Fact or Fiction? Children’s Preferences for Real Versus Make-Believe Stories

Some children and adults are more drawn to the imaginary than others. Here, we examine whether developmental differences also play a role in the degree to which individuals are drawn to make-believe stories over real ones (or vice versa). Experiment 1 explored the influence of the factuality of stories—whether or not stories reflect events that had actually happened—on children’s story preferences. Experiment 2 explored the effect of magical versus realistic content on participants’ story preferences. Age differences were found for both manipulations. The results suggest that despite the surplus of imaginary activity associated with childhood, young children are not more prone to liking “un-real” stories than adults and may in fact like them less.

portion of our waking hours are spent on things that are not real (Bloom, 2010;Dutton, 2004;Gottschall, 2012). At the same time, though, there are considerable individual differences. While a tendency to engage in some degree of imaginary play is a hallmark of typical development (e.g., Charman et al., 1997), it is readily apparent that some children are more prone to fantasy than others. And just as some children create imaginary companions and others do not (Gleason, Sebanc, & Hartup, 2000;Taylor & Mannering, 2007), some adults spend vast amounts of time consuming novels and movies, while others have little taste for fiction at all.
There is a considerable amount of research that focuses on individual differences in the degree to which a person is drawn to fantasy over reality or vice versa, a construct referred to as fantasy orientation (e.g., Sharon & Woolley, 2004;Taylor & Carlson, 2002;Woolley, Boerger, & Markman, 2004). While a great deal of this work has focused on fantasy orientation in children, Taylor and Mannering (2007) suggest that there is continuity between, for instance, a young child with an imaginary companion (Pearson et al., 2001), an adolescent who writes to an imagined other in his or her diary (Seiffge-Krenke, 1997), and adult fiction writers who experience their characters as having minds of their own (Taylor, Doges, & Kohanyi, 2003). From this perspective, the degree to which a person is prone to imaginary pursuits can be conceptualized as an individual difference that likely shows some degree of continuity across the lifespan.
Simultaneously, however, it is possible that a preference for fantasy over reality (or vice versa) may also be subject to developmental differences. This is certainly true of certain types of imaginary activities: Pretend play peaks in the preschool years, then declines between the ages of 5 and 8 (Woolley, 1997), while the percentage of children who have imaginary companions seems to sharply decline after the age of 9 or 10 (Pearson et al., 2001). However, it would be premature to conclude that the disappearance of these specific types of imaginary activity indicates that adults are in general less prone to having a preference for fantasy than children.
One imaginary activity that both adults and children readily engage in involves fictional stories. From the oral traditions of hunter-gatherer tribes to Bollywood, manga, and telenovelas, storytelling is everywhere, and children, like adults, interact with stories on a daily basis. Parents often begin reading to children as early as infancy (Woolley & Cox, 2007), and studies show that children from the ages of 3 to 11 spend between 20 and 40 min per week-day reading (Huston, Wright, Marquis, & Green, 1999). Young children also spend substantial amounts of time viewing television programs: Estimates put children's weekly television consumption at upward of 21 hours per week (Nielsen Media Research, 1998). Given that young children and adults both spend a great deal of time engaged with stories, this may be the ideal domain in which to compare the degree to which children and adults seek out that which they know is not real.
Not all stories are make-believe. In 2010, USA Today's list of the 100 bestselling books of the year featured both stories that purport to be true, such as presidential biographies and Hollywood memoirs, and fiction, such as novels about vampires and Greek gods. Woolley and Cox (2007) distinguish the dimensions of factuality, that is, whether or not something actually exists or has existed, and possibility: whether or not something could exist. The reality show Jersey Shore purports to be factual and follows the melodramatic lives of actual people. Harry Potter and Grey's Anatomy, in contrast, are not factual, but they do differ from each other along the dimension of possibility-Harry Potter, which details the adventures of a boy wizard, is less possible than Grey's Anatomy, which focuses on the lives of a group of fictional surgeons.
