Abstract
Using a questionnaire, the present study examined parents’ beliefs regarding the development of children’s beliefs about science, religion, and the paranormal. The study also investigated parental encouragement of children’s beliefs, as well as parents’ own beliefs within these domains. Results revealed that parents make distinctions between domains in terms of (a) the importance and timing of beliefs in children’s lives, (b) their own beliefs, and (c) what they encourage their children to believe. The results also indicated that parents’ beliefs were often consistent with the beliefs, they encouraged in their children. These findings suggest that parental beliefs provide an important contextual background for children’s development.
Recently, psychologists have begun to recognize that parental beliefs provide important contextual background for children’s development (McGillicuddy-DeLisi & Subramanian, 1994; Super & Harkness, 1986; Tudge et al., 1999). Most of these studies have focused on caregiving practices or on what constitutes an ideal child, and these studies have not attended to parents’ beliefs about what their children should believe. It has been argued that these parental beliefs influence how parents interpret children’s behavior and also shape the types of learning opportunities parents provide to their children (Savage & Gauvain, 1998; Tudge et al., 1999). In the present study, we were interested in exploring a wide range of parental beliefs regarding science, religion, and the paranormal and how various of these beliefs are associated with one another and with the beliefs that parents encourage in their children.
Another reason for examining parental beliefs is to gain a better understanding of how these might potentially influence the emerging belief systems of children. Sociocultural theory suggests that cognitive development is inextricably tied to the child’s social and cultural surroundings (Cole, 1996; Rogoff, 2003; Vygotsky, 1978). Thus, gaining a better understanding of parental beliefs can serve as a first step in exploring how children may appropriate various types of beliefs from their parents and other individuals around them.
Harris (2000, 2002) has suggested that little attention has been given to the idea that children acquire new information through language and cultural artifacts (although see Callanan, 2006). He suggests that this is particularly true for metaphysical questions, such as those involving religious concepts. He suggests that children’s metaphysical credulity stems from three premises: children are not necessarily striving to create a more objective view of reality, there is no “feedback” available in the realm of metaphysics, and as a consequence children must depend on verbal and culturally transmitted information. This reliance is not limited to metaphysics, because there are many scientific phenomena that children cannot directly observe either. For example, children must rely on others to provide information about evolution and the origins of species, the causes and cures of many illnesses, and what happens after death (Evans, Legare, & Rosengren, 2011; Harris, 2002; Harris & Koenig, 2006). Given the influence of adult testimony on children’s cognitive development, it is important to examine links between parents’ beliefs about various domains and the testimony or encouragement they may provide regarding these domains.
Encouraging beliefs in children
Much research has already been conducted on the impact of parental beliefs on children’s development. However, this work has focused primarily on parents’ beliefs about children and how these beliefs impact parenting practices and child development (Sigel, McGillicuddy-DeLisi, & Goodnow, 1992; Super & Harkness, 1986). How parents encourage and support beliefs in science, religion, and the paranormal in their children has received relatively little attention. A main goal of the present study was to investigate parents’ encouragement of these beliefs and how that encouragement is related to parents’ own beliefs.
Several studies offer some evidence that parents influence their children’s beliefs regarding science, the supernatural, and religion. For example, parents’ explanations about scientific principles contribute to children’s understanding of causality in various domains (Callanan & Oakes, 1992). Other research indicates a link between mothers’ and children’s beliefs about the origins of species. Evans (2000) found that fifth through seventh graders’ beliefs about origins were related to the consistency of their mothers’ beliefs. However, no relation between maternal beliefs and the beliefs of first through fourth graders was found. Evans suggests that these age-related differences may have been driven by the development of reasoning skills working in conjunction with maternal beliefs.
