Across childhood, one's thinking becomes more abstract, figurative, and nuanced, and less concrete, literal, and black-and-white. This is also the case for children's beliefs (Turiel, 2006). For example, Misgav et al. (2023) found that values become more focused on internal mental states (e.g., wanting to make someone happy) and less focused on the concrete, observable behaviors associated with that value (e.g., helping someone with carrying groceries) as children age. Likewise, we expect younger children's overall worldviews to consist of more concrete beliefs (e.g., my teacher is nice, my guardian angel is an invisible person who keeps me safe) rather than abstract beliefs (e.g., humans are good by nature, there are divine forces at work in my life; Gabora, 2000).

Despite these early worldviews being more literal and specific, they should function as an adult's worldview, guiding children's interpretations and interactions with the world around them just as any schema would. Research on specific worldview components, such as religious beliefs, supports this hypothesis. For example, children (and adults) who believe that God can answer their prayers are likelier to pray (Froese and Uecker, 2022; Lane, 2020), and children prefer peers who share in their religious beliefs (Heiphetz, 2013). Research has also linked children's and adolescents' values with their behavior (Bacchini et al., 2015; Benish-Weisman, 2015; Paciello et al., 2017; Vecchione et al., 2016). In one study, children with higher self-transcendence values (i.e., greater Universalism and Benevolence values on Schwartz's circumplex model of values) were likelier to share a limited number of chocolate coins with another child (Abramson et al., 2018). Other research has found that children with higher self-transcendence values are less likely to display aggressive behavior (Benish-Weisman, 2019).

Early evidence on adult subjects suggests that one's worldview relates to mental health (Jinkerson, 2016; Walker et al., 2010). One study by Alsancak-Akbulut and Barişkin (2020) investigated the connections between worldviews, mental health (specifically, severity of Obsessive-Compulsive symptoms), and experiences of childhood traumas. They found that those with more negative worldviews (i.e., low self-worth and belief that the world is malevolent and meaningless) had worse OCD symptoms. Furthermore, these negative worldviews partially mediated the relationship between frequency of childhood traumas and OCD symptoms. Thus, it seems that more traumatic childhoods are associated with more negative worldviews, which are then both associated with worse mental health. Further research is needed to examine just how strong and stable these effects are, and whether similar effects exist in child subjects.

Some aspects of children's worldviews appear well-developed by middle childhood. Research on Schwartz et al.'s (2001) model of values has identified the same 10 core values (Universalism, Benevolence, Tradition, Conformity, Security, Power, Achievement, Hedonism, Stimulation, and Self-Direction) in both adults and children as young as 7 years old in various cultural contexts (Cieciuch et al., 2016; Collins et al., 2017; Döring et al., 2015; Liem et al., 2011). Still, value priorities do seem to change as one ages. A review by Döring et al. (2016) suggests that, with age, children tend to value self-transcendence values (i.e., Universalism and Benevolence) and conservatism values (i.e., Tradition, Conformity, and Security) less whereas they value self-enhancement (i.e., Power, Achievement, and Hedonism) and openness to change (i.e., Hedonism, Stimulation, and Self-Direction) more. Furthermore, significant life events can lead to a shift in values. For example, adolescents and adults tend to value security more in the time after a traumatic event (Döring et al., 2016).

The development of metaphysical beliefs, another major component of one's worldview, has been studied extensively over the last 25 years. While adults often believe that children have difficulty determining what is “real,” research suggests that children accurately distinguish between make-believe and physical reality and engage in “fantastical thinking” at similar rates as adults (see Woolley, 1997 for a review). Like adults, children are more likely to believe fantastical stories if they are explained through a religious lens. For example, when stories from the Old Testament Bible, such as Jonah and the Whale, were re-written to remove all references to God and religious characters and locations, children reported significantly lower levels of belief in the story (Vaden and Woolley, 2011). These findings were mirrored in a new study by Davoodi et al. (2023), which found that older children (9–10 years old) were more likely believe religious stories than stories involving magic (e.g., using the “power of God” vs. “magical powers” to solve the same issue). It is important to note that these findings likely depend on the religious context that the participating children are raised in. Children who are raised in a religion that involves magic are more likely to believe in the stories about “magical powers.” Indeed, Goldstein and Woolley (2016) found that children believe more in magical beings (i.e., Santa Claus) when that being is emphasized by their cultural context. In summary, although children tend to be more literal and inflexible in their thinking, they seem to have well-developed values and metaphysical beliefs, two key components of one's worldview.

