Abstract
As the consequences of environmental degradation become more severe, the urgency for timely action increases. The youth have the power to change the future if they are well informed. Preadolescents and adolescents today spend a considerable amount of their time consuming media in many forms. Thus, a way to approach them is through media,
specifically with edutainment content. Previous research has shown that narratives that promote higher enjoyment, identification, and behavioural intentions can facilitate effective learning. Moreover, developmentally, children of different ages may show different
preferences. Building on this, the present paper aims to provide insights into the creation of effective edutainment narratives for environmental education targeted at preadolescents and adolescents. In order to assess preadolescents’ and adolescents’ evaluations of different narrative elements (e.g., protagonist qualities, story settings) and environmental concerns, an online survey of 9- to 17-year-olds (N = 196) was conducted. Overall, the qualities of the protagonist (e.g., sex, age, form) did not affect enjoyment, identification, or behavioural intentions to read similar stories. However, when given the choice, children showed a preference for protagonists who are their age or slightly older, and preadolescents preferred same-sex protagonists more often than adolescents did. As for story setting, the results show that preadolescents enjoy psychologically distant stories more than adolescents do. These results are discussed in light of recent literature on the developmental characteristics of youth and their preferences.
Introduction
Rising temperatures and sea levels, floods, droughts, plastic soup, the extinction of animals and plant life, and even pandemics are among the several serious consequences of environmental degradation that the world population experiences today (M. R. Allen et al., 2018; IPCC, 2014). Given the anthropogenic, or human-made, nature of this degradation, these environmental changes can be reversed through the adoption of well-thought action plans (Masson-Delmotte et al., 2018). Scientists predict that taking no action against
environmental degradation will lead to even more intense, negative consequences. Therefore, effectively educating the public about environmental protection is essential.
Children today, particularly in Europe, have an increasing say in household decisions, such as new purchases or holiday plans. They are encouraged to express their opinions and participate in decision making (Gunter & Furnham, 1998; McNeal, 1992; Valkenburg & Cantor, 2001). Accordingly, this behavioural change among children has great potential to bring the environmental discussion to families. Therefore, educating children on
environmental protection contributes to the awareness of their households (Mira, Cameselle, & Martínez, 2003). Preadolescence and early adolescence are suggested to be particularly important developmental stages during which children display a growing interest in global problems (Blanchet-Cohen, 2008; Chawla & Cushing, 2007; Harker-Schuch, 2019). Thus, targeting these age groups with environmental education is essential. However, “media preferences and media effects are highly dependent on children’s development” (Valkenburg & Piotrowski, 2017, p. 273). Therefore, when addressing children, their developmental differences must be taken into account.
Media is an effective instrument to spread information and promote positive
behavioural changes (Jones, Baldwin, & Lewis, 2012), and it has the potential to educate the youth about environmental protection. The youth engage with entertainment media for six to
seven hours daily (Rideout & Robb, 2019). Delivering educational content within an entertainment context, or edutainment, has been shown to be a useful and persuasive
approach to teaching children as well as promoting prosocial behaviours (Brusse, Fransen, & Smit, 2017; Farrar, 2006; Moyer-Gusé, 2008). Given the intensity with which youth consume media and the societal importance of environmental protection, it is important to understand how media can be used to inform the youth effectively.
Existing research has focused on the flow state, which is a “positive psychological state that is challenging, intrinsically rewarding and enjoyable” (Bressler & Bodzin, 2013, p. 506). Sherry (2004), drawing upon Czikszentmihalyi's (1990) definition of flow (see also Czikszentmihalyi & Nakamura, 2014) emphasises that only educational media produced with the consideration of flow elements can foster engagement. A key element of the flow
experience is enjoyment (Czikszentmihalyi, 1990; Weber, Tamborini, Westcott-Baker, & Kantor, 2009). Edutainment content needs to be enjoyable and engaging to be effective. But what makes it enjoyable for the audience? Although the themes, characters, and settings in stories have been studied to understand what children enjoy, a comprehensive investigation of environmental concerns, story elements, and preferences with a focus on children’s developmental characteristics is missing (Lowther & Sullivan, 1993; Shippen, Houchins, Puckett, & Ramsey, 2007; Summers & Lukasevich, 1983).
The present study will examine the differences between preadolescent and adolescent children in three parts regarding (1) how they experience edutainment narratives in which the sex, age, form, and environmental motive of the protagonist and psychological distance in the narrative are systematically manipulated; (2) the strength of their environmental motivations; and (3) their preferences for the setting as well as the sex, age, and form of the protagonist when constructing their ideal narratives. The present study aims to provide a more delicate
comparison between environmental concerns, the effects of narrative elements, and age groups to understand which elements contribute to effective environmental edutainment.
Environmental Concerns and Youth
Environmental concerns are an essential basis in the formation of environmental attitudes (Schultz, 2000, 2001); therefore, understanding these concerns is a vital step for increasing environmental awareness. One way to understand environmental concerns is by examining the motivations behind the concerns because motives precede behaviour (Lee, Neighbors, & Woods, 2007; Schultz, 2000, 2001). Based on the value-belief-norm model by Stern and Dietz (1994), Schultz (2000, 2001) developed the three-dimensional environmental motive scale (EMS). The dimensions are based on the “relative importance that a person places on themselves (egoistic), other people (altruistic), or nature (biospheric)” (Bruni, Chance, & Schultz, 2012, p. 1; Schultz, 2001). Studies with cross-cultural samples of adults tested and confirmed the three-dimensional structure of environmental motives (de Groot & Steg, 2007; Milfont, Duckitt, & Cameron, 2006). An accurate grasp on individuals’
environmental concerns is instrumental in framing environmental messages effectively such that the right appeals to action are used (Nerlich, Koteyko, & Brown, 2010; Schultz & Zelezny, 2003). Bruni et al. (2012) modified and tested the EMS on children from 10 to 18 years old; their results confirmed that children’s motivations behind their environmental concerns are three dimensional, as well (i.e., egoistic, social-altruistic, and biospheric). Environmental Concerns and Developmental Characteristics
This research focuses on youth, and specifically on 9- to 17-year-olds, as this group is not only the generation to face the worst effects of environmental degradation but also the last generation that can take action against them (UNICEF, 2018). They are the consumers and voters of tomorrow, but they still have the power to evoke public attention to global matters now (e.g., Fridays for Future, School Strike for Climate; Alhendawi, 2014). Thus,
understanding the environmental concerns of 9- to 17-year-olds is important for designing educational content for them.
However, 9- to 17-year-olds are not a homogenous group when their developmental characteristics are considered, and it is vital to know how children within this age group differ to effectively address them. For example, preadolescents are more interested in environmental issues compared to adolescents (Kaplan & Kaplan, 2002; Liefländer & Bogner, 2014; Negev, Sagy, Garb, Salzberg, & Tal, 2008; Uitto & Saloranta, 2010). The period when interest in environmental issues decreases during the ages of 15–16 is called “the adolescent dip”; Olsson and Gericke (2016), in their study of Swedish students (divided into three age groups: 12–13, 15–16, and 18-19), found that the adolescent dip disappears at the age of 18–19. Age-appropriate and relevant messages that take the developmental characteristics and interests of the target group into consideration can achieve effective environmental communications (Miller-Day & Hecht, 2013; Strife, 2012). Since the adolescent dip is a temporary period, it is important to address preadolescents and adolescents properly in alignment with their respective developmental differences.
The period from the age of 8 until the age of 12 is referred to as preadolescence (Valkenburg & Cantor, 2001). Children in this age group begin to outgrow of the influence of their parents (Valkenburg & Piotrowski, 2017), and their interest in real-world experiences increases (Acuff, 1997; Gunter, McAleer, & Clifford, 1991; Mielke, 1983). Preadolescence is, hence, a period in which the individual can make decisions regarding consumption while taking social influences into account (Charry, 2014). Preadolescence is followed by
adolescence, which is divided into two parts, early (age 12 to 15) and late adolescence (age 16 to 19). Throughout adolescence, individuals develop the capacity to understand abstract concepts (Piaget, 2008). After the age of 15, adolescents’ egocentrism decreases (Elkind, 1967), which enables them to perceive issues by placing the issues within broader contexts.