Adults also shape children’s beliefs about the supernatural. Families (and institutions) provide support for children’s beliefs in many supernatural entities (e.g., Santa Claus, tooth fairy, Easter bunny) through various rituals, activities, and artifacts (Clark, 1995; Rosengren & Hickling, 1994; Rosengren, Kalish, Hickling, & Gelman, 1994). According to Clark (1995), parents in the United States encourage these beliefs in part to foster their children’s dependency and to preserve the innocence of childhood.
Numerous studies have demonstrated parallels between parents’ religious beliefs and those of their children (Acock & Bengston, 1978; Hayes & Pittelkow, 1993; Myers, 1996). For example, Acock and Bengston (1978) explored the relation between parental orientations to religion and politics and those of their young-adult children (16–26 years of age). Overall, they found that parents’ orientations and behaviors were highly predictive of those of their children. With respect to religious orientations and behaviors, the correlation between parental responses was .67 for religious behavior, .39 for religiosity, and .57 for traditional religious beliefs. No difference was found in the pattern of relations for mothers and fathers. However, significant differences were obtained across the generations, suggesting that while the orientations and self-described behavior of young adults may be predicted by that of their parents, substantial cohort effects exist. For the present study, we did not directly examine children’s beliefs, but we did examine the connection between parents’ current beliefs and beliefs they encourage in their children. We did this in order to examine the coherence of parental beliefs themselves across diverse areas that are often thought to be incompatible (e.g., science vs. religion). Our goal was to utilize a large enough sample that would enable us to statistically examine how these different beliefs relate in adults. Future work might then examine the influence of consistent or inconsistent beliefs on the children’s belief acquisition.
For the present study, we also investigated parents’ beliefs about when children should learn about science, religion, and the distinction between reality and fantasy. Previous research demonstrates that scientific reasoning and scientific explanations typically emerge in middle childhood, whereas religious beliefs and fantasy–reality distinctions emerge by the end of early childhood (at least for U.S. children). For example, Mantzicopoulos, Patrick, and Samarapungavan (2008) note that kindergartners have very limited experience with scientific activities, and Zimmerman (2000) notes that young children have difficulty with hypothesis testing and evaluating evidence that may contradict their prior beliefs. Focusing on a specific domain of scientific reasoning, Evans (2001) demonstrated that U.S., nonfundamentalist children did not endorse evolutionary theories of species’ origins until the ages of 10 to 12 years. Children’s religious beliefs appear to emerge earlier than their scientific understanding. By 5 to 6 years of age, U.S. Protestant children understand that God has special cognitive abilities (Barrett, Richert, & Driesenga, 2001), and by 5 years of age, most U.S. children may be aware of the concept of prayer (Woolley & Phelps, 2001). Four- to five-year-olds can distinguish between fantasy and reality effectively, although they are sometimes unsure how to categorize fantasy entities (Sharon & Woolley, 2004). In addition, 4- to 6-year-olds understand that machines which violate physical or biological laws are impossible, yet they are often unsure about the reality status of possible but unfamiliar machines (Cook & Sobel, 2011). We were interested in whether parents’ beliefs about (a) the emergence of children’s scientific and religious beliefs and (b) children’s ability to distinguish reality from fantasy parallel empirical findings.
Relationships among different types of parental beliefs
Nonscientific beliefs have often been characterized collectively as involving aspects of magical thinking, and treated as similar in nature and in sharp contrast with rational and scientific approaches (Zusne & Jones, 1989). Zusne and Jones (1989) suggest that this dichotomy may step from two different orientations (“objective” and “subjective” world views; 1989, p. 240) towards events in the world. On the one hand they suggest that some individuals emphasize the importance of subjective, inner experiences and thoughts. On the other hand they suggest that other individuals, holding the second view, emphasize the importance of objective evidence and empirical observations (Zusne & Jones, 1989). From this perspective individuals with religious and paranormal beliefs should be found in the same subjective camp. A related view suggests that beliefs in religion and the paranormal should be related because a general belief in paranormal phenomena or nonnatural causality is common to both. From this perspective, certain parallels can be drawn between religious and paranormal views of life after death and communicating through mental messages (e.g., prayer and telepathy; Wilkinson, 1968).
