2016), it is likely to expect an interaction effect between the message type and the level of involvement.
Researchers found that when highly involved people get aroused, they have increased levels of information processing (Kroeber-Riel, 1979), and that therefore
factual information may be more persuasive compared to narrative information, while people with low levels of information processing might rely more on narrative
information than on factual information. This can be explained with the Elaboration Likelihood Model (ELM) of Petty and Cacioppo (1986). According to this model, information can be processed through a central route or a peripheral route. Whether information is processed through the central or peripheral route depends on the ability and motivation to process the given information. If the ability and motivation to process the information are high because of a high level of involvement, information will be processed through the central route. Contrary to this, information will be processed through the peripheral route when the level of involvement is low, because there is a lack of ability or motivation to process the information. When information is processed through the central route, the person that is exposed to the message will be primarily influenced by rational and factual information like statistics or the strength of the arguments given in the message. When the processing of information goes through the peripheral route, people depend more on other factors, such as emotional appeals.
While factual messages consist of statistical, factual information, narratives rely more on heuristic cues like emotions (Wentzel et al., 2010). Narratives can evoke emotional responses (e.g. Escalas, 2004; Kopfman et al., 1998). People that are less involved will be persuaded more by emotionally driven messages than people that are highly involved (Braverman, 2008). Also, Kronrod, Grinstein and Whatieu (2012) and Cervellon (2013) found that people that find environmental preservation very important will have a more positive feeling towards a message or action that is pro preserving the environment than people that find this topic less important. Therefore, it is plausible to predict that less involved people that are exposed to a narrative message will be more affected by the message than people that are exposed to the factual message. That is why the following is predicted:
H3a: A highly involved person will have a more positive attitude and stronger
intention when exposed to a factual message, while a lower involved person will have a more positive attitude and stronger intention after exposure to a narrative message.
When it comes to the effects of message type on intended message
has a high interest for the topic environmental sustainability, or attributes high value towards it, and is thus highly involved, s/he would have a stronger intention to like, share or comment on a message when exposed to a factual message, according to the ELM of Petty and Cacioppo (1986). On the other hand, when there is a lack of such interest for the topic environmental sustainability, and the person is therefore less involved than someone that attributes high value towards the topic, a narrative message would have a stronger effect on the intention to like, share or comment on a pro-environmental message. That is why the following is expected:
H3b: A narrative will cause a higher level of message engagement when people have a low level of involvement, while highly involved people will have a higher intended message engagement after exposure to a factual message.
Message engagement and attitude, intention
Intended message engagement is an important means to get people to like and spread the content organizations are creating, but how does this intended message engagement relate to the attitude towards pro-environmental behavior and the intention to behave that way? Little is known about these influences, and the present knowledge is somewhat contradicting.In terms of behavioral intentions, there are several studies that underline the effect of contribution ethic: the sense that
individuals have already “done their bit” or feeling as though enough has been done to address a problem (Thøgersen, 1999; Thøgersen & Crompton, 2009; Tiefenbeck, Staake, Roth, & Sachs, 2013). According to this idea, it could be that people think that an act like liking or sharing a message is already enough, resulting in behavior that is contradictory to the target behavior that is presented in the message when a person likes or shares a message. So, if a person likes or shares a message about for instance separating plastic waste, that person might not separate its plastic waste afterwards because s/he has the feeling that s/he has already done their part to help (e.g. Thøgersen, 1999). In contrary to this, Lauren, Smith, Louis, and Dean (2017) found that a high sense of contribution ethics actually causes higher intentions to act in line with the prior behavior, so a person that likes or shares a message concerning separating plastic waste would indeed also have a higher intention to separate waste after liking or sharing the message, according to Lauren et al. (2017). Thus, different researchers have found conflicting results.
Continuing on the positive effects Lauren et al. (2017) have found, also cognitive dissonance could play a role in this process (Festinger, 1962). When people have contradicting ideas and behaviors, they are in a state of cognitive dissonance. For instance, when people know that separating their plastic waste is bad for the environment, but they do not think positively about separating plastic waste and do no intend to separate their plastic waste, their knowledge or attitude and their actual behavior are contradicting, which causes dissonance. In essence, people are trying to find harmony between attitude and behavior, and to reduce the cognitive dissonance (Festinger, 1962). Therefore, it is predicted that prior behavior should be in line with the attitude towards intended behavior and the intention to behave in such a way:
H4: A high level of message engagement will cause a more positive attitude towards separating plastic waste and a higher intention to separate plastic waste than a low level of engagement.
The hypotheses above lead to the following conceptual model:
Figure 1. Conceptual model
Method
Participants and Design
The research question will be answered by conducting an online experimental study with single factor between-subjects design including two levels (Message type: narrative vs. factual) and a continuous predictor (level of involvement) among 234
participants living in the Netherlands. These participants had to speak Dutch and be at least 18 years old. The recruitment of participants took place by asking participants to participate through social media messaging and email. Participants were randomly assigned to one of the two experimental conditions.
Participants who did not agree with the informed consent (n = 1), did not finish the experiment (n = 20), did not have a personal Facebook account (n = 3), were exposed too briefly to the manipulation (n = 72) or did not meet a combination of the requirements were excluded, leaving 162 participants that were included in the analyses. Of these participants, 82 were exposed to the factual message and 80 were exposed to the narrative message. None of the participants knew the exact goal of the study.