For example, they can take broader perspectives on the consequences of environmental degradation. Late adolescents’ media preferences are mostly the same as early adolescents. In general, adolescents become more idealistic in late childhood (Caskey & Anfara, Jr., 2007; Kellough & Kellough, 2008; Scales, 2003). Adolescence is central to identity formation, and adolescents’ worldviews are transformed significantly due to their vulnerability to social influence (Alwin & McCammon, 2003; Wray-Lake, Flanagan, & Osgood, 2010). Similar to preadolescents, the changes in adolescents’ attitudes are considered significant indicators of long-term changes in society (Wray-Lake et al., 2010).
Preadolescents’ pessimism about the future effects of environmental degradation increases as they become teenagers (Ojala, 2012b). For example, an Australian study of pre- and early adolescents revealed that half of the participants were extremely concerned about climate change, and a quarter were worried that the world would end in their lifetime (Tucci, Mitchell, & Goddard, 2007). Such environmental concerns should be carefully addressed by authorities. Another study, this time on the coping mechanisms of 12-year-old Swedish children with regard to climate change, revealed the importance of a nuanced strategy in communicating the consequences of environmental degradation (Ojala, 2012a).
Considering the developmental differences between preadolescence and adolescence, it is vital to examine the two groups, their environmental concerns, and their narrative preferences separately. Henceforth, the following hypotheses are suggested regarding their environmental motivations:
H1a: Preadolescents will have greater egoistic environmental motivation than early adolescents will.
H1b: Adolescents will have greater social-altruistic environmental motivation than preadolescents will.
One way to educate the youth regarding environmental protection is through
edutainment, which refers to a genre blending education with entertainment. Edutainment has been used to educate children on topics such as violence prevention, drug and alcohol
prevention, HIV, and Alzheimer’s awareness (V. D. Allen & Solomon, 2012, 2016; Fritsch, Adams, Redd, Sias, & Herrup, 2006; Glik, Nowak, Valente, Sapsis, & Martin, 2002;
Stephenson & Iannone, 2006) and has been shown to increase information gain and self-efficacy among children. While edutainment has been mostly used in health education, recent further applications of edutainment in environmental education have been observed (Bailey et al., 2015; Bir, Widmar, & Clifford, 2020; Markowitz, Laha, Perone, Pea, & Bailenson, 2018; van Loon, Bailenson, Zaki, Bostick, & Willer, 2018). For instance, Markowitz et al. (2018) examined the effect of an immersive virtual reality experience on learning the consequences of climate change and reported that participants indicated a positive knowledge gain or an increased interest in learning about the impacts of climate change. While, this example highlights a successful application of edutainment, it is important to acknowledge that, for edutainment content to be effective, it should be informative, relevant, and enjoyable for the audience (Borzekowski, 1996; Faiella & Ricciardi, 2015; Rabah, Cassidy, & Beauchemin, 2018). One method to increase engagement with edutainment content is through designing a well-constructed narrative (Bell, 2017; Faiella & Ricciardi, 2015).
Using Narrative in Edutainment
Edutainment can be implemented in many forms of media, including books, video games, films (Cooper & Nisbet, 2017), or plays (Bouchard, 2002). These forms comprise a plethora of components that can make them educational as well as entertaining. One of these components is the narrative, which refers to “any communication that describes specific experiences of characters over time” (Epstein, Thomson, Collins, & Pancella, 2009, p. 595). The use of narrative in edutainment design can increase persuasion by eluding the audience’s
resistance to the message (Cin, Zanna, & Fong, 2004; Slater & Rouner, 2002). Cin et al. (2004) describe two mechanisms to elude resistance through narrative: (1) lessening the magnitude of analytical investigation in the message and (2) increasing self-identification with the characters. Therefore, compared to purely informational approaches, narrative-based approaches can broaden the reach of the message, overcome resistance to the message, and increase an individual’s involvement with the subject (Hopfer, 2012; Kreuter et al., 2010; Larkey & Gonzalez, 2007; Zillmann & Brosius, 2000). Additionally, informational narratives can increase short-term knowledge gains (Epstein et al., 2009).
Several studies have shown the benefits of using narrative with regard to children’s environmental attitudes and prosocial behaviours (Bettmann & Lundahl, 2007; Brusse et al., 2017; Eagles & Demare, 1999; Joronen, Konu, Rankin, & Astedt-Kurki, 2012; Moyer-Gusé, 2008; Valkenburg & Piotrowski, 2017). For instance, one study found that both film and written media can strongly influence environmental attitudes and that long-term media effects are as strong as a week-long camp on increasing environmental awareness (Eagles &
Demare, 1999). Likewise, studies of television content encouraging prosocial behaviour have shown that increased exposure to prosocial media increases prosocial behaviour in children (de Leeuw, Kleemans, Rozendaal, Anschütz, & Buijzen, 2015). Moreover, well-crafted narratives have been shown to increase amusement, excitement, focus, and engagement in early adolescents (Marsh et al., 2011).
The enjoyment of the narrative, identification with the narrative protagonist, and behavioural intentions to read similar narratives are measures to assess the effectiveness of the narrative design. Enjoyment, a pivotal element to induce a flow state in the media user, has been linked to improved learning (Bertacchini, Bilotta, Pantano, & Tavernise, 2012). Besides, based on uses and gratifications theory (Ruggiero, 2000), enjoyment can be a predictor of behavioural intent (Ledbetter, Taylor, & Mazer, 2016; Lou, Chau, & Li, 2005;
O’Keefe & Sulanowski, 1995; Quan-Haase & Young, 2010; Rubin, 2002) and even future behaviour (Ajzen, 1985; Gumulya, 2020). Lastly, identification is a mechanism within a narrative that aids persuasion (de Graaf, Hoeken, Sanders, & Beentjes, 2012; Slater &
Rouner, 2002). Cohen (2001) defines identification with media characters as “an imaginative process invoked as a response to characters presented within mediated texts” (p. 250).
Identification with a likeable protagonist can change beliefs, attitudes, or behaviours (Boon & Lomore, 2001; Robinson & Knobloch-Westerwick, 2017). For example, when video games are considered, identification is perceived as a key mechanism in enjoyment (Christoph, Dorothée, & Peter, 2009; Hefner, Klimmt, & Vorderer, 2007; Vorderer, Klimmt, & Ritterfeld, 2004). Overall, enjoyment, identification, and behavioural intentions can assess the quality of the edutainment content; therefore, one of the main goals of this research is to investigate whether a match between the user’s characteristics and the narrative’s elements affects enjoyment, identification, and behavioural intention outcomes.
Elements of Narrative Design
The present study investigates several factors that contribute to the enjoyment of the narrative, identification with the narrative protagonist, and behavioural intentions to read similar narratives in the future. These factors are: (1) the sex of the protagonist, (2) the psychological distance of the story setting, (3) the age of the protagonist, (4) the form of the protagonist, and (5) the environmental motives of the protagonist. Furthermore, children’s preferences for these narrative factors when creating their ideal story are examined.
Sex of the protagonist. Preadolescents identify with same-sex characters, especially when visual similarities are present. Adolescents prefer same-sex characters more than opposite-sex characters, and this preference becomes more robust for males but weakens for females with age (Beyard-Tyler & Sullivan, 1980). Moreover, adolescents identify with same-sex characters who express societal gender stereotypes (Valkenburg & Piotrowski,
2017). Hence, it is expected that both preadolescents and adolescents will prefer same-sex characters but that this preference will be stronger for adolescents (see H2a and H2b).