Although various nonscientific beliefs are often lumped together and placed in contrast to scientific beliefs, there appears to be substantial data to suggest that religious and paranormal beliefs (i.e., beliefs in nonscientific or nonnatural phenomena) are not systematically linked and should not be treated in this manner (Kristensen, 1999; Williams, Taylor, & Hintze, 1989). Duncan, Donnelly, & Nicholson (1992) found that most paranormal beliefs were unrelated to religious background (see also Thalbourne & O’Brien, 1999). However, the same researchers found that some paranormal beliefs may be more common in some religious groups than in others. For example, Catholics have been shown to be more likely to believe in astrology than Protestants. Other researchers have found that fundamentalist Protestants are less likely to hold paranormal beliefs than either less strict Protestants or nonbelievers (Hillstrom & Strachan, 2000). Very little research has been conducted to explore relations between scientific and nonscientific beliefs. The present study examined such relations and extends previous work on adult beliefs to focus on parents instead of college students (Duncan et al., 1992; Hillstrom & Strachan, 2000) or adults in the general population (Thalbourne & O’Brien, 1999).
Proponents of a different perspective, the functional alternative hypothesis, suggest that for some individuals traditional religious beliefs have been replaced by nonreligious paranormal beliefs (Emmons & Sobal, 1981; Persinger & Makarec, 1990). These researchers suggest that a strong correlation should be found between religious beliefs and belief in religious paranormal phenomena (e.g., beliefs in angels), while negatively related to nonreligious paranormal beliefs (e.g., extrasensory perception [ESP]). These nonreligious paranormal beliefs are thought to function as alternatives to conventional religious beliefs (Thalbourne & O’Brien, 1999). One of the goals of the present study was to examine the relations among different types of beliefs held by parents.
The present study
The present study extends previous research, including the study by Rosengren et al. (1994), by investigating a wide range of parental beliefs. Specifically, we examined scientific, paranormal, and religious beliefs, as well as fantasy beliefs examined in Rosengren et al.’s study. In addition, we examined parents’ currently held beliefs, whereas Rosengren et al. (1994) only measured parents’ childhood beliefs. We measured the extent to which parents of young children endorsed beliefs regarding religion, the paranormal, and science. We also asked parents about what they believed when they themselves were children. Finally, we asked parents what their children believed and what types of beliefs they encouraged in their children. The data we collected enabled us to compare parents’ beliefs across domains and compare the beliefs parents encourage in their children with their own beliefs. To summarize, we addressed three main research questions. First, do parents’ beliefs about (a) the importance of children learning about science, religion, and the paranormal, (b) the age at which children should learn about these domains, and (c) the developmental mechanisms which shape children’s beliefs, differ by domain? Second, are parents’ beliefs related to one another? That is, is there an underlying structure to parental beliefs about scientific, religious, and paranormal entities and phenomena, and are there individual differences in how related these types of beliefs are? Third, how do parents’ current beliefs relate to their encouragement of their children’s beliefs?
Method
Participants
A questionnaire was distributed to 509 parents of children between the ages of 3 and 7 years living in and near a community of approximately 95,000 in the Midwestern United States. Parents were contacted through local schools and from a participant list of parents who are interested in allowing their children to participate in research projects. These individuals were originally contacted from birth announcements and contacts in pediatricians’ offices. We attempted to obtain a rather diverse sample by explicitly targeting a wide range of preschools and elementary schools. These schools included two preschools, two elementary parochial schools, and three public elementary schools. Return rates varied from 31% to 40%, with an overall return rate of 34% (174 total surveys). This is a typical return rate for parental surveys of this nature, especially those where no additional incentive is offered (see, for example, Nguyen & Rosengren, 2004; Rosengren et al., 1994). Table 1 summarizes the demographic characteristics of the sample. Participants reported that they had one (23%), two (48.3%), three (15.5%), four (8.6%), or more than four children (4.6%). A large number of participants reported taking science classes beyond high school (64.9%). The majority of the participants reported being currently affiliated with a religious organization (67.2%). A wide range of religious affiliations were provided, though the majority of the participants were affiliated with either Protestant religious organizations (58.4%) or Catholic religious organizations (31.0%). Additional affiliations mentioned included Jewish (2), Muslim (1), Unitarian Universalist (5), and nondenominational (4). About half of the participants reported that they had the same religious affiliation at present as they had as a child (51.7%). We unfortunately neglected to include the participant’s age in the questionnaire. Based on data from a similar survey we have conducted in the same community, using the same methods of recruitment, we expect that participants were between the ages of 20 and 45 years.