While no previous research has examined the degree to which factuality and possibility affect story preferences across development, there is evidence to suggest that there may be a relationship between age and the degree to which stories are evaluated based on realism. Seven-year-olds, but not 5-year-olds, distinguish between different genres of television based on realism (Wright, Huston, Reitz, & Piemyat, 1994). By 9 to 10 years of age, children show decreased viewing of children's media and increased viewing of general media (Bazalgette & Buckingham, 1995), and one dimension that they evaluate the quality of this media on is the degree to which stories are considered credible or true to life (Nikken & van der Voort, 1997). By adulthood, realism is a strong factor in audience enjoyment, and the ability to become absorbed into a narrative is related to how realistic that narrative seems (Green, 2004).
Simultaneously, however, it does not seem to be the case that adults patently avoid fiction with fantastical elements. The alien epic Avatar is the highest grossing movie of all time, 1 and in a recent poll, one third of adult respondents reported reading some form of science fiction (The National Science Foundation, 2002). Based on these statistics, it could be the case that adults and older children value a specific kind of realism in fiction, namely realistic characters, relationships, and social interactions (e.g., Mar & Oatley, 2008;Mar, Oatley, Djikic, & Mulli, 2011) and that preferences for or against genres that contain fantastical elements are merely a matter of individual preference.
The question explored in this article is whether children and adults systematically vary in the degree to which their story preferences are affected by factuality (Experiment 1) and possibility (Experiment 2). If being drawn to fantastical content is a matter of individual difference, but not subject to developmental influences, we would predict that children and adults, as a group, should not differ in the degree to which they are drawn to fictional or fantastical stories over their more realistic counterparts. In contrast, if the decline of the imaginary activities associated with childhood indicates an increased focus on reality and a decreased interest in the imaginary over the course of development, then we would predict that children should be more drawn to make-believe stories than adults. A third possibility is that adults may actually prefer imaginary stories more than children, perhaps because adults have more limited outlets for imaginary play outside of the realm of stories.
In Experiment 1, children and adults were asked to choose between pairs of stories: One that was described as fictional and one that was described as real. In Experiment 2, we directly manipulated the fantasy content of fictional stories: Some stories contained fantastical elements, such as monsters, aliens, or witches, while others contained no imaginary elements. In this way, we aimed to explore the role of both factuality and possibility in reading preferences across development.

Experiment 1
In this experiment, we offered children and adults a series of choices between pairs of stories that were explicitly labeled as either fictional or true. By controlling story content across participants, we aimed to examine the effect of factuality per se on the preferences of preschool aged and early elementary school aged children, as well as adults.

Participants
Thirty-two children participated in this experiment. Sixteen were 4-and 5-yearold children (mean age: 58.8 months, 11 females), and 16 were 6-to 7-year-old children (mean age: 81.9 months, 12 females). Of the 32 child participants, 11 had parents who elected not to report their child's race. Of the remaining 21 participants, 81% were Caucasian, 9.5% were Hispanic, and 9.5% were African American.
Thirty-two undergraduate participants (16 females) also participated in this experiment. Undergraduate participants were recruited outside an undergraduate dining hall, while child participants were recruited via a laboratory database and through local day care centers and schools.
In addition to the participants documented earlier, 12 children and 1 adult were tested but were excluded from the analysis due to failure to answer control questions (see later).

Stimuli and Procedure
Stimuli were 10 handmade picture books. Each book was 8 Â 8 in., made of colored card stock, and featured a generic title (such as "Buried Treasure" or "Life on the Farm") written on the front. Books were presented to participants in five fixed pairs. Each pair contained one story that was presented as true ("this is a true story about X. It really happened") and one that was presented as fictional ("this is a make-believe story about Y. It's make believe"). The content of the stories was counterbalanced across participants, such that if one child heard a story about "a boy/girl who finds a dinosaur bone" described as being true and one about a "boy/girl who found buried treasure" as being fictional, another participant heard the reverse. For descriptions of the content of all 10 stories, see Table 1.
