Sample size (Lebanon).

In Wave 1, we set a target sample of 40 children and their parents/caregivers for each of the six religious groups recruited for Wave 1 in Lebanon. This was the largest sample we could imagine attaining given the Lebanon data collection staff and budget and local circumstances at the time of Wave 1 data collection. All families who participated in Wave 1 and indicated willingness to be contacted again for further participation are invited for Wave 2 participation 9–15 months after the initial Wave 1 child interview was conducted, with one exception: Due to challenges and delays in recruitment of Protestant Christian families in Wave 1, we are not including this small religious minority group in our Wave 2 sample. This yields a total possible sample of 200 children (across five religious samples) for Wave 2 data collection in Lebanon. Recruitment for this sample began on May 1, 2023.

Eligibility, inclusion, and exclusion criteria (Lebanon).

Lebanon samples are recruited through the American University of Beirut in Beirut, Lebanon. To be eligible for this multi-time point study, a family must live in Lebanon, one of the parents must hold Lebanese nationality, and the child participating in the study must have spent at least three quarters of their life in Lebanon. At the time of their participation in Wave 1, children were required to be between the ages of 4 years 0 months and 10 years 11.99 months. Children must be native speakers of Lebanese Arabic (the language in which the interview is conducted), and parents/caregivers must be able to read and respond to the Parent/Caregiver Survey in writing in Modern Standard Arabic. Participating parents/caregivers must be parents or legal guardians of the participating child; multiple siblings from the same family are allowed to participate. To be included in Wave 2, at Wave 1 parents/caregivers must have reported that children were being raised in one of the following religious traditions: the Druze faith, Maronite Christianity, Orthodox Christianity, Shia Islam, and Sunni Islam.

Prior to participating in Wave 1, parents/caregivers interested in participating in the study were asked to indicate the family’s religious affiliation. This determined which version of the Child Protocol and Parent/Caregiver Survey was administered to the family at Wave 1; Wave 2 is matched to this Wave 1 decision. In cases where the parents/caregivers are affiliated with different religious groups, the parent/caregiver was asked in this pre-Wave 1 questionnaire to indicate the religious affiliation the child identifies with or is most familiar with. At Wave 2 (as in Wave 1), the child is given a version of the protocol that features stimuli drawn from the religion they are most familiar with; in cases of mixed-religion households, the parent/caregiver whose religious affiliation best matches this selection is asked to complete the Parent/Caregiver Survey.

Although the families recruited for Wave 1 varied in their engagement with religious practices, only those families who explicitly identified as members of a particular religious group were recruited for the study.

The research team working in Lebanon has adopted the following exclusion criteria for all waves of data collection. First, children who do not appear to understand the majority of the Child Protocol are excluded, as are children who lose interest or are reluctant to engage productively, and ultimately prefer to opt out after initially providing consent to participate. Second, children whose parents/caregivers either report a formal diagnosis of a developmental disorder or report their own concerns about their child’s comprehension, verbal expression, working memory, or other factors are also excluded from the sample. Third, children whose sessions are frequently interrupted (e.g., by family members making suggestions about how the child should answer) are excluded from the sample; judgments of the frequency and severity of these interruptions are made by data collectors. We have no current plans to exclude data at the trial or task level; that is, if a child’s data is included in the dataset our default will be to include all of their data. However, there might be some exceptions to this made on a case-by-case basis during regular meetings of the data collection team. In keeping with the practices employed by the DBN, all decisions regarding exclusion are made by the local research team in Lebanon.

Recruitment, data collection, and compensation (Lebanon).

For Wave 1 of this study, participants were recruited through schools, youth groups, and community members who agreed to circulate an invitation to participate in the study to parents and through snow-ball sampling from those already recruited. Recruitment was carried out in multiple regions in Lebanon so as to achieve a diverse sample across the different religious groups. This included Beirut, the Chouf Mountains, the South (the city of Sidon), the North (Keserwan and Akkar), and the Bekaa Valley.

Families who participated in Wave 1 and indicated a willingness to be re-contacted for future waves of data collection are invited to participate in Wave 2 using the contact information that they provided when they initially participated.