Characteristics of participants
Survey
The survey contained multiple sections as part of a larger study. Below is a description of the four sections from which we report results within this paper. The first section contained questions asking about the participant. These included questions about the participants’ gender, level of education, and the religious affiliation. A final question in this section asked about the number of science classes the individual had taken since high school.
The second section of the survey contained 14 questions about fantasy and reality, science, and religion. For each of these content areas the same format of questions was used. The first question in each of these areas asked: “How important is it to you that your child knows the difference between fantasy and reality/learns about science/learns about religion?” Participants were given the choice of: unimportant, somewhat unimportant, neutral, somewhat important, and important. The second question in each of the content areas asked: “At what age do you feel your child should be able to make the distinction between fantasy and reality/learn about science/learn about religion?” The next question concerned whether children should make the distinction between fantasy and reality/learn about science/learn about religion on their own or with the help of someone else. Choices included: making the distinction between fantasy and reality/learning about science/learning about religion is not important, on his/her own, with the help of someone, and combination of on his/her own and with help from someone else.
The third section of the survey asked about the participants’ current beliefs (“Please indicate whether you believe in each of the following by circling one of the following options…”). The beliefs examined included: faith healing, angels, miracles, witches, ghosts, power of crystals, spirits, God, extrasensory perception, astrology, reincarnation, communication with the dead, psychics or fortune-tellers, UFOs, mind reading, alien kidnapping, good-luck charms, the power of prayer, extraterrestrials (aliens), and magic. For each of these items, participants were asked to circle one of the following responses: definitely yes, probably yes, not sure, probably no, definitely no.
The fourth section asked the participant about the beliefs that adults in their household might actively encourage or discourage in the target child (“Which of the following beliefs do adults in your household actively encourage or discourage in your child?”). The beliefs examined included: belief in magic as a supernatural power, belief in divine creation as the most accurate view of how humans came into existence, belief that angels exist, belief in the power of prayer, belief that the devil exists, belief in evolution as the most accurate view of how humans came into existence, belief in miracles, belief that science can provide an accurate view of events in the world, belief that witches exist, belief in God, belief that ghosts exist, belief that science provides the best explanation for most events, belief in Santa Claus, belief in the tooth fairy, belief in magic, and belief in fairies. For each of these items, the participant was asked to circle one of the following responses: encourage a great deal, encourage somewhat, neither, discourage somewhat, discourage a great deal.
Results
We begin the results with analyses of parents’ beliefs about the importance of children learning about different domains, the age at which children should learn about these domains, and about the mechanisms which shape children’s learning. Next we present the results of factor analyses which condensed the wide arrays of beliefs (both encouraged by parents and held by parents) into a handful of factors. Lastly, we present analyses which investigated relations between beliefs parents encourage and beliefs parents themselves hold.