The sample of participants consisted of 43,2% male and 56,8% female participants, between the age of 19 and 74 with a mean age of 26,84 (SD = 8.03). Ninety-two percent lived in the city, 1,9% in the suburbs and 6,2% lived in the countryside. Because 51,9% has finished or is currently doing a Master’s degree, 25,3% bachelor’s degree, 17,3% university of applied sciences, 1,9% vocational education and 3,1% higher general education, the sample of participants was relatively highly educated. On the question about how much days per week the participants used Facebook, the average score lied between daily and a few times per week (M = 5.51, SD = 1.02).
Procedure
The participants were asked to participate in the experiment through email or social media messages. When starting the experiment, participants would first see a factsheet in which the procedure and goal of the experiment were explained.
Participants were told that the experiment was about their social media usage and personal interests as a coverstory that was used to try to cover the actual goal of the study. This way, participants’ answers would not be influenced by knowledge about the goal of the study. After this, participants were shown the informed consent form, which they had to sign before being able to participate. First, participants were asked several general questions about their age, gender, level of education and also about their social media usage to remain in line with the cover story. Following this, participants had to answer to what extent they were involved with environmental sustainability through four items. These items were embedded in other questions
about social issues like gay marriage and gender inequality. Examples of these questions are my actions impact the environment and I think women get fewer opportunities than men in our society. The four items were embedded in the other questions as part of the coverstory.
After the social issues questionnaire that included the involvement items, participants were instructed that they would see a Facebook post of the Ocean Cleanup that they had to read carefully. Participants were also told what this
organization does to try to preserve the environment. They were then exposed to one of the two (narrative or factual) messages. To check whether participants had skipped the experimental manipulation, the time spent on the page with the manipulated message was measured. On the page following the post by the Ocean cleanup, containing either the factual or narrative message, participants were asked to answer several questions regarding message engagement, attitude towards separating plastic waste and the intention to separate plastic waste from other waste (in this order). Hereafter, participants were asked about their perceived goal of the study and a manipulation check was conducted. At the last page, a debriefing and a small space where participants could leave comments on the experiment were shown.
Manipulation
Participants were either exposed to a narrative message, or a factual message. The message format was a Facebook post by the Ocean Cleanup, an organization with the main goal of cleaning the ocean from plastic waste. Both messages were kept the same as much as possible, in terms of length and meaning. The narrative message was written from the perspective of Robin (34), a citizen that has been separating plastic waste for years, and was 50 words longer than the factual message. The factual message was written as if the Ocean Cleanup summed up several facts about separating plastic waste.
The messages took into consideration findings from prior research on the effectiveness of narrative messages. According to De Graaf et al. (2016) and Winterbottom, Bekker, Connor, and Mooney (2008), a first-person perspective is likely to be more persuasive than a third-person perspective. This is why the main character in the narrative message, Robin, speaks from a first-person perspective. No particular person is mentioned in the factual text. De Graaf et al. (2016) found that the persuasive effects could be more positive if the message could evoke emotional
responses and the main character in the narrative performed the target behavior. In the narrative text, Robin talks about certain emotions that occur while thinking about plastic pollution, and about the fact that s/he has been separating plastic waste for years. In the factual text, these emotions are not mentioned because the text is not from the perspective of a particular person. Also, the target behavior should be
performed by the character in the narrative (De Graaf et al., 2016). In this case, Robin tells about how he always separates plastic waste from the rest. Similarity between the participant and the character in the narrative was not shown to play a crucial role in the effectiveness of narratives, but dissimilarity was not studied. Therefore to avoid feelings of dissimilarity, a gender-neutral name was chosen for the character (Robin). Both texts can be found in Appendix A.
Measures
At the start of the experiment, the control variables age (open), gender (man, woman, different), demographic situation (city, suburbs, countryside), highest finished or current level of education (None, Primary education, Pre-vocational education, Higher general education, Preparatory scholarly education, Vocational education, University of applied sciences, Research bachelor’s degree, Research master’s degree) and whether they had a Facebook account (yes, no) were measured. Also, to substantiate the cover story, some questions were asked about the frequency with which Facebook users actually used Facebook, and how much they liked, commented on or shared Facebook posts from friends or organizations.
Involvement. The independent variable involvement with environmental sustainability was measured by presenting participants with a list of four statements (Schuhwerk & Lefkoff-Hagius, 1995), in which the participants had to declare
through a 7-point Likert scale whether they totally agreed, or totally disagreed: ‘’I am concerned about the environment’’; ‘’the condition of the environment affects the quality of my life’’; ‘’I am willing to make sacrifices to protect the environment’’; and ‘’my actions impact on the environment’’. This scale was not reliable with a
Cronbach’s α of .32, but was reliable after deleting the item ‘’I am concerned about the environment’’ (Cronbach’s α = .73, M = 5.19, SD = 1.04). After the scale was computed, the sample was divided in two groups, high involvement and low involvement, on the basis of the median level of involvement (Median = 5.33). 70 participants scored below 5.33 and were placed in the group lower involvement, while
92 participants scored 5.33 or higher. The latter were placed in the group higher involvement.
Message engagement. Also message engagement was measured using a 7-point Likert scale, with answers ranging from totally disagree to totally agree. First, participants were asked to imagine seeing the Facebook post when logged on to their own Facebook account. Then, three questions were asked about their willingness to engage with the message: ‘’if this message would be on my timeline, I would like / comment on / share this message through my own Facebook profile’’. Before the variable was constructed, a reliability analysis was conducted, showing that the items would form a reliable scale with a Cronbach’s α of .77 (M = 2.73, SD = 1.28).