Worldviews appear to be resistant to change (Carey, 1999; Greenberg and Arndt, 2011) and generally cohesive, with beliefs within a worldview being more compatible with one another than contradictory (Cohen, 2001; Oishi et al., 1998; Schwartz et al., 2001). Just as adults are more receptive to information that fits within the framework of their existing worldviews, teenagers appear to have the same inclination. Zummo et al. (2021) found that adolescents with more conservative worldviews (operationalized as worldviews focused on interpersonal competition and individualism) are less receptive to new climate change research findings. Additionally, abstract, fundamental beliefs inform specific, context-dependent beliefs (Narvaez et al., 1999; Rattan et al., 2012). A study by Lane in 2020 found that children and adults who endorse religious beliefs are more likely to believe prayer can affect future events but not wishes. Still, the stability and consistency of children's overall worldviews are yet to be tested.

In addition to understanding the content of children's beliefs, researchers are increasingly examining the origins of these beliefs. Beliefs can form through direct experience with the world (a la “theory-theory;” see Gopnik, 2011) and through social learning (Arrow and Burns, 2004; Conway and Schaller, 2007; Harton and Bourgeois, 2004; Kashima et al., 2015; Kasser et al., 1995; Van Lange et al., 1997; Vandello and Cohen, 2004). Legare (2019) identified five processes of belief development: exploration, observation, participation, imitation, and instruction. Consider beliefs about human nature as an example. A child may come to believe that people are inherently good through their exploration of the world (e.g., others have helped the child when they asked for it), observation (e.g., seeing their parents exchange hugs), participation (e.g., a teacher explaining to the child that their classmate did not yell because they were mean, but because they were frustrated), imitation (e.g., the child imitating pro-social actions), and instruction (e.g., the child being read a story that explicitly says people are good and kind). Beliefs across all worldview dimensions can form through these five processes, but it is yet to be explored which processes are more foundational to worldview development.

Furthermore, children seem to rely on different sources of information for different types of beliefs (Harris and Corriveau, 2021; Davoodi et al., 2020). For example, in research by Payir et al. (2024), a sample of Turkish children relied more heavily on testimonial source to justify their religious beliefs (e.g., “My grandma told me about heaven”) but used more elaboration to justify their scientific beliefs (e.g., “If there was no electricity, then my tablet would not work right now”). Thus, the source of children's various worldview beliefs may differ depending on the domain of the belief.

Parents have a particularly strong influence on their children's beliefs, especially when their children are young. Research has identified strong overlaps between parents' and children's values (Kohn et al., 1986) and beliefs regarding religion, fate, and social cynicism (Oceja, 2009). Parents with authoritative parenting styles, who talk openly about their beliefs, and who have secure attachments with their children seem to transmit their beliefs to their children the most faithfully (Barni et al., 2011; Cheung and Pomerantz, 2015; Fraley et al., 2012; Pratt et al., 2003; Whitbeck and Gecas, 1988). From this research, we can hypothesize that parents have a strong influence on children's overall worldviews. However, instruments that directly measure children's and adults' worldviews are needed to examine their alignment as well as the factors that moderate this relationship.

In addition to environmental and social influences, genetic factors may also play a role in shaping worldviews. Much of this research has centered on Schwartz's circumplex model of values (though see Olson et al., 2001 for a study on heritability of attitudes). A study by Uzefovsky et al. (2015) on Israeli 7-year-olds found that genetic factors explained 29% of variance in self-transcendence values, 47% in self-enhancement values, and 31% in conservatism values. Only openness to change values appeared to be fully explained by environmental factors. Thus, it appears that more survival-oriented beliefs (e.g., the values of security, power, and care for others) have the strongest genetic component. These findings mirror other studies on children and adults, which have also identified genetics as a significant predictor of values (Button et al., 2011; Knafo and Spinath, 2011; Schermer et al., 2008; Waller et al., 1990). Other research, however, suggests that variance explained by genetic factors is often overestimated (Liu et al., 2015; Sauce and Matzel, 2018).