All child interviews are conducted in person, usually in the family’s home but sometimes in private spaces of a house of worship or community center or at the American University of Beirut. If a parent/caregiver and the child consent, the interview is video or audio recorded for later analysis. During the interview, the researcher also logs most of the child’s responses in real-time via Qualtrics survey software. After the interview, the designated parent/caregiver is sent a link to the Parent/Caregiver Survey which they complete online via Qualtrics survey software.

To thank participants for their time and their contributions to this study, families are given 10 USD for each participating child upon completion of the Child Protocol. For families with more than one child participating, parents/caregivers receive an additional 10 USD for each child.

Familiarity task (formerly “Sorting task”).

Rationale for retention and revision of task: As in Wave 1, the Wave 2 Child Protocol begins with a Familiarity task (referred to as the “Sorting task” in [1]), designed to assess children’s familiarity with religious agents, fictional agents, and natural entities.

Including a familiarity assessment in all waves of data collection allows for observation of age-related changes in the recognition of targeted religious concepts at the individual level, complementing the cross-sectional insights afforded by the Wave 1 dataset alone. In Wave 2, the Familiarity task also serves as a screening tool, both within a testing session (allowing us to skip over extended questions about agents and entities with which an individual child is unfamiliar) and for future analyses (allowing us to explore variability between children who did versus did not indicate familiarity with a given agent/entity).

The Wave 2 Familiarity task includes the agents and entities that are featured in later tasks in the Wave 2 protocol, rather than the full range of items included in the Wave 1 Sorting task. This allows us to track changes in familiarity within subjects for agents of particular interest to researchers, while reducing the overall length of the protocol and eliminating overlapping questions between the Familiarity task and the Reality Status task (see “Reality Status task,” below).

Task design and procedure: To begin the task, the researcher says, “We’re going to talk about a bunch of different things. Your first job is to tell me if you’ve ever heard of the thing I mention.” Next, the child is presented with two practice trials designed to familiarize children with the task format and to establish that the task can include both familiar and unfamiliar agents. All examples from the child protocol are in Lebanese Arabic, the language in which the Child Protocol was conducted; as a result, the words written in Arabic may differ from Modern Standard Arabic. One practice trial involves a familiar agent (e.g., “ الحكما/الدكاترة,” “doctors” in Lebanese Arabic), and the second practice trial involves the use of a nonsense word that children should indicate is unfamiliar (e.g., “ الهع خع,” “el-ha3kha3”, a nonsense word in Lebanese Arabic). These practice trials are presented in a random order. Finally, the child is presented with six to seven test trials in a random order. On each trial, the child is asked “Have you ever heard of [target agent/entity]?” (response options: “yes” or “no”).

To preserve cultural and religious specificity and appropriateness while maintaining some degree of standardization across samples, each version of the protocol includes a total of six to seven target agents/entities within three categories, as follows.

Each version of the protocol contains two to three religious agents that are widely considered “real” in the cultural-religious setting, matched to the religious agents included in the Wave 1 protocol. This includes one prominent human or human-like religious figure, such as a divine human or a human prophet (e.g., Jesus, the Prophet Muhammad), referred to in this manuscript as Religious Agent 1 (and matched to Agent 1 in the Wave 1 protocol for that sample). It also includes the “Biggest” god in that cultural-religious setting [e.g., God, Allah; see 21], referred to in this manuscript as Religious Agent 2 (again, matched to Agent 2 in that sample’s Wave 1 protocol). Teams can also opt to include an additional prominent religious being in this task, referred to in this manuscript as Religious Agent 3. These agents are also featured later in the protocol in the Property Attributions and Reality Status tasks (described in later sections).

Each version of the protocol also includes two fictional agents widely considered “fictional” or “fantastical” in the cultural-religious setting. These fictional agents are a subset of the fictional agents included in Wave 1. Research teams were advised to choose one agent that is culturally endorsed for children by adults—that is, when adults talk about the agent around children, they tend to use language consistent with it being real (e.g., “بابا نويل/سانتا كلوز” [“Papa Noel/Santa Claus”]); and one agent that is not typically endorsed even for children—that is, an agent for which adults do not go to any special efforts to “make real” for children (e.g., “الغول” [“Ghoul/Monster”]).