Attitude towards separating plastic waste. The attitude was measured based on six items used by Ajzen (2006), on which the participants had to answer through a 7-point bipolar Likert scale whether they thought separating plastic waste was:
unpleasant – pleasant, bad – good, worthless – valuable, uncomfortable –
comfortable, harmful – pleasant, unwise – wise. The Cronbach’s α of all six items
together was .90. That is why the construct variable Attitude was formed from these six items (Cronbach’s α = .90, M = 5.78, SD = 1.13).
Intention to separate plastic waste. Intention to separate plastic waste was also based on prior research by Ajzen (2006). Three items, answered through a 7-point Likert scale ranging from highly unlikely to highly likely together created the construct variable Intention. The three items were: ‘’I intend to separate my plastic waste the coming month’’; ‘’I want to separate my plastic waste the coming month’’; and ‘’I expect to separate my plastic waste coming month’’. The scale was very reliable (Cronbach’s α = .94; M = 4.62, SD = 1.80) and was therefore constructed. Manipulation check and goal of the study
To check whether the manipulation worked as intended, participants had to answer to what extent they totally disagreed, or totally agreed with two statements: ‘’I found the text in the message of the Ocean Cleanup lively’’; and ‘’I found the text in the message of the Ocean Cleanup factual’’. These items were based on two out of three items out of prior research by Das, Kerkhof and Kuiper (2008) and were answered by using a 7-point Likert scale. The third item measured the level of abstractness, which is not relevant in this study. One of the two items was recoded because they were measured in opposite direction. A correlation analysis was run to
test whether the variables Liveliness of the text and Factuality of the text correlated with each other. The variables correlated significantly, r = .31, p < .001. The variable Factuality of the text was recoded so the scale variable Liveliness could be
constructed (M = 5.40, SD = .05).
After this, participants were asked to tell what they thought the goal of the study was, to check for demand characteristics.
Analysis plan
To test the predicted hypotheses for attitude and intention, two PROCESS (model 8) analyses with 5000 bootstrap samples were conducted. In the first analysis, attitude was included as dependent variable. In the second analysis, intention was included as dependent variable. In both analyses, message type was included as independent variable, involvement as moderator variable and message engagement as mediator variable. For the analyses with Attitude as dependent variable, gender was also included as covariate. For the analyses with intention as dependent variable, gender and demographic situation were included as covariates. The full results of the PROCESS regression analyses are shown in Table 2.
Results
Randomization check
First, a randomization check was done to test if the participants in the different condition would differ from each other in terms of gender, age, educational level and demographic situation. A chi-squared analysis was conducted to test the
randomization of gender and demographic situation. People from different genders were equally distributed between both experimental conditions (Pearson’s χ2 (1) = 0.60, p = .440), as well as people from different demographic situations (Pearson’s χ2 (2) = 2.98, p = .225). To analyze whether age and educational level were equally distributed between both experimental conditions, a One-Way ANOVA was
conducted. This showed that age was equally distributed between both experimental conditions, F (1, 160) = 3.76, p = .054, as well as educational level, F (1, 160) = 0.01, p = .911. The groups did not differ with regard to gender, age, educational level and demographic situation.
Correlation analysis
Correlation analyses were conducted to check whether certain background variables (gender, age, demographic situation and educational level) correlated with the dependent variables message engagement, attitude and intention. If so, they could be included in the analyses as covariates if necessary. Attitude correlated with gender, and intention correlated with gender and demographic situation. These control
variables are included in the analyses testing the hypotheses. The results are shown in Table 1.
Table 1. Correlation analysis of the control variables
Control variables
Gender Age Demographic
situation
Educational level
Attitude .32** .05 .15 .12
Intention .20** .12 .27** .11
** Correlation is significant at the .01 level
Manipulation check
To check if participants in the factual group indeed perceived the factual text as more factual than the narrative group, an independent samples t-test was
conducted. The t-test was conducted with message type (factual vs narrative) as independent variable, and Liveliness as dependent variable. There was a significant difference between message types on perceived Liveliness of the message, t(160) = 3.02, p = .003, d = .00, 95% CI [.11, .53]. As expected, the participants in the factual group indeed perceived the factuality of the message (M = 5.78, SD = .86) as more factual than the narrative group (M = 5.33, SD = .86).
Moderated Mediation model
Hypothesis 1a stated that participants in the narrative group would have a more positive attitude and intention towards separating plastic waste than people who saw the factual message, while Hypothesis 1b predicted that a narrative would also have a stronger effect on intended message engagement. Hypothesis 2 predicted that people that have a higher level of involvement also have a more positive attitude towards separating plastic waste and a higher intention to separate plastic waste than
people with a lower level of involvement. In Hypothesis 3a, an interaction effect between Hypothesis 1 and 2 was predicted, stating that a high level of involvement would increase the effect a factual message would have on attitude and intention, and a low level of involvement would increase the effect a narrative would have on attitude and intention. Hypothesis 3b was built on the same idea as Hypothesis 3a, but then with intended message engagement as outcome variable. Hypothesis 4 predicted that the higher the intended message engagement would be, the more positive the attitude and the higher the behavioral intention would be.