H2a: Adolescents will show greater (i) identification, (ii) enjoyment, and (iii) behavioural intentions for a story with same-sex characters compared to preadolescents.
H2b: Adolescents will prefer the same-sex character from a story more frequently compared to preadolescents.
Psychological distance in narration. According to construal level theory (CLT) (Liberman & Trope, 2008), as the distance of stimuli (e.g., events, objects) becomes more removed from one’s immediate reality, they become more psychologically distant. While psychologically close stimuli are represented concretely and in detail, psychologically distant stimuli are described more generally and abstractly.
Environmental problems such as climate change typically have a high psychological distance for people (Spence, Poortinga, & Pidgeon, 2012), as people generally think the consequences of these problems will occur in the future, in distant places, or among other people. Specifically, when participants believe that climate change will affect their local area, affect people they believe are similar to them, and occur in the near future, the greater their concern about climate change. Diminishing psychological distance through experimental manipulation, media, or experiences can result in an increase in individual prosocial behaviours (Hardy, Bhattacharjee, Reed II, & Aquino, 2010; Levine, Prosser, Evans, & Reicher, 2005; Loewenstein & Small, 2007; Small & Simonsohn, 2008).
The appreciation of narratives with greater psychological distance requires skills such as planning, abstraction, and perspective-taking (Liberman & Trope, 2014); such skills significantly improve as children age. In preadolescence, children start to look for realistic fantasy. In other words, they can enjoy “content that happens in a fantasy setting” but in
which the characters’ experiences follow “the rules of logic” (Valkenburg & Piotrowski, 2017, p. 72). When their developmental characteristics are considered, adolescents are more skilled than preadolescents are at traversing psychological distances. Adolescents enjoy more complex content that relies on “more abstract ideas and problems” (Valkenburg &
Piotrowski, 2017, p. 83). Based on their increased ability to think abstractly, adolescents are expected to engage better with psychologically distant events (see H3a and H3b).
H3a: Preadolescents will show greater (i) identification, (ii) enjoyment, and (iii) behavioural intentions for psychologically close (i.e., temporally and spatially) narratives compared to early adolescents.
H3b: Preadolescents will prefer psychologically close (i.e., temporally and spatially) narratives more frequently than early adolescents will.
Age of the protagonist. During and after preadolescence, children “identify primarily with real-world human idols, such as sports heroes and movie stars”, who tend to be older than they are (Valkenburg & Piotrowski, 2017, p. 72). Thus, both preadolescents and
adolescents identify with characters of a similar age, and they identify with adults even more (Valkenburg & Piotrowski, 2017). Fifty-five per cent of preadolescents’ favourite reported characters were younger than 16 (Hoffner, 1996). Hence, it is expected that the characters preadolescents prefer will be younger compared to those adolescents prefer (see H4a and H4b).
H4a: Preadolescents will show higher (i) identification, (ii) enjoyment, and (iii) behavioural intentions for narratives with younger characters compared to adolescents.
H4b: Preadolescents will prefer younger characters than adolescents will choose for their stories.
Form of the protagonist. Identification with characters with similar attributes is common (Bandura, 1989). However, identification is not limited to anthropomorphic
characters; individuals can identify with a block in a virtual reality game, as well (Aymerich-Franch, 2012). Likewise, children and adolescents can “identify with both people and
characters” (Cohen, 2001, p. 249). In a study of 183 children from 5 to 12 years old, it was observed that, as children age, they prefer television characters who “can be imagined without resorting to fantastical thinking” (i.e., are socially real) and form stronger parasocial interactions with more realistic characters (Rosaen & Dibble, 2008, p. 147). Valkenburg and Piotrowski (2017) explain this through the demystification of fantasy characters and
children’s increased admiration for the real world. Thus, all participants are expected to prefer human characters (see H5a and H5b).
H5a: Children will show higher (i) identification, (ii) enjoyment, and (iii) behavioural intentions for narratives with the human characters than those with non-human characters.
H5b: Children will prefer a human character more frequently than a non-human character.
Environmental motives of the protagonist. The match between the protagonist’s and audience’s values (e.g., moral judgements) can affect identification, enjoyment, and
behavioural intentions (Eden, Daalmans, & Johnson, 2017; Janicke & Raney, 2015). The (egoistic, social-altruistic, or biospheric) environmental motive of the protagonist may influence how different audiences appreciate the story; thus, the following are hypothesised:
H6a: Preadolescents will show higher (i) identification, (ii) enjoyment, and (iii) behavioural intentions than adolescents will when the protagonist has an egoistic environmental motive.
H6b: Adolescents will show higher (i) identification, (ii) enjoyment, and (iii) behavioural intentions than preadolescents will when the protagonist has a social-altruistic environmental motive.
To summarise, the present study examines the difference between preadolescents and adolescents regarding (1) their environmental motivations; (2) their enjoyment, identification, and behavioural intentions to read similar narratives based on different narrative elements (i.e., the sex, age, form, and environmental motive of the protagonist and the psychological distance in the setting); and (3) their preferred narrative (i.e., the sex, age, form, and
environmental motive of the protagonist and the setting of the story) when they are given the chance to build their own story.
Methods Participants
Two hundred and three participants were recruited for an online survey via convenience sampling. The survey invitation was placed on social media (e.g., Twitter, Facebook) and announced on two Turkish radio programmes. The questionnaire was
available in Turkish, English, and Dutch. Seven participants dropped out of the study because they did not want to participate, but their parents provided consent for their participation. The data were collected from May 13–June 9, 2020. The final sample consisted of 196
participants, (M Age = 13.32, SD Age = 2.05, 117 females, 79 males). Most participants (82.65%) were Turkish speakers (see Table 1).
Table 1
Distribution of Age, Sex, Language, and Conditions across the Sample (N = 196)
Characteristics n % Age 9–12 (Preadolescents) 13–17 (Adolescents) 67 129 34.18 65.82 Sex Female Male 117 79 59.69 40.31 Language Turkish English Dutch 162 22 12 82.65 11.22 6.12
Characteristics n % Story 1
Male – Low Psychological Distance Female – Low Psychological Distance Male – High Psychological Distance Female – High Psychological Distance
50 49 45 52 25.51 25.00 22.96 26.53 Story 2 Younger Older 97 99 49.49 50.51 Story 3 Human – Self Human – Others Human – Nature Robot – Self Robot – Others Robot – Nature 33 35 33 32 33 30 16.84 17.86 16.84 16.33 16.84 15.31
Note. The protagonist’s sex and the psychological distance were manipulated in Story 1. Only the protagonist’s age was manipulated in Story 2. The protagonist’s form and environmental motive were manipulated in Story 3.
Procedure
The data were collected through an online questionnaire using Qualtrics. The session started with a briefing about the researcher, the aim and outline of the study, and lastly, asking for participants’ consent. Initially, the participants’ parents were asked to provide their consent for their children’s participation as well as the sex and birth years of the participants (see Figure 1). After acquiring parental consent, the parents were asked to invite their
children to complete the survey. Similarly, the children were provided with information about the researcher, the topic of the study, and the method of participation.
After providing their consent, the participants were informed that they would be presented with excerpts from three stories and would be asked to answer questions about these stories. The first part of the survey examined the relationship between narrative elements (i.e., the sex, age, form, and environmental motive of the protagonist and
psychological distance in the narration) and their perception of the stories, such as their ease of understanding of the story, their enjoyment, their intention to read similar narratives, and
their identification with the protagonist(s). Each participant was randomly assigned to one story from three sets of stories. The first set included four stories in which the sex of the protagonist (i.e., female or male) and the psychological distance of the story (i.e., distant or close) were manipulated. The second set had two stories in which the age of the protagonist (i.e., younger or older) was manipulated. The last set included six stories in which the protagonist’s form (i.e., human or robot) and environmental motive (i.e., egoistic, social-altruistic, or biospheric) were manipulated (see Appendix A). Because each participant would have to evaluate too many stories in a within-subject design and due to difficulties in creating comparable narratives, all manipulations were done between subjects.