Finally, each version of the protocol includes two target entities that are intended to be more natural or scientific in nature. One of these target entities is germs, viruses, or culturally appropriate equivalents (in Lebanon: “الجراثيم;” the research team also provides French or English translations to children as needed); this is matched to the “scientific entity unobservable by the naked eye” featured in Wave 1. The second target entity—newly introduced in Wave 2—is electricity (in Lebanon: “الكهربا”), selected because it is also considered unobservable by the naked eye in previous research [22,23]; the consensus among research team leaders was that this is a term that children would be familiar with in most if not all field sites.

Adaptation for samples in Lebanon: The final set of agents featured in the Wave 2 Familiarity task for the five samples from Lebanon is presented in Table 1.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Sample-specific item selection for the Familiarity task, for the five samples from Lebanon. See main text for descriptions of network-wide guidelines for item selection.

https://doi.org/10.1371/journal.pone.0330727.t001

Planned analyses: This task primarily serves as a familiarity check for other tasks in the protocol (see Reality Status task, below).

Nonetheless, as in Wave 1, we also hypothesize that children’s familiarity with the target agents/entities included in this task will increase with age, and that this increase will be more pronounced for religious agents. Beyond this, we plan to assess age-related changes in familiarity via within-subjects models of changes in familiarity judgments across waves (for target entities that were included in more than one wave of data collection).

Memory span tasks.

Rationale for retention and revision of task: As in Wave 1, the Wave 2 Child Protocol also includes a working memory span task. Memory span tasks are widely used as a measure of short term auditory memory, one aspect of broader cognitive development and executive function [24,25].

Wave 1 featured both “digit span” and “word span” adaptations of the “Memory for Digit Span assessment: Digits Forward and Digits Backward” components of the Wechsler Intelligence Scales for Children-Revised (WISC-R) [26]. To reduce the overall length of the Wave 2 protocol we include only the “forward” set of trials, for only one of the two versions of the task (digit span or word span), depending on which version was more culturally appropriate.

Although the length of the protocol restricts the ability to incorporate a full battery of measures of general cognitive development or executive function, we include an assessment of working memory at each timepoint to allow researchers to better understand how changes in children’s responses across waves of data collection might be shaped by domain-general changes in working memory (known to be associated with other general cognitive developmental changes) versus domain-specific changes in religious beliefs and behaviors (due, e.g., to religious experiences or education that the child may have received between timepoints).

Task design and procedure: The Wave 2 Child Protocol includes the “forward” block of one version of the Memory Span task, either the “digit span” version, featuring the numbers 1–9, spoken in the language of study administration; or the “word span” version, featuring nine familiar words roughly matched in syllable length [see 1]. In Wave 2, research teams were given the option to select either digits or words; to date, all research teams have decided to use the digit span.

Our administration of this forward digit span task is identical to the description in [1]. We describe it briefly here for completeness; see the previous manuscript for more details and discussion.

At the beginning of the Memory Span task, the researcher provides an introduction and set of practice trials, specially designed for use in the range of samples represented by the DBN. The researcher then begins the test trials, following the instructions for the “Memory for Digit Span assessment: Digits Forward” component of the WISC-R [26]. There are a total of 8 “levels” test trials a child can complete, fixed in ascending order of difficulty (Level 1 = sequences of two digits; Level 8 = sequences of nine digits). The researcher reads each sequence aloud at a pace of approximately 1 digit per second; then the child attempts to repeat back the sequence in the same order. Children are not given any feedback about their responses, regardless of accuracy. Responses that are incomplete and responses in which children begin the sequence incorrectly but then correct themselves are both counted as incorrect. If the child repeats at least one of the two sequences back correctly on a given level, they move onto the next level of the task. The task ends when the child provides two incorrect responses within the same level.

Adaptation for samples in Lebanon: This task does not require any sample-specific customization, beyond translation.

Planned analyses: Analyses of the Memory Span task will follow the standards for analyzing “Memory for Digit Span assessment: Digits Forward” components of the WISC-R [26], including analyses of assessing individual differences and developmental differences.