Attitude. To test if a narrative message would have a more positive effect on attitude towards separating plastic than a factual message, and whether a high level of involvement would have a more positive effect on attitude than a low level of
involvement, a PROCESS (model 8) with 5000 bootstrap samples was conducted. Also, the interaction effect between message type and involvement was tested. There was no significant relation between message type and attitude, b = -1.30, SE = 84, t = -1.55, p = .655. Also, contrary to expectations, exposure to a narrative message frame did not result in higher levels of intended message engagement, compared to a factual message, b = .47, SE = 1.04, t = 45, p = .655. While it was expected that there would be a relation between the level of involvement and attitude, no significant effect was found, b = -.21, SE = .25, t = -.87, p = .124. The interaction effect was not significant either, b = .24, SE = .16, t = 1.52, p = .130, which means that participants that were highly involved did not have a more positive attitude after exposure to a factual message than less involved participants, and less involved participants did not have a more positive attitude towards separating plastic waste after exposure to a narrative than highly involved participants. Nor was the effect of message type on message engagement moderated by involvement levels, b = -.09, SE = .20, t = -.48, p = .635. In line with expectations the intended message engagement did correlate positively with attitude, b = .21, SE = .07, t = 3.20, p = .002. When levels of intended message engagement increased, attitude increased as well, and vice versa.
Intention. The same PROCESS analysis showed that a narrative did not have a stronger effect on intention than a factual message, b = -.41, SE = 1.33, t = -.31, p = .757, nor did a narrative have a stronger effect on intended message engagement than a factual message, b = .37, SE = 1.04, t = 36, p = .721. No interaction effect between message type and the level of involvement on intention was found either, b = .13, SE = .25, t = .51, p = .609. Also, involvement and message type did not significantly
interact with each other while aiming at intended message engagement, b = -.07, SE = .20, t = -.37, p = .710. Participants that were highly involved did not have a higher intention to separate plastic waste, nor a higher intended message engagement when exposed to a factual message than less involved participants, while the same effect did not occur for participants that were exposed to the narrative message, compared to the participants that were highly involved. However, the intended message engagement and intention to separate plastic waste did significantly correlate, b = .32, SE = .10, t = 3.16, p = .002.
No significant relations were found between message type and attitude, intention and intended message engagement. A narrative message did therefore not effectuate a stronger effect on attitude, intention or intended message engagement, so Hypothesis 1a and 1b are rejected. Also, no effect was found for the level of involvement on attitude and intention, so a higher level of involvement did not have a stronger effect on attitude and intention than a lower level of involvement, therefore rejecting Hypothesis 2.
The interactions between message type and level of involvement did also not cause significant relations; participants that were highly involved did not have a more positive attitude and a higher intention and intended message engagement than less involved participants after exposure to a factual message, while participants that were less involved did not have a more positive attitude and a higher intention and intended message engagement than highly involved participants after exposure to a narrative message. Therefore, Hypotheses 3a and 3b were rejected. On the other hand, significant relations between intended message engagement and attitude and intention were found. When intended message engagement increased, the attitude towards separating plastic became more positive and the intention to separate plastic waste got stronger, and vice versa. Therefore, Hypothesis 4 was accepted.
Table 2. Regression results of the moderated mediation model of message type on attitude and intention through message engagement, with involvement as moderating variable in the correlation between message type and message engagement, attitude and intention.
Predictor b SE t p BC 5.000 BOOT
LL95 UL95
Attitude Message engagement (mediator)
Constant 1.73 1.62 1.07 .286 -1.46 4.93
Message type 0.47 1.04 0.45 .655 -1.59 2.52
Involvement 0.16 0.30 0.52 .602 -0.44 0.75
Message type x -0.09 0.20 -0.48 .635 -0.48 0.30
Involvement
Intention Message engagement (mediator)
Constant 1.42 1.62 0.88 .383 -1.78 4.62 Message type 0.37 1.04 0.36 .721 -1.68 2.42 Involvement 0.14 0.30 0.45 .650 -0.46 0.73 Message type x -0.07 0.20 -0.37 .710 -0.46 0.31 Involvement Attitude Constant 5.42 1.32 4.12 .000 2.82 8.01 Message engagement 0.21 0.07 3.20 .000 0.08 0.33 Message type -1.30 0.84 -1.55 .124 -2.97 0.36 Involvement -0.21 0.25 -0.87 .384 -0.70 0.27 Message type x 0.24 0.16 1.52 .130 -0.07 0.56 Involvement Intention Constant 0.70 2.08 0.34 .738 -3.41 4.80 Message engagement 0.32 0.10 3.16 .002 0.12 0.53 Message type -0.41 1.33 -0.31 .757 -3.03 2.21 Involvement 0.12 0.39 0.31 .759 -0.64 0.88 Message type x 0.13 0.25 0.51 .609 -0.37 0.62 Involvement
Conditional direct effects of message frame on attitude at values of involvement
Involvement Effect SE t p LL95 UL95
5.19 (medium) -0.04 0.17 -0.27 .789 -0.37 0.28
6.23 (high) 0.21 0.23 0.89 .374 -0.25 0.67
Conditional direct effects of message frame on intention at values of involvement
Involvement Effect SE t p LL95 UL95
4.14 (low) 0.12 0.37 0.33 .741 -0.61 0.85
5.19 (medium) 0.26 0.26 0.99 .326 -0.26 0.77
6.23 (high) 0.39 0.37 1.06 .292 -0.34 1.12
Conditional indirect effects of message type on attitude when message engagement is low, medium or high
Message involvement Bootstrap indirect effect Bootstrap SE BootLLCI BootULCI
4.14 (low) 0.02 0.06 -0.09 0.15
5.19 (medium) -0.00 0.04 -0.09 0.08
6.23 (high) -0.02 0.06 -0.17 0.09
Conditional indirect effects of message type on intention when message engagement is low, medium or high
Message involvement Bootstrap indirect effect Bootstrap SE BootLLCI BootULCI
4.14 (low) 0.02 0.10 -0.17 0.22
5.19 (medium) -0.00 0.07 -0.14 0.14
6.23 (high) -0.03 0.10 -0.24 0.16
Note. N = 162. Bootstrap sample size = 5.000
Conclusion & Discussion
The goal of this study was to examine if a narrative message would have a stronger effect on attitude towards separating plastic waste, the intention to separate plastic waste and the intention to like, share or comment on a message, and if a higher level of involvement would effectuate a more positive attitude and a stronger intention than a lower level of involvement. Also, the interaction effects between these two factors were studied. It was hypothesized that people that had a higher level of intention would have a more positive attitude, a stronger intention and a higher level of
intended message engagement when exposed to a factual message, while people who were less involved would have a more positive attitude, stronger intention and higher
level of intended message engagement when exposed to a narrative message. It was also predicted that a higher level of intended message engagement would cause a more positive attitude towards separating plastic waste and a stronger intention to separate plastic waste. The author tried to answer these questions by means of an experiment. Participants saw either one of two online Facebook messages ostensibly posted by the Ocean Cleanup. One was written in a narrative format, the other one as a factual text.