After children had heard the description of each book in a pair, they were asked, "which story do you think sounds like a better story? Which one is the best?" Participants responded by pointing to one of the two books. In this way, children were asked to indicate which of two books sounded like a better book based on a one-sentence description provided to them by the experimenter, much as a parent might hold up two library books for a child to see and ask the child if they would rather read the book about X or the book about Y. In the current experiment, the choice was phrased in terms of which book the child thought "sounded the best," rather than which they "wanted to read" to minimize participants' frustration about not being allowed to read the book they had chosen before proceeding on to the next story pair.
Half of the adult participants were run on stimuli identical to those used in the child experiment, and half were run on adult equivalents, with the words fiction and nonfiction substituted in for real and make-believe, and the content of the stories were aged up to be appropriate for adults. For example, while the child stimuli set included a story about a child finding a dinosaur bone, the adult version of the story focused on an archeologist.
Before being asked to indicate their preference between the two stories, participants in all age groups were asked control questions to verify that they had encoded which story was which (e.g., "Can you show me which story was a true The Great Discovery About a boy/girl who discovered a dinosaur bone 2 Really Big Family About a boy/girl with lots of brothers and sisters 2 Life on the Farm About a boy/girl who lives on a farm 3 The Very First Day About a boy/girl's very first day of school 3 Summer Vacation About a boy/girl on summer vacation 4 Caught About a boy/girl who caught a bank robber 4 TV Star About a boy/girl who got to be on TV 5 Lost in the Woods About a boy/girl who got lost in the woods 5 Afraid of the Dark About a boy/girl who is afraid of the dark story about a boy who discovered a dinosaur bone? And which story was a make-believe story about a boy who finds buried treasure?"). After completing all five trials, participants who had successfully answered all of the control questions were assigned two scores: a numerical score (number of times, out of the five trials, that the participant chose the "make-believe" story option over the realistic story option) and a categorical score, indicating which of the two options, "real" stories or make-believe stories, they chose the majority (at least three fifths) of the time.

Results and Discussion
A preliminary comparison revealed that adult preferences did not vary depending on whether adults were given the child stimuli (M ¼ 2.19, SD ¼ 1.22) or the adult equivalent (M ¼ 2.19, SD ¼ 1.11, t(30) ¼ 0, and p ¼ 1). The adult data were then collapsed across condition, and a one-way analysis of variance (ANOVA) was conducted with choices to the make-believe option (out of five) as the dependent measure and age (4-5 vs. 6-7 vs. adult) as the between-subjects variable.
We then ran a planned comparison between the older and younger child groups and found no significant difference in the mean number of choices of the make-believe option (younger: M ¼ 1.31 and SD ¼ 1.20; older: Collapsing across the child groups, we ran a planned comparison between children and adults. Children chose the make-believe story option (M ¼ 1.59 and SD ¼ 1.04) significantly less often than adults (M ¼ 2.19, SD ¼ 1.15, t(62) ¼ 2.19, p ¼ .033, and d ¼ .55). A similar pattern of results was found with the categorical data: While 28 out of 32 children chose real stories a majority of the time, only 19 out of 32 adults did so, yielding a significant between-group difference ( 2 (1, N ¼ 64) ¼ 5.13 and p ¼ .025).