Beliefs regarding children learning about fantasy, science, and religion
Parents reported that it was important to learn about science (M = 4.6 out of 5.0, SD = 0.7), learn about the fantasy–reality distinction (M = 4.5, SD = 0.9), and learn about religion (M = 4.4, SD = 1.0). A Friedman test revealed that there were significant differences in the encouragement of these different beliefs, χ2(2) = 7.49, p < .05, Kendall’s W = .02. A series of follow-up Wilcoxon tests revealed that parents reported that it was more important to them that their children learn about science than learn about religion, Z = −2.62, p < .01. Paired comparisons between (a) the importance of learning fantasy versus reality and the importance of learning about religion and (b) the importance of learning about fantasy versus reality and the importance of learning about science were nonsignificant, Z = −1.53, p = .13, and Z = −1.19, p = .23, respectively.
We next examined the ages at which parents thought a child should: (a) learn the distinction between fantasy and reality; (b) learn about science; and (c) learn about religion. On average, parents reported that their children should learn about religion at a younger age (M = 4.9 years, SD = 1.7) than about fantasy versus reality (M = 5.9 years, SD = 1.6), with learning about science in the middle (M = 5.4 years, SD = 1.7). A repeated measures analysis of variance indicated that the domain differences were statistically significant, F(2, 294) = 16.98, p < .001, η2 p = .10. A Tukey’s least significant difference (LSD) post hoc analysis revealed that the ages at which parents thought children should learn about these three domains all differed from one another (ps < .05). Follow-up analyses examined whether the ages at which parents believe children should learn about science, religion, and the paranormal varies as a function of the importance parents assign to these domains. Ages at which children should learn about science and religion were negatively correlated with ratings of the importance of these domains: r = −.47, p < .001 for science, and r = −.46, p < .001 for religion. The correlation between the age at which children should distinguish reality from fantasy and the importance of children making this distinction did not quite reach conventional levels of statistical significance, r = −.16, p = .051.
We also asked parents whether they thought their child should learn about the fantasy–reality distinction, science, and religion, on their own, with help, or by a combination of on their own and with help. We computed a 3 (domain: fantasy vs. reality, science, religion) × 4 (mechanism: not important, on their own, with help, combination) generalized estimating equation analysis which tested for a main effect of mechanism and a Domain × Mechanism interaction. The analysis revealed a significant Domain × Mechanism interaction, Wald χ2(6) = 9049.36, p < .001, and a main effect of mechanism, Wald χ2(3) = 976.02, p < .001. Parents most often believed that children learn on their own and with help, and parents least often agreed that these mechanisms are irrelevant (see Table 2). Inspection of the model’s parameter estimates revealed that the percentage of parents indicating that mechanisms were not important to them did not vary by domain. However, analysis of parameter estimates revealed that parents were more likely to believe that children learn about science on their own and with help than about religion, Wald χ2(1) = 11.06, p < .01. Parents were more likely to believe that children learn about fantasy–reality distinction only on their own than about science on their own, Wald χ2(1) = 1791.53, p < .001. Parents were more likely to believe that children learn about religion with help than learn about science with help, Wald χ2(1) = 4.51, p < .05. All other comparisons of parameter estimates were nonsignificant.
Beliefs about the mechanisms that shape children’s beliefs
Structure of beliefs encouraged by parents
For each of the factor analyses that follow, we performed principal components analyses with promax rotation. A 4-factor solution appeared to best capture adult encouragement of their own children’s beliefs (see Table 3). The first component appeared to be best captured by encouragement of religious beliefs (eigenvalue = 4.68, variance = 31.2%) and included encouragement of beliefs in God, the devil, angels, divine creation, and the efficacy of prayer. The second component, corresponding to encouragement of positive fantasy beliefs (eigenvalue = 2.86, variance = 19.1%), had high loadings for encouragement of beliefs in Santa, the tooth fairy, magic, and fairies. The third component, corresponding to a belief in scientific explanations (eigenvalue = 1.55, variance = 10.3%), had high loadings on encouragement that science provides the most accurate view of events in the world, that science provides the best overall type of explanation, and that evolution best explains human existence. The fourth component, corresponding to negative fantasy figures (eigenvalue = 1.37, variance = 9.1%) had high loadings on encouragement in witches and ghosts.