Following the stories, the participants’ environmental motives and how included they feel in nature were measured. In the final part of the survey, the participants were asked to describe their ideal stories. On this page, the participants were asked to choose the sex, age, and form (e.g., human, robot, animal) of the protagonist as well as the time frame and location in which they want the story to take place. After the children selected their ideal stories, the survey ended with a short debrief.
The study was approved by the Ethics Committee of the Amsterdam School of Communication Research, University of Amsterdam (2020-YME-12132).
Figure 1. Flow chart of the survey sections. In Part One, each participant was assigned to one of the white boxes per story.
Stimuli
Story 1: Sex and psychological distance. In the first narrative, the sex of the protagonist (female or male) and the psychological distance of the story (close or distant) were manipulated between participants. The story featured a protagonist, living on a farm, who needs to find a way to grow sufficient crops for their family, as they are struggling
through a drought. The protagonist’s sex was manipulated by using gender-specific names and pronouns (e.g., Olivia for female and Oliver for male). The psychological distance of the story was manipulated by specifying the time (modern day or the year 3050) and location (Earth or Mars). Besides the name of the protagonist and the psychological distance, all four stories were identical.
Story 2: Age. In the second narrative, the protagonist’s age was manipulated between two versions. The protagonist was either 15 years old (adolescent) or 25 years old (adult), but besides this difference, both stories were identical. This narrative featured a girl, Julia, who suffered from the effects of a heat wave in the city. Julia wanted to motivate people to fight against climate change by educating them via an initiative she started with her friends.
Story 3: Protagonist form and environmental motive. In the third and final story, the protagonist’s form (robot or human) and environmental motive (egoistic, social-altruistic, or biospheric) were manipulated between stories. The protagonist had to migrate from their drowning home island due to rising water levels to a city built on the sunken remains of London. The protagonist decides to travel back in time to educate people about the future of the planet and save either himself, his family, or the planet’s animals from the effects of environmental degradation. Besides the protagonist’s form and motive, all stories were identical.
Ease of understanding the stimuli. In order to ensure all stories were comparable in difficulty, the ease of understanding each story was measured with one item (Appendix B). Overall, all three stories were easy to understand (M1 = 4.36, SD1 = .80; M2 = 4.31, SD2 = .76; M3 = 4.23, SD3 = .81) and there was no significant difference between stories in ease of understanding based on a repeated measures ANOVA test (F [2, 390] = 1.94, p = .145, η2 = .01).
Measures
All variables reported were measured with five-point scales (e.g., 1 = Not at all important, 5 = Very important; 1 = Very difficult to understand, 5 = Very easy to understand) (see Appendix B).
Control variables.
Environmental motives. The original 12-item Children’s Environmental Motives Scale (ChEMS) by Bruni et al. (2012) was used to assess the participants’ environmental motives. ChEMS is based on the idea that “children’s environmental concerns for the consequences of polluting the environment” are tripartite (i.e., egoistic, social-altruistic, and biospheric) (Bruni et al., 2012, p. 12). For example, the biospheric component of ChEMS entails concern for animals, birds, plants, and trees. The 12 items formed a reliable (α = .80, M = 4.62, SD = .38) scale. Since our hypotheses regarding environmental motives concern different dimensions, we calculated the mean scores separately for each dimension (αSELF = .67, MSELF = 4.58, SDSELF = .48; αOTHERS = .79, MOTHERS = 4.53, SDOTHERS = .56; αNATURE = .84, MNATURE = 4.75, SDNATURE = .51).
Inclusion of self in nature. Participants’ perceptions of their inclusion in nature was measured with a single graphical item based on Larson et al. (2019), as inclusion can be an important factor that affects their interest in environmental issues and consequences. Respondents showed above-average inclusion in nature (M = 3.58, SD = 1.00).
Dependent variables.
Enjoyment. Enjoyment of the narrative was measured by adding one item (i.e., “I think this story is pleasant”) to the original four-item enjoyment scale by Ghani et al. (1991). Overall, the respondents enjoyed the stories (α1 = .85, M1 = 3.38, SD1 = .81; α2 = .86, M2 = 3.38, SD2 = .83; α3 = .89, M3 = 3.82, SD3 = .86).
Behavioural intentions. Participants’ future intentions to read (i.e., the rest of the story and similar stories in the future) were measured with two items based on Koufaris (2002) (α1 = .79, α2 = .85, α3 = .85). The items were merged to create a behavioural intentions measure (M1 = 3.80, SD1 = 1.01; M2 = 3.51, SD2 = 1.10; M3 = 3.95, SD3 = 1.05).
Identification. Identification with the protagonists in the narratives was measured with four items (e.g., “While reading the story, I forgot myself and was fully absorbed”) based on Cohen (2001). For each story, the scale showed adequate internal consistency (α1 = .71, α2 = .76, α3 = .80). Four items were merged into the variable “Identification” by averaging the items (M1 = 3.56, SD1 = .80; M2 = 3.76, SD2 = .78; M3 = 3.86, SD3 = .78).
Own narratives. At the end of the survey, participants were asked to form their ideal storyline by choosing different story elements. Participants chose among several options (including “no preference”) for the sex, age, and form of their protagonist as well as the time and place of the story.
Results Randomisation Checks
Chi-square tests were conducted to see whether the participants were similarly distributed into different conditions based on their age group (i.e., preadolescents and adolescents) and sex (Table 2). Overall, the analyses demonstrated that the randomisation of the groups was successful.
Table 2
Story and Condition by Age Group and Sex
Age Group Chi-square
tests of independence for Age Group
Sex Chi-square
tests of independence
for Sex
Condition Preadolescents Adolescents Male Female
Story 1
Male – Low psychological distance Female – Low psychological distance Male – High psychological distance Female – High psychological distance
15 15 19 18 35 34 26 34 χ2 (3) = 1.96, p = .580 24 17 18 20 26 32 27 32 χ2 (3) = 1.95, p = .584 Story 2 Younger Older 33 34 66 63 χ2 (1) = 0.06, p = .800 43 36 56 61 χ2 (1) = 0.81, p = .367 Story 3 Human – Self Human – Others Human – Nature Robot – Self Robot – Others Robot – Nature 11 11 10 10 15 10 22 24 23 22 18 20 χ2 (5) = 2.35, p = .800 9 19 18 11 13 9 24 16 15 21 20 21 χ2 (5) = 9.76, p = .082
Relationship between the Variables
Bivariate correlations were conducted to gain an overview of the relationship between the variables. The results show that age had a significant correlation with ease of
understanding, enjoyment, behavioural intentions, and environmental motives. In general, older children understood the stories better, while younger children showed greater
enjoyment, identification, and intentions to read the stories (See Table 3). Moreover, younger children showed higher overall environmental motive scores, proving the “adolescent dip” is present in this sample. In line with Bruni et al. (2012), children who felt more inclusion in nature also showed significantly higher environmental motives. Environmental motives and inclusion in nature also showed a positive correlation with each other and with ease of understanding, identification with the protagonist, and enjoyment of the narrative. The dependent variables positively correlated with each other, as well. For example, a participant who enjoyed the story more also showed greater identification and behavioural intentions than a participant who enjoyed the story less.
Table 3
Correlation Matrix for Outcome Variables and Age (N = 196)
1 2 3 4 5 6 7 1 IOSIN - 2 Age - .09 - 3 Ease of understanding .01 .21** - 4 Identification .18* - .12 .30** - 5 Enjoyment .18* - .17* .27** .60** - 6 Behavioural Intentions .25** - .18* .22** .58** .71** - 7 Environmental Motives .28** - .17* .25** .28** .27** .38** - Note. IOSIN = Inclusion of self in nature. * p < .05 (two-tailed), ** p < .01 (two-tailed). See Appendix C for a further analysis of correlations per story and condition.