Hypothesis 1a stated that a narrative message would have a more positive effect on attitude towards separating plastic waste and intention to separate plastic waste than a factual message (e.g. Braddock & Dillard, 2016), because they can reduce the reactance that people experience when being exposed by a persuasive message (Moyer-Gusée, 2008). A narrative message did not have a stronger effect on attitude and intention, so this hypothesis had to be rejected. Also, according to
Hypothesis 1b, a narrative should have a stronger effect on the intended message engagement, because narratives can evoke emotions (e.g. Escalas, 2004), and people tend to like, share or comment on emotional content a lot more than content that does not trigger emotional responses (e.g. Dobele et al., 2007). Unfortunately, this effect was not found in this study, so Hypothesis 1b had to be rejected.
Hypothesis 2 stated that people with a higher level of involvement would have a more positive attitude and intention than people with a lower level of involvement, because people with high levels of involvement find it more important to think and act pro-environmentally than people who have a lower level of involvement (Vermeir & Verbeke, 2006). This effect was not found for attitude, nor for intention, so
Hypothesis 2 had to be rejected.
According to the ELM of Petty and Cacioppo (1986), people that are highly involved should have a more positive attitude, stronger intention and higher level of intended message engagement when exposed to a factual message. On the other hand, people that are less involved should have a more positive attitude, stronger intention and higher level of intended message engagement when exposed to a narrative message (e.g. Braverman, 2008). Unfortunately, neither of these effect occurred, so Hypotheses 3a and 3b had to be rejected.
Lastly, it was predicted that a higher level of message engagement would effectuate a more positive attitude and a stronger intention, because people tend to reduce their feeling of cognitive dissonance (Festinger, 1962). Indeed, there was a
significant positive relation between the intended message engagement and the attitude and intention. When the intended message engagement increased, the attitude and intention would also increase, and vice versa. Therefore, Hypothesis 4 was accepted. However, no pronunciations could be done about the causality of the effect because the effect was measured by means of a regression analysis.
It seems that a narrative does not have a stronger effect on attitude and
intention, even when mediated by the level of intended message engagement or when the level of involvement moderates the direct and mediation effects. Several prior studies already concluded that narrative and factual messages might not differ that significantly in terms of persuasive outcomes like attitude and intention or might only differ slightly from each other (Zebregs, et al., 2015; Zebregs, et al., 2015), so maybe narratives are not as much of an important topic as predicted before. The narrative text was based on findings of prior studies, so it is hard to conclude whether the narrative text was or was not effective enough or whether participants are just not affected more by narratives than by factual messages. However, it would be possible to add more emotional content in the narrative, because in this study the level of emotional coverage might have been somewhat low. This was because it was hard to implement the same amount of emotions in the factual text written from the
perspective of the Ocean Cleanup, while the narrative was written from a person’s perspective. If it were possible to, for instance, increase the length of the narrative and add a story and images of plastic-covered beaches and dead animals with bodies full of plastic, the level of emotion would be a lot higher, which could positively influence the effect narratives could have on attitude, intention and intended message
engagement (Van Laer et al., 2013).
The level of involvement did also not have a significant effect on attitude and intention, nor did it interact significantly with the type of message. The fact that there was no significant interaction effect on attitude, intention and intended message engagement could be because the level of involvement concerning environmental sustainability was very high among this group of participants. When looking at the ELM of Petty and Cacioppo (1986), people that are highly involved should be more affected by a factual message, while people that are less involved should be more affected by a narrative message. In this study, the least involved person scored already very positive. To divide the participants in two groups, a median split was conducted but the median was already so high that there was actually no big difference between
the group that had a high level of involvement and the group that had a low level of involvement. This could explain the fact that there were no significant differences between these groups. Finding a sample of participants that are not so homogeneous in terms of level of involvement could increase the difference between samples in future research.
Some general limitations could also be argued. Because the experiment took place online, people were not checked while participating. Participants may have been distracted, unmotivated, or under time pressure, which could have influenced the outcomes negatively. The decrease of reactance only takes place when people can fully focus on the narrative (Moyer-Gusée, 2008). When distracted, that major advantage for narratives decreases. Also, a study by Van den Hende, Dahl,
Schoormans and Snelders (2012) underlines the importance of instructing participants to actively imagine being in the key actor’s position before reading the narrative text, so that the feeling of similarity between reader and actor could be increased. This was not done in the current study, but is an easy improvement to implement in further studies.