Pattern matrix for principal components analysis of parental encouragement of children’s beliefs
Note. Factor loadings ≥ .4 are shown in boldface.
The factor loadings of the factor analyses described above were used to create composite factor scores by summing the scores for each factor and dividing the sums by the numbers of items in each factor. We then computed Spearman correlation coefficients in order to determine how types of encouraged beliefs may be related to one another. Encouraging scientific beliefs was negatively correlated with encouraging religious beliefs, ρ = −.48, p < .001. Also, we found a significant but weak positive correlation between encouraging positive fantasy figures and encouraging negative fantasy figures, ρ = .24, p < .01. Correlations between encouraging religious beliefs and encouraging positive fantasy and negative fantasy beliefs were nonsignificant, ρ = −.06, p = .46, and ρ = −.02, p = .81, respectively. Correlations between encouraging scientific beliefs and encouraging positive fantasy and negative fantasy beliefs were nonsignificant, ρ = .03, p = .67, and ρ = .00, p = .99, respectively.
Structure of parental beliefs
A 4-factor solution appeared to best capture the adults’ current beliefs (see Table 4 ). One component was primarily represented by New Age beliefs (eigenvalue = 7.23, variance = 36.2%). This component had high loadings on beliefs in witches, ghosts, spirits, ESP, reincarnation, communication with the dead, and psychics. The second component involved religious beliefs (eigenvalue = 3.39, variance = 17.00%), with high factor loadings for beliefs in faith healing, angels, miracles, God, and the efficacy of prayer. The third component appeared to primarily capture science fiction beliefs (eigenvalue = 1.61, variance = 8.0%), with high factor loadings on beliefs in UFOs, mind reading, alien kidnapping, and aliens. The last component seemed to primarily capture beliefs related to magic (eigenvalue = 1.13, variance = 5.6%). Beliefs in good-luck charms and magic loaded on this component. Belief in astrology loaded moderately on the New Age and magic components, and belief in spirits loaded moderately on the New Age and religious components. In these cases, we omitted both beliefs from the composite factor scores described in the following lines. Scores for New Age, science fiction, and magic components were positively correlated with one another. Correlations between religious beliefs and other types of beliefs were not significantly correlated, on the other hand. Table 5 presents Spearman correlation coefficients for the four belief factors for the sample.
Pattern matrix for principal components analysis of parental beliefs
Note. Factor loadings ≥ .4 are shown in boldface. ESP = extrasensory perception.
Correlation coefficients for relationships among parents’ own beliefs
Note.
* p < .01;
** p < .001.
Relationships among parents’ current beliefs and beliefs encouraged in their children
The factor loadings of the factor analyses described previously were used to create composite factor scores by summing the scores for each factor and dividing the sums by the numbers of items in each factor. These factor scores were used in the analyses that follow.
We examined the relation between parents’ current beliefs and the factors regarding encouragement of beliefs in their children. Adults’ New Age, science fiction, and magical beliefs each were positively correlated with encouragement of belief in positive and negative fantasy figures. Adults’ religious beliefs were positively correlated with encouraging religious beliefs and negatively correlated with encouraging scientific beliefs (see Table 6 ).
Correlation coefficients for relationships between parents’ own beliefs and encouraged beliefs
Note.
* p < .01;
** p < .001.
Discussion
Parental beliefs about scientific, religious, and paranormal phenomena
The results of the present study provide a detailed view of parental beliefs regarding scientific, religious, and paranormal phenomena. One key finding was that there are domain-specific variations in (a) parental beliefs about the importance and emergence of children’s beliefs, (b) beliefs parents endorse, and (c) beliefs parents encourage in their children. For example, parents indicated that it was more important to them that their children learn about science than learn about religion, although parents indicated that children should learn about religion before they learn about science. A second key finding was that parents’ currently held beliefs are often (but not always) consistent with beliefs they encouraged in their children. For example, stronger religious beliefs were associated with stronger encouragement of religious beliefs and weaker encouragement of scientific beliefs. However, there was no relationship between nonreligious paranormal beliefs and encouragement of scientific beliefs.