Given the significant correlations between the control variables ease of understanding and inclusion of self in nature and the dependent variables, we further examined their
these variables was successful (Table 4). Nevertheless, whenever control variables correlated with dependent variables in a given story, they were used as covariates for subsequent
analyses in hypotheses testing.1 Table 4
One-Way Analysis of Variance of Ease of Understanding and IOSIN by Condition
Ease of Understanding IOSIN
Activity F df p F df p
Story 1 0.30 3,192 .829 1.49 3,192 .219
Story 2 0.39 1,194 .534 0.01 1,194 .934
Story 3 0.95 5,190 .453 0.59 5,190 .704
Note. IOSIN = Inclusion of self in nature. Hypothesis Testing
Environmental motivations of the participants. The relationship between age group and environmental motivation (EM) was measured with one-way ANOVA tests. The results show that age group had a significant effect on the overall EM scores such that
preadolescents had higher EM scores than adolescents (F[1, 194] = 4.40, p = .037, η2 = .02). As for the different types of EM, age group had a significant effect on egoistic (F[1, 178.34] = 8.57, p = .027, η2 = .04) and biospheric motives (F[1, 193.31] = 5.31, p = .022, η2 = .02). Once again, preadolescents had higher egoistic and biospheric EMs than adolescents had (see Table 5). The difference in the social-altruistic dimension was insignificant.
Table 5
One-Way Analysis of Variance of Environmental Motives by Age Group
Measure F df p Effect size M (SD) Preadolescents Adolescents Environmental Motives 4.40 1,194 .037 .02 4.69 (0.32) 4.58 (0.40) Egoistic a 8.57 1,178.34 .027 .04 4.71 (0.36) 4.52 (0.52) Social-altruistic 0.06 1,194 .800 .00 4.54 (0.57) 4.52 (0.56) Biospheric a 5.31 1,193.31 .022 .02 4.85 (0.29) 4.70 (0.59) Note. Effect size = η2a Welch’s ANOVA was used because Levene’s test for equality of variances showed that egoistic (F[1,194] = 13.50, p < .001) and biospheric (F[1,194] = 8.83, p = .003) were heteroscedastic. The effect size for Welch’s ANOVA was calculated via the adjusted omega-squared formula (ω2 = df1 [F-1] / df1 [F-1] + NT).
Overall, H1a suggesting that preadolescents would show greater egoistic EM compared to adolescents was supported. At the same time, the results failed to support H1b, which stated that adolescents would show greater social-altruistic EM compared to the preadolescents.
Effects of story elements on identification, enjoyment, and behavioural intentions. Table 6 at the end of this section demonstrates the means per age group and manipulated narrative element.
Story 1: Sex of the protagonist and the psychological distance in narration. Three three-way ANCOVAs examined the effects of age group (preadolescent or adolescent), sex of the protagonist (male or female), and psychological distance (close or distant) on
enjoyment, identification, and behavioural intentions after controlling for participants’ EM and inclusion in nature. For these analyses, the protagonist’s sex was recoded in two groups: the same or the opposite sex as the participant.
Enjoyment. Age group had no main effect on enjoyment (F[1, 187] = 2.96, p = .087, η2 = .02). There was no main effect of protagonist sex (F[1, 187] = 2.39, p = .124, η2 = .01) or psychological distance (F[1, 187] = 0.00, p = .969, η2 = .00) on enjoyment. No interaction effect was found between age group and protagonist sex (F[1, 187] = 2.16, p = .143, η2 = .01). However, age group and psychological distance interacted with enjoyment (F[1, 187] = 4.04, p = .046, η2 = .02); preadolescents enjoyed psychologically close stories less than adolescents did (MPreadolescents = 3.41, SDPreadolescents = 0.95; MAdolescents = 3.43, SDAadolescents = 0.66). No significant three-way interaction on enjoyment was found, (F[1, 187] = 0.25, p = .623, η2 = .00). The covariate, participant’s EM, was insignificant (F[1, 187] = 3.62, p = .059, η2 = .02).
Identification. There was no significant effect of age group (F[1, 187] = 1.41, p = .237, η2 = .01), protagonist sex (F[1, 187] = 0.14, p = .707, η2 = .00), or psychological distance (F[1, 187] = 0.83, p = .364, η2 = .09) on identification. No interaction was found
between age group and protagonist sex (F[1, 187] = 0.21, p = .649, η2 = .00). Similarly, age group and psychological distance did not interact to identification (F[1, 187] = 1.93, p = .166, η2 = .01). No significant three-way interaction on identification was found (F[1, 187] = 0.95, p = .332, η2 = .01). The covariate, participant EM, was significant (F[1, 187] = 7.54, p = .007, η2 = .04).
Behavioural intentions. There was no significant main effect of age group (F[1, 186] = 1.56, p = .213, η2 = .01), protagonist sex (F[1, 186] = 0.19, p = .747, η2 = .00) or
psychological distance (F[1, 186] = 0.86, p = .356, η2 = .00) on behavioural intentions. No interaction was found between age group and protagonist sex (F[1, 186] = 2.21, p = .139, η2 = .01). Similarly, age group and psychological distance did not interact to behavioural intentions (F[1, 186] = 2.76, p = .098, η2 = .01). No significant three-way interaction on behavioural intentions was found (F[1, 186] = 0.14, p = .713, η2 = .00). Participant EM (F[1, 186] = 4.95, p = .027, η2 = .03) was significant as a covariate, while inclusion in nature was not significant (F[1, 186] = 2.97, p = .086, η2 = .02).
H2a suggesting that adolescents show greater identification, enjoyment, and
behavioural intentions for same-sex characters compared to preadolescents is rejected. On the other hand, preadolescents enjoyed psychologically distant stories more than adolescents in contrast to what part (i) of H3a posits. Hence, H3a suggesting that preadolescents show greater identification, enjoyment, and behavioural intentions for psychologically close narratives than adolescents do is rejected.
Story 2: Age of the protagonist. Three two-way ANCOVA tests were conducted to examine the effects of protagonist age (adolescent or adult) and participants’ age group on enjoyment, identification, and behavioural intentions after controlling for participant’s EM, inclusion in nature, and ease of understanding in Story 2.
Enjoyment. Age group had a significant main effect on enjoyment (F[1, 189] = 7.56, p = .007, η2 = .04); preadolescents enjoyed the story more than adolescents did (MPreadolescents = 3.60, SDPreadolescents = 0.10; MAdolescents = 3.30, SDAadolescents = 0.07). Protagonist age had no main effect on enjoyment (F[1, 189] = 1.65, p = .200, η2 = .01). Age group and protagonist age did not interact to enjoyment (F[1, 189] = 1.19, p = .276, η2 = .01). Participant EM (F[1, 189] = 4.72, p = .400, η2 = .00) was not a significant covariate, while inclusion in nature (F[1, 189] = 5.77, p = .017, η2 = .03) and ease of understanding (F[1, 189] = 15.68, p < .001, η2 = .08) were.
Identification. No significant effect of age group (F[1, 189] = 0.35, p = .557, η2 = .00) or protagonist age (F[1, 189] = 0.29, p = .594, η2 = .01) on identification was found. Age group and protagonist age did not interact to identification (F[1, 189] = 1.06, p = .306, η2 = .01). The covariates participant EM (F[1, 189] = 5.67, p = .019, η2 = .03), inclusion in nature (F[1, 189] = 4.65, p = .003, η2 = .02), and ease of understanding (F[1, 189] = 19.32, p < .001, η2 = .09) were significantly related to identification.