Prior research found very contradicting results concerning the influence intended message engagement could have on the attitude towards separating plastic waste and the intention to separate plastic waste. No prior research on this topic was conducted, so hypotheses were based on other theories like the cognitive dissonance. If participants’ intention to like, share or comment on an online Facebook message increased, they wanted to feel and act accordingly. A reason for this could be to reduce the feeling of cognitive dissonance (Festinger, 1962). These findings
contradict the theory of contribution ethics, which states that the conscious of people is satisfied when someone has liked, shared or commented on a message, therefore feeling no need to behave according to the shown behavior anymore.
This is a very useful finding, because this could mean that organizations like the Ocean Cleanup or other organizations that try to make people behave in a pro-environmental way can persuade people into doing so by creating an online message that is liked, shared or commented on many times. The only thing that has yet to be discovered is how to increase the intended message engagement. The use of a
narrative instead of a factual message did not increase intended message engagement, neither when involvement moderated the effect. Beverland et al. (2015) found that people could have a stronger intention to like, share or comment on a message when
they recognize a part of their identity in the message (e.g. nationality, race, culture). It could be interesting to examine whether a factor like identity could influence the level of intended message engagement.
What is also an important factor to point out concerning message engagement is that the level of message engagement was measured as a form of intention in this study. Normally, people like, share and comment on online Facebook messages every day. It is therefore interesting to conduct an experiment while using a real Facebook site where participants have to behave online like they would normally. Also, the need for reducing cognitive dissonance could increase because in that case the participants actually conducted the behavior – liking, sharing or commenting on an online
message, instead of only answering if they might like, share or comment (Festinger, 1962). Maybe people do not even think about sharing a message when they are online, but only responded they would because they were asked about it now. This also could have led to socially desirable answers, because much of the participants knew the researcher directly and more than half of the participants finished or is currently finishing their Master’s degree, which could mean that more than half of the participants has even a slight understanding of answering questions in a socially desirable way. Therefore, the actual behavior concerning message engagement should be measured instead of intended message engagement.
Because the desired behavior – separating plastic waste - is related to sustainable, altruistic behavior, the significant results concerning message
engagement in this study could also be used for optimizing campaigns that want to effectuate a behavior change in terms of recycling, decreasing energy usage at home, eating less meat and many other types of behavior that are pro-environmental, but harder to realize than the undesired behavior.
Although not many effects have been found, the current study has contributed to the available scientific knowledge by demonstrating that narrative and factual messages do not always differ in terms of the persuasive outcomes that could be caused by exposure to them, nor when the level of involvement is included as moderator variable. On the other hand, it is very important that the current study contributed to a field of research wherein almost no scientific research has been conducted yet. In an era of digitalization, likes and shares on social media like
Facebook it is of utmost importance to examine the effects that online behavior could have on attitudes and behavioral intentions. To further examine the effects that online
behavior like liking, sharing or commenting on online Facebook messages could have on these persuasive outcomes, future research should focus on the actual behavior instead of the behavioral intention and the identity factor, because identity might be a relevant predictor of the level of message engagement.
References
Adaval, R., & Wyer, R. S. (1998). The role of narratives in consumer information processing. Journal of Consumer Psychology, 7(3), 207–245. doi:
https://doi.org/10.1207/s15327663jcp0703_01
Ajzen, I. (2006). Constructing a TpB questionnaire: Conceptual and methodological considerations. Downloaded from
https://people.umass.edu/aizen/pdf/tpb.measurement.pdf
Beverland, M., Dobele, A., & Farrelly, F. (2015). The viral marketing metaphor explored through Vegemite. Marketing Intelligence & Planning,33(5), 656– 674. doi: https://doi.org/10.1108/MIP-08-2014-0146
Braddock, K., & Dillard, J. P. (2016). Meta-analytic evidence for the persuasive effect of narratives on beliefs, attitudes, intentions, and behaviors. Communication Monographs, 83(4), 446–467. doi:
https://doi.org/10.1080/03637751.2015.1128555
Braverman, J. (2008). Testimonials versus informational persuasive messages: The moderating effect of delivery mode and personal involvement.
Communication Research, 35(5), 666–694. doi: https://doi.org/10.1177/0093650208321785
Cervellon, M. C. (2012). Victoria's dirty secrets: Effectiveness of green not-for-profit messages targeting brands. Journal of Advertising, 41(4), 133–145. doi: https://doi.org/10.1080/00913367.2012.10672462
Das, E., Kerkhof, P., & Kuiper, J. (2008). Improving the effectiveness of fundraising messages: The impact of charity goal attainment, message framing, and evidence on persuasion. Journal of Applied Communication Research, 36(2), 161–175. doi: https://doi.org/10.1080/00909880801922854
De Graaf, A. D., Sanders, J. M., & Hoeken, H. (2016). Characteristics of narrative interventions and health effects: A review of the content, form, and context of narratives in health-related narrative persuasion research. Review of
Communication Research, 4, 88–131. doi: 10.12840
Derraik, J. G. (2002). The pollution of the marine environment by plastic debris: a review. Marine Pollution Bulletin, 44(9), 842–852. doi:
https://doi.org/10.1016/S0025-326X(02)00220-5
Dobele, A., Lindgreen, A., Beverland, M., Vanhamme, J., & Van Wijk, R. (2007). Why pass on viral messages? Because they connect emotionally. Business Horizons, 50(4), 291–304. doi: https://doi.org/10.1016/j.bushor.2007.01.004
Eckler, P., & Bolls, P. (2011). Spreading the virus: Emotional tone of viral advertising and its effect on forwarding intentions and attitudes. Journal of Interactive Advertising, 11(2), 1–11. doi:
https://doi.org/10.1080/15252019.2011.10722180
Eco-Cycle (2016). Be straw free campaign. Retrieved from http://www.ecocycle.org/bestrawfree/
Escalas, J. E. (1998). Advertising narratives: What are they and how do they work. RepresentingCconsumers: Voices, Views, and Visions, 267–289.