We anticipate that researchers will apply both single-timepoint (cross-sectional) and multi-time point analysis techniques to the data from this task. As in Wave 1, analyses might include examinations of relationships between responses to this task and responses to other tasks included in the Wave 2 Child Protocol within a given cultural-religious sample; assessments of the validity of the memory span task in a given field site; and careful explorations of how responses to the memory span task might shed light on any observed similarities and differences within and across samples for other tasks in the Child Protocol. Applying this logic across timepoints, analyses might include examinations of the degree to which individual-level changes in memory span performance across timepoints might explain changes in responses to other parts of our protocols.

Property Attributions task.

Rationale for retention and revision of task: As in Wave 1, the Wave 2 Child Protocol includes a Property Attributions task, designed to assess children’s concepts of religious agents, with a particular focus on how these agents are understood to be constrained or unconstrained by the “laws” of folk physics, folk biology, folk psychology (including both cognitive-epistemic and social-emotional aspects of folk psychology), and folk sociology. For similar tasks, see [27–29].

Including such an assessment in all waves of data collection allows us to observe age-related changes in concepts of religious agents at the individual level, to complement the cross-sectional insights afforded by the Wave 1 dataset alone. For this task in particular, we consider within-subjects studies of development to provide important insights into the ways in which children’s understanding of the physical, biological, psychological, and sociological constraints on religious beings might depend on children’s personal exposure to religious doctrine and representational practices (e.g., hearing stories about Jesus walking on water; being told that Allah is completely different from humans; seeing pictures and statues), as well as their personal engagement in religious practices (e.g., opening the door for Elijah during Passover, leaving offerings of food for ancestors).

Researchers across many of the field sites and samples included in the network report that the Wave 1 Property Attributions task was very long for children, with many participants losing interest or focus over the course of the task. To address this in Wave 2, we adopt a “planned missingness” approach [30]: For each child, we randomly select two of the four items included within each domain, reducing the length of the task by 50% for each child. This approach allows us to gather data on all domains for each of the target agents at each time point from each child, while distributing the individual properties within domains assessed at Wave 2 across the whole sample.

Task design and procedure: The Wave 1 Property Attributions task consisted of a 20-item close-ended battery intended to assess participants’ representations of the physical, biological, cognitive-epistemic, social-emotional, and sociological properties and constraints of two to three religious agents, as compared to the properties and constraints of ordinary human agents. For a complete description, including the full set of items, see [1].

In Wave 2 each child is shown a random subset of two of the four items for the physical domain, two of the four items for the biological domain, two of the four items for the cognitive-epistemic domain, two of the four items for the social-emotional domain, and two of the four items for the sociological domain. Put another way, each child assesses half of the 20 original items, and each of the 20 original items are assessed by roughly half of the children within each sample.

For each item, the child responds to a prompt about an ordinary human (“a person”), and two to three religious agents (Religious Agent 1, Religious Agent 2, and, if relevant, Religious Agent 3; see description under “Familiarity task” above). These religious agents are matched to the agents featured in Wave 1; they are also featured in the Wave 2 Familiarity task (described in a previous section) and in the Wave 2 Reality Status task (described in a later section). If a child has indicated in the Familiarity task that they are not familiar with one of these agents, this agent is omitted from the Property Attributions task for that child.

Adaptation for samples in Lebanon: The set of religious agents featured in the Property Attributions task for the five samples from Lebanon is described in the “Familiarity task” section above (see Table 1, Religious Agents).

Planned analyses: For single-timepoint (cross-sectional) analyses, see the planned analyses in [1]. We note, however, that in Wave 2 each child contributes only half of the data points that children contributed in Wave 1, per the “planned missingness” approach described above. This makes item-level analyses even more critical for interpreting response patterns and age-related differences.

Bearing in mind the constraints just described, multi-time point analyses might be conducted either at the item level, with each participant contributing data for a subset of items in the original Wave 1 protocol; or at the level of the domain (physical, biological, and so on), with each participant contributing a “score” for each domain.