Behavioural intentions. Age group had a significant main effect on behavioural intentions (F[1, 189] = 5.13, p = .025, η2 = .03); preadolescents showed greater behavioural intentions compared to adolescents (MPreadolescents = 3.70, SDPreadolescents = .13; MAdolescents = 3.39, SDAadolescents = .09). Protagonist age had no significant main effect on behavioural intentions (F[1, 189] = 3.15, p = .077, η2 = .02). Age group and protagonist age did not interact to behavioural intentions (F[1, 189] = 3.24, p = .073, η2 = .02). The covariates, participant EM (F[1, 189] = 5.66, p = .018, η2 = .03), inclusion in nature (F[1, 189] = 7.04, p = .009, η2 = .04), and ease of understanding (F[1, 189] = 7.72, p = .006, η2 = .04) were significant.
The results failed to support H4a, which stated that preadolescents show higher identification, enjoyment, and behavioural intentions for younger characters than adolescents do.
Story 3: EM and form of the protagonist. Three three-way ANCOVA tests examined the effects of the participants’ age group (preadolescent or adolescent), the form of the protagonist (robot or human), and the EM of the protagonist (egoistic, social-altruistic, or biospheric) on enjoyment, identification, and behavioural intentions after controlling for participant’s EM, inclusion in nature, and ease of understanding in Story 3.
Enjoyment. There was no significant main effect of age group (F[1, 182] = 3.74, p = .055, η2 = .02), protagonist form (F[1, 182] = 0.32, p = .571, η2 = .00), or protagonist EM (F[2, 182] = 0.83, p = .438, η2 = .01) on enjoyment. Age group and protagonist form did not interact to enjoyment (F[1, 182] = 0.74, p = .391, η2 = .00). Similarly, no interaction was found between age group and protagonist EM (F[2, 182] = 0.48, p = .621, η2 = .01). No significant three-way interaction on identification was found (F[2, 182] = 0.06, p = .945, η2 = .00). The covariates participant EM (F[1, 182] = 3.95, p = .048, η2 = .02) and ease of
understanding (F[1, 182] = 25.79, p < .001, η2 = .12) were significant.
Identification. There was no significant main effect of age group (F[1, 182] = 1.47, p = .228, η2 = .01), protagonist form (F[1, 182] = 0.91, p = .928, η2 = .00), or protagonist EM (F[2, 182] = 1.11, p = .331, η2 = .01) on identification. Age group and protagonist form did not interact to identification (F[1, 182] = 0.02, p = .891, η2 = .00). Similarly, no interaction was found between age group and protagonist EM (F[2, 182] = 0.18, p = .838, η2 = .00). No significant three-way interaction with identification was found (F[2, 182] = 0.68, p = .509, η2 = .01). The covariates participant EM (F[1, 182] = 5.71, p = .018, η2 = .03) and ease of understanding (F[1, 182] = 24.27, p < .001, η2 = .12) were significant.
Behavioural intentions. There was no significant main effect of age group (F[1, 181] = 3.65, p = .058, η2 = .02), protagonist form (F[1, 181] = 0.02, p = .876, η2 = .00), or
protagonist EM (F[2, 181] = 0.55, p = .580, η2 = .01) on behavioural intentions. Age group and protagonist form did not interact to behavioural intentions (F[1, 181] = 0.15, p = .700, η2
= .00). Similarly, no interaction was found between age group and protagonist EM (F[2, 181] = 0.60, p = .550, η2 = .01). No significant three-way interaction with behavioural intentions was found (F[2, 181] = 1.16, p = .315, η2 = .01). The covariates participant EM (F[1, 181] = 5.74, p = .018, η2 = .03) and ease of understanding (F[1, 181] = 14.56, p < .001, η2 = .07) were significant, but inclusion in nature was not (F[1, 181] = 2.83, p = .094, η2 = .02).
H5a stating that children show greater identification, enjoyment, and behavioural intentions for human characters than for non-human characters is rejected. Likewise, H6a claiming that preadolescents show greater identification, enjoyment, and behavioural
intentions for the egoistic motive than adolescents did is rejected. Similarly, the results failed to support H6b stating that adolescents show greater identification, enjoyment, and
Means and Standard Deviations of Dependent Variables by Age Group per Manipulated Element in Narration
Enjoyment Identification Behavioural Intentions
Narrative
Manipulation Preadolescents Adolescents
All
Participants Preadolescents Adolescents
All
Participants Preadolescents Adolescents
All Participants EM of the Protagonist Egoistic 4.20 (0.94)a 3.79 (0.87)a 3.92 (0.91) 4.01 (1.01) 3.78 (0.70) 3.86 (0.81) 4.10 (1.12) 3.82 (0.97) 3.91 (1.02) Social-altruistic 3.89 (0.82)a 3.77 (0.91)a 3.82 (0.87) 3.98 (0.81) 3.91 (0.65) 3.94 (0.71) 4.10 (1.03) 3.95 (1.15) 4.01 (1.10) Biospheric 3.65 (0.76)a 3.76 (0.81)a 3.73 (0.79) 3.64 (0.80) 3.83 (0.84) 3.77 (0.83) 4.13 (0.90) 3.84 (1.08) 3.93 (1.03) Sex of the Protagonist Same-sex 3.55 (0.72) 3.47 (0.76) 3.50 (0.74) 3.74 (0.67) 3.52 (0.80) 3.58 (0.77) 3.86 (1.22) 3.79 (0.93) 3.81 (1.02) Opposite-sex 3.52 (0.98) 3.08 (0.75) 3.25 (0.86) 3.64 (0.82) 3.48 (0.84) 3.54 (0.83) 4.09 (0.85) 3.61 (1.03) 3.79 (0.99) Age of the Protagonist Younger 3.72 (0.95)b 3.27 (0.67)b 3.42 (0.81) 3.90 (0.87) 3.70 (0.75) 3.77 (0.80) 4.05 (1.03)c 3.36 (0.99)c 3.59 (1.05) Older 3.45 (0.87)b 3.27 (0.83)b 3.33 (0.84) 3.74 (0.75) 3.77 (0.78) 3.76 (0.76) 3.50 (1.21)c 3.38 (1.12)c 3.42 (1.14) Form of the Protagonist Human 3.80 (0.90) 3.78 (0.76) 3.78 (0.80) 3.86 (0.98) 3.84 (0.74) 3.84 (0.82) 4.05 (1.08) 3.88 (0.98) 3.96 (1.01) Robot 4.02 (0.82) 3.77 (0.96) 3.86 (0.92) 3.91 (0.79) 3.85 (0.73) 3.87 (0.75) 4.16 (0.95) 3.85 (1.16) 3.96 (1.09) PD in Narration Low 3.41 (0.98)d 3.43 (0.66)d 3.42 (0.75) 3.67 (0.65) 3.63 (0.84) 3.64 (0.79) 3.78 (1.09) 3.80 (0.84) 3.76 (0.91) High 3.64 (0.77)e 3.16 (0.87)e 3.34 (0.86) 3.71 (0.82) 3.34 (0.76) 3.48 (0.80) 4.14 (1.00) 3.61 (1.11) 3.81 (1.10)
Note. PD = Psychological distance. EM = Environmental motive. a, b, c Significant difference between age groups but not the manipulated element. d, e Significant difference between age groups and the manipulated element.
Narrative construction: Associations between story element preferences and age groups. The final part of the survey measured participants’ preferences for different story elements. Six chi-square tests were conducted to measure the relationship between these preferences and participants’ age groups. The results showed a significant relationship between age group and the preferences for the sex (χ2 [2] = 6.66, p = .036) and age (χ2 [4] = 38.09, p < .001) of the protagonist. No association was observed between age group and preferences for protagonist form or story setting (Table 7).