Escalas, J. E. (2004). Imagine yourself in the product: Mental simulation, narrative transportation, and persuasion. Journal of Advertising, 33(2), 37–48. doi: https://doi.org/10.1080/00913367.2004.10639163
European Environmental Agency (2017). Recycling rates in Europe. Retrieved from
https://www.eea.europa.eu/about-us/competitions/waste-smart-competition/recycling-rates-in-europe/
Festinger, L. (1962). Cognitive dissonance. Scientific American, 207(4), 93–106.
Fulgoni, G. M. (2015). How brands using social media ignite marketing and drive growth: Measurement of paid social media appears solid but are the metrics for organic social overstated? Journal of Advertising Research, 55(3), 232– 236. doi: 10.2501/JAR-2015-004
Gourmelon, G. (2015, January 27). Global plastic production rises, recycling lags. Retrieved from http://www.worldwatch.org
Green, M. C., & Brock, T. C. (2000). The role of transportation in the persuasiveness of public narratives. Journal of Personality and Social Psychology, 79(5), 701–721. doi: http://dx.doi.org/10.1037/0022-3514.79.5.701
Jansen, N. (2014). Plastic cijfers. Retrieved from https://www.kijkmagazine.nl/artikel/plastic/
Kabadayi, S., & Price, K. (2014). Consumer–brand engagement on Facebook: Liking and commenting behaviors. Journal of Research in Interactive Marketing, 8(3), 203–223. doi: https://doi.org/10.1108/JRIM-12-2013-0081
Kollmuss, A., & Agyeman, J. (2002). Mind the gap: why do people act
environmentally and what are the barriers to pro-environmental behavior?. Environmental Education Research, 8(3), 239–260. doi:
https://doi.org/10.1080/13504620220145401
Kopfman, J. E., Smith, S. W., Ah Yun, J. K., & Hodges, A. (1998). Affective and cognitive reactions to narrative versus statistical evidence organ donation messages. Journal of Applied Communication Research, 26, 279–300. doi: 10.1080/00909889809365508
Kroeber-Riel, W. (1979). Activation research: Psychobiological approaches in consumer research. Journal of Consumer Research, 5(4), 240–250. doi: https://doi.org/10.1086/208736
Kronrod, A., Grinstein, A., & Wathieu, L. (2012). Go green! Should environmental messages be so assertive? Journal of Marketing, 76(1), 95–102. doi:
https://doi.org/10.1509/jm.10.0416
Lakoff, G. (2010). Why it matters how we frame the environment. Environmental Communication, 4(1), 70–81. doi:
https://doi.org/10.1080/17524030903529749
Lauren, N., Smith, L. D. G., Louis, W. R., & Dean, A. J. (2017). Promoting spillover: How past behaviors increase environmental intentions by cueing
self-perceptions. Environment and Behavior, 1–24. doi: https://doi.org/10.1177/0013916517740408
Laville, S., & Taylor, M. (2017, June 28). A million bottles a minute: world's plastic binge 'as dangerous as climate change. The Guardian. Retrieved from
https://www.theguardian.com
Mattila, A. S. (2000). The role of narratives in the advertising of experiential services. Journal of Service Research, 3(1), 35–45. doi:
https://doi.org/10.1177/109467050031003
Mazzocco, P. J., Green, M. C., Sasota, J. A., & Jones, N. W. (2010). This story is not for everyone: Transportability and narrative persuasion. Social Psychological and Personality Science, 1, 361–368. doi: 10.1177/1948550610376600
McQueen, A., Kreuter, M. W., Kalesan, B., & Alcaraz, K. I. (2011). Understanding narrative effects: the impact of breast cancer survivor stories on message processing, attitudes, and beliefs among African American women. Health Psychology, 30(6), 674. doi: http://dx.doi.org/10.1037/a0025395
Moyer-Gusé, E. (2008). Toward a theory of entertainment persuasion: Explaining the persuasive effects of entertainment-education messages. Communication Theory, 18, 407–425. doi: https://doi.org/10.1111/j.1468-2885.2008.00328.x
Murphy, S. T., Frank, L. B., Chatterjee, J. S., & Baezconde‐Garbanati, L. (2013). Narrative versus nonnarrative: The role of identification, transportation, and emotion in reducing health disparities. Journal of Communication, 63(1), 116– 137. doi: https://doi.org/10.1111/jcom.12007
Oskamp, S. (2002). Environmentally responsible behavior: Teaching and promoting it effectively. Analyses of Social Issues and Public Policy, 2(1), 173–182. doi: https://doi.org/10.1111/j.1530-2415.2002.00036.x
Petty, R. E., & Cacioppo, J. T. (1986). The elaboration likelihood model of
persuasion. Communication and persuasion (pp. 1-24). Springer New York. doi: https://doi.org/10.1007/978-1-4612-4964-1_1
Rothschild, M. L. (1984). Perspectives on involvement: Current problems and future directions. Advances in Consumer Research, 1, 216–217.