Finally, we examined whether or not each group had a significant preference either for or against make-believe stories. A one-sample t test revealed that child participants chose the make-believe option significantly less often than predicted by chance (50% of the five trials) (M ¼ 1.59, SD ¼ 1.04, t(31) ¼ 4.92, p < .001, and d ¼ .875), and an effect could be seen in both the 4 to 5 and 6 to 7 age range individually (younger group: t(15) ¼ 3.97, p ¼ .001, and d ¼ .99; older group: t(15) ¼ 3.10, p < .01, and d ¼ .78). In contrast, adults did not choose the makebelieve stories at a rate different from chance (M ¼ 2.19, SD ¼ 1.15, t(31) ¼ 1.54, p ¼ .13, and d ¼ .27). See Figure 1. Categorically, within the child group, about half of the participants (N ¼ 16) demonstrated a strong preference, choosing their favored option at least four out of five times, while the other half (N ¼ 16) demonstrated a 2 to 3 split. Among children who showed a strong preference, 14 out of 16 showed a strong preference for nonfiction, while only two showed a strong preference for fiction (binomial probability test, p < .01).
An identical pattern was found among those who demonstrated a weaker preference, with 14 out of 16 choosing the nonfiction option more frequently than the fiction option (binomial probability test, p < .01).
In sum, children in this experiment showed a bias for selecting real over make-believe stories and chose real stories significantly more often than adults, who showed no preference for either fiction or nonfiction. While only a subset of children showed a strong (at least four out of five) preference for stories that really happened, the tendency to choose real stories more often than make-believe stories was shown individually by over 85% of the children in our sample. In many ways, this result is surprising. Children's media is rife with anthropomorphized animals, cartoons, and fantastical creatures, but there are relatively few blockbusters or bestsellers that would correspond to categories like documentary and nonfiction. So what can account for the significant overall preference for reality found in the child groups here?
One possibility is that, all other things being equal, children are biased in favor of true stories. In many ways, this type of preference would be logical. Given that children are likely to read stories in part to learn about the world, why would they not prefer stories that are true to those that might contain misleading depictions of the world? Perhaps the adults in our sample, in contrast, are not so biased in favor of what is true because they do not need to be, given their wealth of experience and knowledge about the real world. However, it is also possible that the results obtained in Experiment 1 are an artifact of the particular stimuli used. For example, perhaps children-but not adults-shied away from stories described as make-believe because that term seemed babyish and children were attempting to appear more adult-like in their responses. A second concern with Experiment 1 involves the relatively high rate of children failing the manipulation check question. While data were only analyzed from the 32 child participants who passed the manipulation check, that so many children had difficulty identifying which story was which suggests that the presentation of the stimuli-and particularly the labeling of stories as real or make-believe-may have been somewhat confusing, particularly for the youngest children in our sample. Could the difficulty of the task somehow account for children's significant pattern of preference?
Finally, all of the stories presented in this experiment were realistic-whether or not they were described as having actually happened, they could have happened. Perhaps children preferred realistic stories that did, in fact, happen to the fictional alternative because the reality status of the former's contents is more straightforward and clear cut. Given that young children tend to judge implausible events as impossible (Shtulman & Carey, 2007), it is possible that stories that could have happened, but are nonetheless make-believe because they did not, may be somewhat confusing and off-putting, even for children who passed the manipulation check.
One way to explore these alternatives is to set aside the dimension of factuality and ask children and adults to choose between fantasy stories and realistic stories, a categorical distinction they can make with much more expertise (e.g., Corriveau et al., 2009;Woolley & Van Reet, 2006). If the results obtained in Experiment 1 reflect a genuine age difference, where adults prefer un-real stories more than children do, then we might expect to see a similar pattern of results when participants are asked to choose between realistic (possible) and fantastical (impossible) stories, with adults preferring fantasy stories more than children.

Experiment 2
In this experiment, we explored children's and adults' preferences for realistic versus fantastical stories. We contrasted two distinct types of fantasy stories with realistic stories: fantasy stories that focus on fantastical characters (e.g., Richert, Shawber, Hoffman, & Taylor, 2009) and fantasy stories that focus on fantastical events (e.g., Richert & Smith, 2011).