Table 7
Story Element Preference Count and Percentage by Age Group Preadolescents (n = 67) Adolescents (n = 129) Total (N = 196) Chi-square tests of independence Protagonist Sex Same-Sex 24 (35.82%) 31 (24.08%) 55 (28.06) χ2 (2) = 6.66, p = .036, Goodman and Kruskal's tau = .02 Opposite-Sex 3 (4.48%) 1 (.78%) 4 (2.04%) No Preference 40 (59.70%) 97 (75.19%) 137 (69.90%) Protagonist Age Younger than 12 21 (31.30%) 4 (3.10%) 25 (12.80%) χ2 (4) = 38.09, p < .001, Goodman and Kruskal's tau = .04 13–16 27 (40.30%) 47 (36.40%) 74 (37.80%) 17–20 11 (16.40%) 49 (38.00%) 60 (30.60%) Older than 21 2 (3.00%) 14 (10.90%) 16 (8.20%) No Preference 6 (9.00%) 15 (11.60%) 21 (10.70%) Protagonist Form Human 41 (61.20%) 92 (71.30%) 133 (67.30%) χ2 (4) = 2.59, p = .629, Goodman and Kruskal's tau = .01 Robot 6 (9.00%) 9 (7.00%) 15 (7.70%) Animal 4 (6.00%) 8 (6.20%) 12 (6.10%) Other 7 (10.40%) 9 (7.00%) 16 (8.20%) No Preference 9 (13.40%) 11 (8.50%) 20 (10.20%) Story Setting – Time
Past 5 (7.50%) 19 (14.70%) 24 (12.20%) χ2 (3) = 3.02, p = .388, Goodman and Kruskal's tau = .00 Now 21 (31.30%) 32 (24.80%) 53 (27.00%) Future 36 (53.70%) 65 (50.40%) 101 (51.50%) No preference 5 (7.50%) 13 (10.10%) 18 (9.20%)
Preadolescents (n = 67) Adolescents (n = 129) Total (N = 196) Chi-square tests of independence Story Setting – Location My city 7 (10.40%) 15 (11.60%) 22 (11.20%) χ2 (7) = 1.89, p = .966, Goodman and Kruskal's tau = .00 My country 19 (28.40%) 31 (24.00%) 50 (25.50%) Desert 2 (3.00%) 3 (2.30%) 5 (2.60%) Rainforest 8 (11.90%) 19 (14.70%) 27 (13.80%) Underwater 3 (4.50%) 6 (4.70%) 9 (4.60%) Space 11 (16.40%) 17 (13.20%) 28 (14.30%) Other 13 (19.40%) 25 (19.40%) 38 (19.40%) No preference 4 (6.00%) 13 (10.10%) 17 (8.70%) Note. Percentages within the age group are given.
H2b, which stated that adolescents prefer a same-sex character for their story more frequently than preadolescents, is rejected, as the opposite was observed. Furthermore, H4b suggesting that preadolescents choose younger characters more than adolescents do is supported. Preadolescents more frequently chose protagonists to be younger than 16 (71.60%), while adolescents more frequently chose protagonists older than 13 (74.40%). Thirdly, H5b, which suggested that all participants would prefer the human character to the non-human characters, is supported with 67.30% of participants choosing a human
protagonist regardless of age.
Independent of age, the majority of participants chose their own countries and a future time as their story setting. Although the future is temporally distant, the country they reside in is spatially close. Overall, these results failed to support H3b, which suggested that
preadolescents would prefer psychologically close (i.e., both spatially and temporally) narratives more than adolescents.
Discussion
The present study aims to provide insights into the production of developmentally appropriate edutainment content to educate preadolescents and adolescents about
environmental protection by systematically analysing their perceptions and preferences for narrative design elements and their EMs.
Environmental Motives
We hypothesised that there would be differences between preadolescents and adolescents regarding their EMs. As expected, preadolescents had a higher egoistic EM, meaning that they care more about the consequences of environmental degradation for themselves than adolescents do. The decrease in egoistic EM for adolescents can be explained through decreased egocentrism and an increased ability to grasp broader
perspectives after late childhood (Aboud, 1988, 2008; Vezzali et al., 2018). However, we find no support for the expectation that adolescents would show higher social-altruistic EM
compared to preadolescents. Lack of a significant difference between preadolescent’ and adolescent’ social-altruistic EM could be explained by Kahn et al.’s work (2009), which states that children might not be able to differentiate between the effects of climate change on themselves and other humans.
Furthermore, in line with the conclusions of other authors (Kaplan & Kaplan, 2002; Liefländer & Bogner, 2014; Olsson & Gericke, 2016; Tugurian & Carrier, 2017),
preadolescents were more concerned about the environment in general. Moreover, preadolescents showed higher biospheric EM compared to adolescents. The decrease in biospheric EM among adolescents is not surprising considering their decreased inclusion in nature (Krettenauer, Wang, Jia, & Yao, 2019) and the so-called “adolescent dip” (Olsson & Gericke, 2016). Yet, it should be acknowledged that younger children’s self-reported environmental attitudes can be affected more by social desirability than older children’s are (Liefländer & Bogner, 2014; Oerke & Bogner, 2013).
Elaboration likelihood model (Petty & Cacioppo, 1986; Slater & Rouner, 2002) claims that individuals are more likely to elaborate on messages regarding issues that are relevant to
them. Therefore, tackling relevant environmental concerns can increase audience engagement with the content. When younger children are addressed, the effects of environmental
degradation should be framed through the effects on the child and, for example, animals. Narrative Design Elements
We first examined the effects of different narrative elements (i.e., sex, age, and form of the protagonist and psychological distance) on preadolescents’ and adolescents’
experiences of the narrative (i.e., how much the story is enjoyed, how much the reader
identified with the protagonist, and how much they would like to read similar content). Later, we asked children to construct their own narratives by selecting their preferred narrative elements.
The results of the first part of the study showed that, as opposed to our expectations, the sex, age, form, and EM of the protagonist did not affect the individual’s experience of the narrative. While we expected that preadolescents would have a better experience with a narrative with a same-sex, younger protagonist compared to adolescents, the results did not confirm this assumption. Moreover, the analyses of the preferred narrative elements (Part Three) showed that although both preadolescents and adolescents predominantly showed no preference for protagonist sex, contrary to our expectations, preadolescents preferred same-sex characters more frequently than adolescents did when making their ideal stories. On the other hand, our hypothesis regarding protagonist age was confirmed, as both age groups preferred protagonists either around their age or a couple of years older than they are, which is in alignment with Valkenburg and Piotrowski (2017).
Our expectation for Part One that all participants would have a better experience of the narrative with a human protagonist rather than a robot protagonist was not confirmed. However, as expected, more than half of the participants preferred a human protagonist independent of age group in Part Three. Considering that the similarities between character
attributes can increase engagement with the character, our finding that children prefer human characters that are at least their age is in line with social cognitive theory (Bandura, 1989; Hoffner & Buchanan, 2005; Marsh et al., 2011). Based on this finding, we suggest that content creators design narratives with human protagonists to appeal to children from 9 to 17 years old.
Contrary to our expectations, preadolescents enjoyed psychologically distant stories more compared to adolescents; yet, this finding was in line with Parker and Lepper (1992). Although age group and psychological distance affected enjoyment, the analysis of narrative preference did not show an association between age group and preferred psychological distance elements. Regardless of age, more than half of the participants preferred their story to take place in the future, and the most-preferred location for the story was the participant’s home country.
Children’s increased enjoyment from the narrative could result in greater behavioural intentions and improved learning (Ajzen, 1985; Charsky, 2010; Gumulya, 2020; Ledbetter et al., 2016; Lou et al., 2005; O’Keefe & Sulanowski, 1995; Quan-Haase & Young, 2010; Rubin, 2002). After integrating our findings, we have several suggestions to content creators on designing enjoyable narratives. For preadolescents, narratives with same-sex and similarly aged human protagonists in a psychologically distant setting would be more enjoyable and preferable. On the other hand, adolescents would enjoy and prefer a psychologically close story with a human protagonist who is slightly older than they are (Valkenburg & Piotrowski, 2017).