Pooling across the diverse religious traditions represented in the network, we hypothesize that, on average, attributions of human-like properties to religious agents will decrease across waves; that is, as children grow older they will be more likely to say that religious agents will violate constraints. To test this hypothesis, we will examine the main effect of participant age within subjects, considering only religious agents. We also anticipate that, on average, children’s perceptions of the difference between religious agents and the human will increase across waves; to test this, we will also examine a statistical interaction between agent type (religious versus human) and participant age within subjects.

However, as the examples from Lebanon illustrate, we expect substantial variability across religious groups and religious agents—not only in the degree of these (hypothesized) developmental trends, but also in their direction. For example, in many Muslim settings, the Prophet Muhammad is not considered “divine” from a theological perspective, so we would expect increased recognition of his human status with the acquisition of more religious knowledge. In this case, we would predict that as Muslim children grow older they will be less likely to say that this particular religious agent (the Prophet Muhammad) will violate constraints, and that the perceived difference between the Prophet Muhammad and the (ordinary) human agent will decrease across waves. Exploring similarities and differences in these patterns across cultural-religious settings will be critical to making sense of, adding nuance to, and noting exceptions to any general trends in the pooled sample.

Additional exploratory analyses will investigate whether these hypothesized trends vary across domains (e.g., whether these age-related changes are more pronounced in the physical and biological domains than the cognitive-epistemic, social-emotional, and sociological domains) and across agents (e.g., whether these age-related changes are more pronounced for “Big Gods” than for other religious/supernatural agents).

Reality Status Task (formerly “Justifications task”).

Rationale for retention and revision of task: As in Wave 1, the Wave 2 Child Protocol includes a Reality Status task (referred to as the “Justifications task” in [1]). This task is designed to evaluate children’s judgments about the reality status of a variety of religious, fantastical, and natural entities, as well as their understanding of their own source knowledge regarding these entities (i.e., how they might justify their reality judgments). For similar tasks, see [23,27–29,31–33].

Including the Reality Status task in all waves of data collection allows us to observe age-related changes in concepts of religious agents at the individual level, to complement the cross-sectional insights afforded by the Wave 1 dataset. Again, we consider children’s concepts of religious agents to be particularly sensitive to personal experience, personal exposure to testimony and religious doctrine, and personal engagement in religious practices—all of which underline the importance of within-subjects work across multiple time points.

We have made several changes to the task to address feedback raised by network members. First, to address questions of how children’s reasoning about religious agents compares to their reasoning about other agents and entities, we now ask these questions not only about religious agents, but also about the fictional agents and natural entities in the Wave 2 Familiarity Task. To compensate for the additional time this added to the task, we exclude the block of questions included in Wave 1 focusing on children’s reasoning about “normal, everyday people.” Second, we focus primarily on forced-choice responses, rather than open-ended responses; this is intended to address concerns raised by network members about systematic variability across field sites in children’s willingness and comfort providing open-ended responses to these kinds of questions. Finally, this task includes one new question (and a follow-up) designed to assess an additional aspect of children’s learning about these entities: their strategy for seeking further information about the entities in question.

Task design and procedure: In Wave 2, the Reality Status task features all of the target agents/entities included in the Familiarity Task (described in an earlier section): two to three religious agents (Religious Agent 1, Religious Agent 2, and, if relevant, Religious Agent 3; see description under “Familiarity task” above); two fictional agents, one of which is generally endorsed by adults when they speak to children, while the other is not; and two target entities that are intended to be more natural or scientific in nature (“germs” or “viruses”, and “electricity”). If a child has indicated in the Familiarity task that they are not familiar with one of these agents/entities, it is omitted from the Reality Status task for that child.