Sashi, C. M. (2012). Customer engagement, buyer-seller relationships, and social media. Management Decision, 50(2), 253–272. doi:
https://doi.org/10.1108/00251741211203551
Schuhwerk, M. E., & Lefkoff-Hagius, R. (1995). Green or non-green? Does type of appeal matter when advertising a green product? Journal of Advertising, 24(2), 45–54. doi: https://doi.org/10.1080/00913367.1995.10673475
Thøgersen, J. (1999). Spillover processes in the development of a sustainable consumption pattern. Journal of Economic Psychology, 20, 53–81. doi: 10.1016/S0167-4870(98)00043-9
Thøgersen, J., & Crompton, T. (2009). Simple and painless? The limitations of spillover in environmental campaigning. Journal of Consumer Policy, 32, 141–163. doi: 10.1007/s10603-009-9101-1
Tiefenbeck, V., Staake, T., Roth, K., & Sachs, O. (2013). For better or for worse? Empirical evidence of moral licensing in a behavioral energy conservation campaign. Energy Policy, 57, 160–171. doi: 10.1016/j.enpol.2013.01.021
Van den Hende, E., Dahl, D., Schoormans, J., & Snelders, D. (2012). Narrative transportation in concept tests for really new products: The moderating effect of reader-protagonist similarity. Journal of Product Innovation Management, 29, 157–170. doi:10.1111/j.1540-5885.2012.00961
Van Laer, T., De Ruyter, K., Visconti, L. M., & Wetzels, M. (2013). The extended transportation-imagery model: A meta-analysis of the antecedents and consequences of consumers' narrative transportation. Journal of Consumer Research, 40(5), 797–817. doi: https://doi.org/10.1086/673383
Vermeir, I., & Verbeke, W. (2006). Sustainable food consumption: Exploring the consumer “attitude–behavioral intention” gap. Journal of Agricultural and Environmental Ethics, 19(2), 169–194. doi: https://doi.org/10.1007/s10806-005-5485-3
Wentzel, D., Tomczak, T., & Herrmann, A. (2010). The moderating effect of
manipulative intent and cognitive resources on the evaluation of narrative ads. Psychology & Marketing, 27(5), 510–530. doi:
https://doi.org/10.1002/mar.20341
Wirtz, B. W., Schilke, O., & Ullrich, S. (2010). Strategic development of business models: implications of the Web 2.0 for creating value on the Internet. Long range planning, 43(2-3), 272–290. doi:
Yuki, T. (2015). What makes brands' social content shareable on facebook?: An analysis that demonstrates the power of online trust and attention. Journal of Advertising Research, 55(4), 458–470. doi: 10.2501/JAR-2015-026
Zaichkowsky, J. (1985). Measuring the involvement construct. Journal of Consumer Research, 12, 341–352. doi: https://doi.org/10.1086/208520
Zebregs, S., van den Putte, B., Neijens, P., & de Graaf, A. (2015). The differential impact of statistical and narrative evidence on beliefs, attitude, and intention: A meta-analysis. Health Communication, 30(3), 282–289. doi:
https://doi.org/10.1080/10410236.2013.842528
Zebregs, S., van den Putte, B., de Graaf, A., Lammers, J., & Neijens, P. (2015). The effects of narrative versus non-narrative information in school health
education about alcohol drinking for low educated adolescents. BMC Public Health, 15(1), 1085. doi: https://doi.org/10.1186/s12889-015-2425-7
Appendix 1 – Factsheet, Informed Consent, Debriefing
1.1 Factsheet Beste participant,
Bedankt voor je deelname aan dit onderzoek. Voor dit onderzoek is het noodzakelijk dat je een persoonlijk Facebook account hebt. Is dit niet het geval, dan kun je niet deelnemen aan dit onderzoek. Het onderzoek gaat over persoonlijke interesses en social media. In het onderzoek worden je een aantal vragen gesteld, onder andere over jouw social media gebruik en interesses en krijg je een Facebook post te zien. Ook hierover zullen enkele vragen gesteld worden. Het onderzoek zal ongeveer 5 minuten in beslag nemen.
De antwoorden zullen te allen tijde anoniem blijven en enkel gebruikt worden voor dit onderzoek. Ook zullen zich geen noemenswaardige ongemakken voordoen tijdens of door dit onderzoek. Het staat je vrij te weigeren aan dit onderzoek mee te doen. Ook mag je ten alle tijden stoppen met het onderzoek, zonder opgaaf van redenen. Je kan achteraf, binnen 24 uur, jouw toestemming voor het gebruik van de
onderzoeksgegevens intrekken.
Mochten er naar aanleiding van je deelname aan dit onderzoek klachten of
opmerkingen bij je zijn, dan kun je contact opnemen met de Commissie Ethiek van de afdeling Communicatiewetenschap, per adres: ASCoR secretariaat, Commissie
Ethiek, Universiteit van Amsterdam, Postbus 15793, 1001 NG Amsterdam; 020-525 3680; ascor-secr-fmg@uva.nl. Een vertrouwelijke behandeling van je klacht of opmerking is daarbij gewaarborgd. Hartelijk bedankt voor je deelname, je kunt nu beginnen door naar de volgende pagina te gaan.
Met vriendelijke groet,
Bart Tetteroo