Finally, we found positive correlations between EM, ease of understanding, enjoyment, identification, and behavioural intentions. This suggests that children who are more concerned with environment engaged with the stories more and hence had a better understanding and experience of the narratives. However, EMs were measured after the
narratives were read, and the stories might have stimulated the children’s awareness (Roy, 2018; Singh & Khanna, 2014). Therefore, it is also possible that the participants who had a better understanding and experience of the narratives became more aware of the
environmental degradation covered in the stories, such as drought, heat waves, and rising sea levels. Lastly, the correlations showed that, although younger children found the stories harder to understand, they had a better experience of the stories than the older children did. Having a better experience of the stories, albeit finding them less easy to understand contradicts with Leavitt and Christenfeld (2013). They found a positive effect of ease of understanding on the enjoyment of narratives. Nevertheless, based on the flow theory, cognitive challenges that are proportionate to the individual’s capacity can make the media more enjoyable, and induce a flow state (Sherry, 2004).
Limitations and Recommendations for Future Research
Although this research delivers a broad examination of the effects of many narrative elements, environmental concerns, and age groups, the domination of Turkish-speaking participants in the sample makes contextual comparisons harder. Nevertheless, since the results are in line with the existing research that reported lower environmental interest and motives as well as inclusion in nature in adolescents, our results are generalisable to other children aged 9 to 17 years. However, because the definitions of age groups might change within the literature, making smaller groups (i.e. age 9–10, age 11–12) and then interpreting the results within the developmental phase these groups belong to could improve the
generalisability of the results.
Another limitation of the research was the broad scope of the application of narrative design (e.g., games, films), making the findings less generalisable across media other than text. Building on the results of this research, future researchers can test whether the
Future studies should also research how the visual appearances of characters affect identification and hence the effectiveness of the narrative (Gulz, 2004).
The nature of the relationships between enjoyment, identification, and behavioural intentions (i.e., main effect, moderation, or mediation) calls for further attention. Similarly, the correlation between ease of understanding and the dependent variables has alternative explanations and needs further investigation.
Conclusion
When it comes to producing content for children, the study showed that the separation of age groups based on developmental characteristics is necessary for story elements such as protagonist sex, protagonist age, and narrative psychological distance. Since a powerful narrative can result in greater engagement, immersion, motivation, and learning (Naul & Liu, 2020), the results of this research have practical implications for educators and narrative designers. This study also confirmed the “adolescent dip”, as adolescents showed lower inclusion in nature and EM. Preadolescents enjoy psychologically distant stories more than adolescents do; thus, the design of narratives with greater psychological distance or fantasy elements is suggested for preadolescents, while the opposite is recommended for adolescents. The majority of children have no preference for the protagonist’s sex. Children want to see protagonists of a similar age; hence, the age of the target group should be reflected in the design of the protagonist. All in all, environmental degradation is a global concern that needs to be communicated effectively. This study builds on previous work with a systematic
examination of the effects of narrative elements on children, children’s environmental concerns, and children’s preferences for narrative elements. Content creators would benefit from the findings of this research when designing effective narratives for environmental protection.
Acknowledgements
First, I would like to thank the AUC Scholarship Fund and the Amsterdam Excellence Scholarship, as their support allowed me to study at the University of Amsterdam and learn from the best for the last five years.
I want to thank Dr Sindy Sumter for introducing me to Dr Hande Sungur. Dr Hande let this paper be my work while steering me in the right direction when she felt I needed it. My dear friend, Tess Bakker, was always there to translate my materials into Dutch, even though she was going through a hectic period herself. For this, I am incredibly grateful.
I also would like to thank my friends and their families. They put great effort into helping me collect the data. In addition, the Turkish radio stations Radyo Karavan and Açık Radyo helped me reach as many people as possible by sharing my survey in every way they could.
I cannot imagine the past months without the constant understanding, encouragement, and support I received from my boyfriend, Martijn. He was also there to share his knowledge with me whenever I was in need.
Finally, I must express my profound gratitude to my family and especially to my parents, Vildan and Engin Ovalıoğlu, for providing me with unconditional support and continuous encouragement throughout my years of study. This accomplishment would not have been possible without them. Thank you.
References
Aboud, F. E. (1988). Children and prejudice. Cambridge, MA: Addison-Wesley. Aboud, F. E. (2008). A social-cognitive developmental theory of prejudice. In S. M.
Quintana & C. McKown (Eds.), Handbook of race, racism, and the developing child (pp. 55–71). John Wiley & Sons Inc.
Acuff, D. (1997). What kids buy and why: The psychology of marketing to kids. New York, NY: Free Press.
Ajzen, I. (1985). From intentions to actions: A theory of planned behavior. In J. Kurhl & J. Beckman (Eds.), Action control: From cognitions to behaviors (pp. 11–39). doi:
10.1007/978-3-642-69746-3_2
Alhendawi, A. (2014). Why young people are key to tackling climate change. Retrieved from
https://tinyurl.com/y8abn2es
Allen, M. R., Dube, O. P., Solecki, W., Aragón-Durand, F., Cramer, W., Humphreys, S., … Zickfeld, K. (2018). Framing and context. In Valerie Masson-Delmotte, P. Zhai, H.-O. Pörtner, D. Roberts, J. Skea, P. R. Shukla, … T. Waterfield (Eds.), An IPCC Special Report on the impacts of global warming of 1.5°C above pre-industrial levels and related global greenhouse gas emission pathways, in the context of strengthening the global response to the threat of climate change, sustainable development. Retrieved from https://tinyurl.com/y896f3tq
Allen, V. D., & Solomon, P. (2012). Educational-entertainment as an intervention with black adolescents exposed to community violence. Journal of Prevention & Intervention in the Community, 40(4), 313–324. doi: 10.1080/10852352.2012.707452
Allen, V. D., & Solomon, P. (2016). EVIP—Edutainment violence intervention/prevention model. Journal of Human Behavior in the Social Environment, 26(3–4), 325–335. doi:
Alwin, D. F., & McCammon, R. J. (2003). Generations, cohorts, and social change. In J. T. Mortimer & M. J. Shanahan (Eds.), Handbook of the life course (pp. 23–49). New York: Plenum.
Aymerich-Franch, L. (2012). Can we identify with a block? Identification with non-anthropomorphic avatars in virtual reality games. International Society for Presence Research Annual Conference, 24–26. Philadelphia.
Bailey, J. O., Bailenson, J. N., Flora, J., Armel, K. C., Voelker, D., & Reeves, B. (2015). The impact of vivid messages on reducing energy consumption related to hot water use. Environment and Behavior, 47(5), 570–592. doi: 10.1177/0013916514551604
Bandura, A. (1989). Social cognitive theory. In R. Vasta (Ed.), Annals of child development (Vol. 6, pp. 1–60). Retrieved from https://tinyurl.com/ycagbqlw
Bell, K. (2017). Game on!: Gamification, gameful design, and the rise of the gamer educator. Baltimore: Johns Hopkins University Press.
Bertacchini, F., Bilotta, E., Pantano, P., & Tavernise, A. (2012). Motivating the learning of science topics in secondary school: A constructivist edutainment setting for studying Chaos. Computers & Education, 59(4), 1377–1386. doi:
10.1016/j.compedu.2012.05.001
Bettmann, J. E., & Lundahl, B. W. (2007). Tell me a story: A review of narrative assessments for preschoolers. Child and Adolescent Social Work Journal, 24(5), 455–475. doi:
10.1007/s10560-007-0095-8
Beyard-Tyler, K. C., & Sullivan, H. J. (1980). Adolescent reading preferences for type of theme and sex of character. Reading Research Quarterly, 16(1), 104–120. doi:
10.2307/747350
Bir, C., Widmar, N. O., & Clifford, M. (2020). The intersection of “natural” edutainment and perceptions of natural resource uses. Environmental Communication, 14(2), 168–183.