Method Participants
Seventy-two children participated in this experiment. Thirty-six of the children were 4-and 5-year-olds (mean age ¼ 57.8 months, 18 females), and 36 were between the ages of 6 and 8 (mean age ¼ 88.2 months, 18 females). Of the 72 participants, 18 had parents who declined to report their child's race. Of the remaining 54 participants, 83% were white, 11% were of Asian, east Indian, or Middle Eastern descent, and the remaining 6% were Native American (two participants) and Hispanic (one participant). In addition to those documented earlier, three children were run but excluded from the analysis: two for failing the manipulation check questions and one who opted not to complete the experiment.
Thirty-six undergraduate participants (18 females) also participated. Adult and child participants were recruited as in Experiment 1.

Stimuli and Procedure
Ten 8 Â 8 in. card stock books, with generic titles written on the cover, were used as stimuli. As in Experiment 1, participants were presented with pairs of books and told a single sentence about each book. They were then asked memory control questions, as in Experiment 1, to verify that they had encoded which book was which. Finally, children were asked, "which story do you think sounds like a better story? Which one is the best?" The process was repeated for each of the five story pairs. Child participants were assigned to either the fantasy character condition (N ¼ 40) or the fantasy plot condition (N ¼ 32). Children in the fantasy character condition were asked to choose between stories that focused on a fantastical character, such as an alien, an invisible child, or a baby monster, and stories that focused on realistic character (a boy or a girl). Children in the fantasy plot condition were asked to choose between two stories in which the characters were realistic (a boy or a girl), but one story was described as having a realistic plot (e.g., "this story is about a boy/girl with lots of brothers and sisters") while the other had a plot with a fantastical element (e.g., "this story is about a boy/ girl who lives on an invisible farm").
Story content was counterbalanced across participants. For example, if one participant saw a story about an alien who plays hide and seek (fantastical) and a boy or girl who eats cookies (realistic), another participant was asked to choose between stories in which a boy or girl plays hide and seek (realistic) and an alien who eats cookies (fantastical). Undergraduate participants (N ¼ 18 fantasy character condition and N ¼ 18 fantasy plot condition) were run on survey versions of the exact same items as child participants but were not presented with card stock books as visual aids.

Results and Discussion
A 2 (condition: fantasy character vs. fantasy plot)Â 3 (age group: 4-5 year olds vs. 6-8 year olds vs. adult) ANOVA was conducted with choices to the fantastical option (out of five) as the outcome variable. A main effect was found for condition, such that stories that featured normal characters in fantastical plots (M ¼ 3.56 and SD ¼ 1.39) were chosen more frequently than stories that featured a fantasy character (M ¼ 2.98, SD ¼ 1.46, F(1,102) ¼ 4.10, and p ¼ .045). A second main effect was found for age group (F(2,102) ¼ 9.22 and p < .001). There was no significant interaction between the two.
We then conducted three planned comparisons, looking for differences in the frequency with which participants in each age group chose fantastical stories. Collapsing across the fantasy character and fantasy plot conditions, older children chose fantasy stories more often than younger children, but this difference did not reach significance (6-8 year olds: M ¼ 3.11, SD ¼ 1.55; 4-5 year olds: M ¼ 2.64, SD ¼ 1.25), t(70) ¼ 1.42, p ¼ .16, and d ¼ .33). In contrast, adults chose fantasy stories significantly more often than either preschool or early elementary school aged children (4-5 year olds: t(70) ¼ 4.69, p < .001, and d ¼ 1.12; 6-8 year olds: t(70) ¼ 2.71, p ¼ .008, and d ¼ .65). While 33 out of 36 adult participants chose fantastical stories more often than realistic stories, only 41 out of 72 children showed this pattern ( 2 (1, N ¼ 108) ¼ 11.85 and p ¼ .0006).