The first aim of the present study was to investigate domain-related differences in parents’ beliefs about children learning about science, religion, and the paranormal (e.g., distinguishing between reality and fantasy). We found that parents believed that children should learn about religion at a younger age than they should learn about science. This result contrasts with parents indicating that it is more important for their children to learn about science than learn about religion. It may be that parents think that scientific reasoning becomes more important only when children begin to learn about science. However, the present data do not allow us to address this explanation directly. These findings also raise some interesting issues with respect to distinguishing science from other domains. If it is important for children to learn about religion first, then science, and then the fantasy–reality distinction, this would make it difficult for children to distinguish science from other forms of thought or belief. Regardless of the apparent paradoxes, beliefs about the importance of and age for learning about religion, science, and fantasy–reality distinction provide additional dimensions to parental beliefs as a contextual background for children’s development.
A second aim was to describe the underlying structure of parental beliefs involving a wide variety of paranormal, scientific, and religious phenomena. These beliefs are related to one another in complex ways. Parents’ current beliefs tended to center around four main categories—religious, New Age, science fiction, and magic. The 4-factor solution generated by the data demonstrates that paranormal and religious beliefs are distinct. The factor analysis results also demonstrate that not all paranormal beliefs should be lumped together. New Age and science fiction beliefs both involve phenomena and entities beyond everyday experience, yet they are quite distinct. Zusne and Jones (1989) noted that the term “paranormal” is vague, and given the present results, distinctions should be made in future studies of such beliefs.
A third aim of the present study was to examine parental encouragement of various beliefs in their children in relation to parents’ own beliefs. Several developmental theories recognize the importance of parental beliefs on children’s development. For example, beliefs constitute one of three facets of Super and Harkness’ (1986) developmental niche model. The macrosystem in Bronfenbrenner’s (1979) ecological systems theory also involves beliefs. Beliefs may shape child-rearing practices, the opportunities parents give to children, and how parents teach children about the world. Results from the present study suggest that parents place different value on religious, positive and negative fantasy, and scientific beliefs and have different notions about when and how children should acquire particular beliefs.
The results of this study demonstrated that parents’ own beliefs are related to the beliefs which they report cultivating in their children. For example, parents’ religious beliefs were positively correlated with religious beliefs they encouraged in their children and were negatively correlated with encouraging scientific beliefs. Furthermore, parents’ magical beliefs were positively correlated with encouragement of fantasy figures (both positive and negative). Adult encouragement of fantasy figures may strengthen children’s beliefs in these figures (Woolley, Boerger, & Markman, 2004). It is important, however, to note that parental encouragement does not guarantee that children will endorse magical beliefs, given that children’s conceptual and knowledge development also contribute to their beliefs about magic (Rosengren & Hickling, 2000). The present results demonstrated that New Age and science fiction were positively correlated with encouraging beliefs in positive and negative fantasy figures. These correlations are likely due to the fact that New Age and science fiction beliefs themselves tended to center around positive or benign entities or forces. However, it is unclear why New Age, science fiction, and magic beliefs were positively correlated with encouraging beliefs in negative fantasy figures.
The results of the present study have implications for both formal and informal education. Teachers often are faced with children who bring different models of how the world works, and some children may be uncomfortable with scientific explanations which may contradict their worldviews (Evans, Legare, & Rosengren, 2011). Parents who hold strong religious or paranormal beliefs and who strongly encourage such beliefs in their children may contribute to the mismatch between children’s worldviews and explanations children encounter in school. It is also possible that these parents could help children develop ways of reconciling different worldviews or creating hybrid worldviews. In more informal educational settings, parents who hold and encourage religious beliefs may provide explanations and learning opportunities which support their religious views and which avoid scientific explanations. Our data indicated a negative correlation between holding religious beliefs and encouraging scientific beliefs in children. We did not, however, find a significant correlation between holding nonreligious paranormal beliefs and encouraging scientific beliefs. Thus religious and nonreligious paranormal beliefs may impact children in different ways.