Finally, we examined the degree to which each of the three age groups demonstrated significant preferences for stories with fantasy content. Elementary school aged children chose fantasy stories significantly more often than chance (M ¼ 3.11, t(35) ¼ 2.37, p ¼ .023, and d ¼ .39), but the 4-and 5-yearold participants showed no significant preference for fantasy stories (M ¼ 2.64, t(35) ¼ .67, p ¼ .51, and d ¼ .11). Like the older children, adult participants chose the fantastical option significantly more often than predicted by chance (M ¼ 4.0, SD ¼ 1.22; t(35) ¼ 7.38, p < .001, and d ¼ 1.23). See Figure 2.
In sum, as in Experiment 1, children in Experiment 2 chose "un-real" stories-in this case, stories with fantastical content-significantly less often than adults. Across manipulations, 4-and 5-year-old children chose the un-real option (the make-believe stories in Experiment 1 and the fantastical stories in Experiment 2) the least often of the three groups, and adults chose that option most frequently. However, in contrast to Experiment 1, children in Experiment 2 did not avoid the un-real option. Four-and five-year-old children had no significant preference either in favor of or against fantasy stories, while older children significantly preferred fantasy stories but showed a smaller bias in favor of these stories than adults. The type of fantastical content-fantastical character or fantastical plot-had no effect on this pattern.
Certain problems discussed earlier were addressed in this current study. In contrast to Experiment 1, the rate of children failing the manipulation check questions in Experiment 2 was quite low-less than 3%. Also, in Experiment 2, the term make-believe, which may have childish connotations, was never explicitly used. It seems unlikely that children would see stories with magical content as babyish, when such a large percentage of the media they see targeted to adults contains fantastical content. Further, in both Experiments 1 and 2, there was a trend toward older children choosing un-real stories more frequently than preschoolers. It seems unlikely that 4-year-olds are more sensitive to appearing immature than elementary schoolers. Rather, the results of Experiment 2 seem to reflect converging evidence that stories more removed from reality become more appealing with age.

General Discussion
In two experiments, we asked child and adult participants to choose between pairs of books based on simple descriptions. In Experiment 1, we examined the effect of factuality-whether or not a story actually happened-on participants' preferences, while in Experiment 2, we examined the effect of possibility, or whether or not a story could happen (in a way that stories featuring fantastical or magical content could not). While genre preferences are often seen as a matter of individual preference, age differences were nonetheless found in this task. Across experiments, a pattern emerged in which adults chose both fictional (Experiment 1) and fantastical (Experiment 2) stories significantly more often than children did. In Experiment 1, child participants showed a significant bias in favor of real stories over make-believe stories, while adults did not distinguish between the two. In contrast, in Experiment 2, both 6-to 8-year-old children and adults chose fantasy stories significantly more often than realistic stories, but the strength of this preference increased with age.
In many ways, this pattern of results is surprising. There are obvious individual differences in the degree to which people are drawn to different media genres: Some adults love fantasy stories, while others would rather read War and Peace than Lord of the Rings. Similarly, children vary from each other on the dimension of fantasy orientation (e.g., Richert & Smith, 2011;Sharon & Woolley, 2004). While we saw some evidence of individual differences here (e.g., some child participants showed a stronger preference in favor of real stories in Experiment 1 than others), the general patterning of results was remarkably consistent within age groups: 85% of child participants chose real stories more than make-believe stories in Experiment 1, while over 90% of adults tested in Experiment 2 chose fantasy stories over their realistic counterparts.
One factor that may have contributed to these results is the method used. Children and adults were not asked to read stories and then indicate which of the two they enjoyed more, but rather, to choose which story sounded better based on a brief description. While this method is analogous to many of the decisions people make about media in the real world-such as choosing to read a book based on the brief description on the back cover, perusing the New York Times list, which provides a one-sentence log line about each book, or deciding which movie to see based on a trailer-it likely minimizes the role of idiosyncratic preferences, because participants are given relatively little information on which to base their choices. For this reason, these data should not be taken to mean that individual differences in fantasy orientation do not influence story preferences in the real world, but rather, to indicate that-at least in some circumstances-developmental differences may also play a role in determining the degree to which individuals seek out un-real stories over their more realistic counterparts. Relatedly, the exact phrasing of the question used in this experiment asked participants which story sounded the best, rather than which story they would like to read. It could be that adults and children differ in their idea of what constitutes a "good story" rather than in preference per se.
It is also worth noting that the story descriptions used in these experiments-particularly Experiment 2-were generally devoid of the conflict and stakes that characterize the kind of stories we see in the adult fiction market. For example, consider the item where participants were offered a choice between "a boy/girl who plays hide and seek" and "an alien who eats cookies" (or vice versa). In this case, the realistic story appears to depict a fairly mundane every day activity. If given the choice between a less mundane realistic story and a fantasy story, it is likely that adult participants would have shown more variation in their responses.
Why, then, were children more attracted than adults to these relatively mundane stories? One explanation for these results is that young children still have a great deal to learn about the real world. Perhaps stories that sound mundane to adults are more interesting to children because they have had less experience with equivalent real-world scenarios themselves. It is possible that fantasy stories would gain some advantage with the younger children in our sample if they heard each story multiple times-giving them ample experience with the mundane realistic story-before they were asked to choose which story they would like to re-read again. Previous research has shown that children are more likely to learn problem-solving techniques from realistic stories than from fantasy stories (Richert & Smith, 2011), supporting the idea that children's story choices in these tasks may have been influenced by their ability to learn from stories that are, in some sense, real.
An alternative interpretation of these data relies on the fact that other forms of imaginary activity, such as pretend play and the creation of imaginary friends, decline over the age range tested here (Pearson et al., 2001;Woolley, 1997). Perhaps both children and adults are prone toward fantasy and engaging in imaginary pursuits, but for adults, this type of activity is more firmly bound to the realm of stories. In most circumstances, it would not be socially acceptable for adults to engage in elaborate games of pretend, but they can spend a great deal of time engaged in the imaginary world of books and movies without anyone batting an eye. While adults generally do not have imaginary friends, many feel as if they have some kind of relationship with fictional characters (Klimmt, Hartman, & Schramm, 2006). It could be that engagement with fiction-and particularly fantasy stories-may increase across development to fill the void left as other forms of imaginary play decrease. A recent study of pretending in middle childhood and beyond found that while a majority of undergraduate participants had ceased pretending, some undergraduates still engaged in pretend games (Smith & Lillard, 2012). Notably, one of the examples cited in this article involves an undergraduate pretending to be the protagonist from the Bourne movies, supporting the idea that fiction may serve as a "prop" in adult games of make-believe (Walton, 1990).
Interestingly, in Experiment 2, participants showed a stronger preference for stories that focused on realistic characters engaged in fantastical plots than for stories that featured fantasy characters, such as aliens or wizards, engaged in realistic plots. This could be because both children and adults put a premium on a specific kind of realism in fiction, namely realistic characters, relationships, and social interactions (e.g., Mar & Oatley, 2008;Mar et al., 2010). It may be easier to identify with a human protagonist, even if that protagonist lives in a world in which magic and impossible things exist. Alternatively, participants could prefer fantastical plots to mundane ones because those plots provide a greater level of escape from the every day world. An alien who eats cookies at school is still going to school, while a child who plays a game in outer space is having an adventure. Future research is needed to distinguish between these possibilities.
In sum, both children and adults spend a great deal of time engaging with stories, but very little research has attempted to investigate the way that preferences for different kinds of stories change with age. Here, we presented cross-sectional findings that suggest that both fictional stories (those that could have happened, but did not) and fantastical stories (stories that contain impossible content and therefore could not have happened) become more appealing with age.
These results suggest that a preference for imaginary stories is not merely a matter of individual difference but may also be influenced by developmental influences. Despite the decline of pretend play in childhood, the current results suggest that adults are not less interested in the imaginary than children are, at least within the domain of stories, and may in fact be more drawn to stories that are not-and could not-be real.