Limitations and future directions
One limitation of the present study involves the representativeness of the sample. Participants tended to be more highly educated than the general population. Sixty-seven percent of participants reported having at least a college degree; approximately 27% of the U.S. population reported having a college degree in 2006–2008 (U.S. Census Bureau, 2009). In addition, our sample resided in a particular location in the United States. This fact could limit the generalizability of the results, although it is not clear that the structure of and relationship between different types of beliefs we studied would vary by geographic region. It is far more likely that differences may be seen cross-culturally or across groups with very different religious, scientific, and folklore traditions. Even within this sample a wide range of nonscientific beliefs were evident. Both the range of beliefs and the overall structure are likely to show variation across groups. Both the types of beliefs and how they are structured in the minds of parents may be important factors that influence what children in a particular family or culture come to believe.
Our study was also limited by the methodology used to collect parental self-reports. We relied on parents’ perceptions of beliefs they encouraged in their children, but we did not measure encouragement through other means. As our goal was primarily to examine the structure of parental beliefs, our methods required a relatively large sample. Follow-up studies could investigate in more detail how parental beliefs and their structure influence both parental encouragement of beliefs and children’s actual beliefs. Additionally, adding more open-ended questions to the survey might have allowed us to examine more closely how parents encourage beliefs or to capture and provide more nuanced differences in how parents think about religion, science, and the paranormal. Unstructured or semistructured interviews would have made this possible—although these methods would have limited the feasibility of the sample size of the study.
Given that parents endorse a wide variety of beliefs, additional research is needed to examine how parents reconcile very different, potentially conflicting types of beliefs. Previous research with children and adults demonstrates that individuals sometimes shift between belief systems and sometimes try to reconcile them (Evans et al., 2011; Schult & Wellman, 1997). Future studies need to determine how (and under what circumstances) parents deal with scientific, religious, New Age, and science fiction beliefs in particular. Recent evidence suggests that both children and adults often hold seemingly incompatible explanations of closely related phenomena (Evans et al., 2011). Future studies should address how parents reconcile these types of beliefs when encouraging beliefs in their children and might also examine the strategies parents offer their children for navigating competing belief systems.
Future studies should extend our findings, which relied on parental self-reports, and should investigate how parents encourage children’s beliefs using systematic observations of parents interacting with their children. For example, one could study types of explanations parents give to children, while dyads read books or engage in tasks that relate to science, religion, and fantasy. In-depth interviews with parents and children regarding their beliefs could serve as a basis for additional research as well.
There are numerous directions in which additional research may expand the present results. Additional research could explore the reasons why parents attempt to encourage different beliefs in their children (e.g., parenting guides, family members, formal schooling). Also, future research could investigate the relative importance of parental encouragement of beliefs in relation to other sources of children’s beliefs. Personal experience and direct observation are emphasized by constructivist theories of development, although others have acknowledged social mechanisms. Peer culture, for example, may serve as an important source for children’s beliefs (Opie & Opie, 1959; Vyse, 1997), and Harris and his colleagues (Harris, 2000, 2002; Harris & Koenig, 2006) note that parental “testimony” may be key when children cannot directly observe or experience phenomena on their own (e.g., evolution, earth revolving around the sun). Teachers (particularly in terms of scientific beliefs), mass media, and other social and cultural factors may also shape children’s beliefs in significant ways. The present study provides the bases for further exploring the relations among parental belief systems regarding science, religion, and specific types of paranormal phenomena and for further understanding how parents encourage such beliefs in their children.
Footnotes
This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.
Acknowledgments
We thank the parents who participated in the present study. We also gratefully acknowledge the help of Christine Hahn in the collection and coding of this data.