The task consists of a separate block of questions for each agent/entity. The two to three religious agents are always presented one after another (Religious Agent 1, followed by Religious Agent 2, followed by Religious Agent 3 if applicable in the sample); the two fictional agents are always presented one after another (endorsed, followed by non-endorsed); and the two natural entities are always presented one after another (germs/viruses, followed by electricity). Beyond this, the order of presentation is random: Some children see the blocks about religious agents first, others see the blocks aboutfictional agents first, and so on. Each block begins with a very brief introduction—“Now I’m going to ask you some questions about [agent/entity]”—followed by questions about children’s judgments about the reality status of the agent. First, the researcher asks the child to make a reality judgment: “Is/are [agent/entity] real or not real?” Then the researcher asks the child to assess their certainty about this reality judgment by asking, “How sure are you that [agent/entity] is/are [child’s answer: real/not real]? A little bit sure, or really sure?” This is followed by an assessment of the child’s source knowledge regarding this reality judgment: “How do you know that [agent/entity] is/are [child’s answer: real/not real]?” The final question in this set of questions is about the child’s sense of community consensus regarding this reality judgment: “Would most people in your [city/town/village] agree with you and say that [agent/entity] is/are [child’s answer: real/not real]?” This set of questions is identical to the questions used in Wave 1.

Finally, the child is asked two additional questions designed to shed light on their information seeking regarding the target agent/entity: “If you wanted to know more about [agent/entity], what would you do? Would you read, ask someone, or figure things out for yourself?” This question is followed by an open-ended question designed to prompt children to elaborate on their previous response; as appropriate, the experimenter asks one of the following questions: “What would you read?” or “Who would you ask?” or “How would you figure things out?”

Adaptation for samples in Lebanon: The set of religious agents, fantastical entities, and natural entities featured in the Reality Status tasks for the five samples from Lebanon is described in the “Familiarity task” section above (see Table 1).

Planned analyses: For single-timepoint (cross-sectional) analyses, see the planned analyses in [1,2]. We anticipate implementing similar coding schemes to characterize trends and variability in the content of children’s responses (e.g., identifying certain content as “religious”; counting mentions of “God” or other religious agents; coding different types of explanations) and their manners of speaking (e.g., noting the use of generic sentences). For similar approaches, see [32,34].

Wave 2 offers a new opportunity to compare children’s responses to religious agents versus fantastical agents versus natural entities; other single-timepoint analyses will focus on these differences.

Additional single-timepoint analyses will focus on children’s forced-choice responses to the new set of questions of the form, “If you wanted to know more about [target agent/entity], what would you do? Would you read, ask someone, or figure things out for yourself?” Such analyses will be exploratory in nature.

Multi-time point analyses will only be possible for the target agents/entities included at both time points: Religious Agent 1, Religious Agent 2, and Religious Agent 3 (where applicable). We hypothesize that, for all open-ended questions, children will offer richer and more detailed examples, descriptions, and explanations of what they know and how they know it at Wave 2 [see 32, 35, 36]. To test this hypothesis, we will examine the main effects of participant age within subjects using any relevant coding schemes we have developed to quantify and characterize children’s responses.

Child as Transmission Agent task.

Rationale for retention and revision of task: The Wave 2 Child as Transmission Agent task is a modified version of the qualitative, exploratory task used in [1], designed to capture the ways in which children are themselves vectors of religious information transmission. Wave 1 included 2 separate blocks related to the celebration of a religious holiday versus practices at home; in Wave 2, we include only the religious holiday block of this task.

As research teams began collecting data for Wave 1, they reported a wide variety of responses to the open-ended questions included in this task in Wave 1. In some settings, children had less experience engaging in one-on-one conversations with adults, especially conversations involving open-ended and abstract questions, and were hesitant to offer more than a few words in response to our questions. In other settings, such conversations were well-received and highly familiar to children. In response to this feedback, in Wave 2 we eliminate one open-ended question and employ a forced-choice format for another question, yielding a slightly shorter task that is better suited to the range of cultural-religious settings represented by the DBN, while allowing for within-subjects exploration of changes in how children talk about their religious holidays over time.

Task design and procedure: As in Wave 1, children are introduced to the task with the following script: “Let’s imagine [character] is a new kid in your neighborhood who is younger than you who doesn’t know what happens at your house. Now let’s say it’s a special day like [religious holiday]. [Character] needs your help to remember what you do when celebrating this special holiday.”

Children are then invited to provide open-ended instructions and explanations about the things that the character should do to observe the religious holiday the right way and prompted to explain why the character should behave this way. For each cultural-religious sample, Wave 2 features the same sample-specific holiday and the same two sample-specific prompts about that holiday that were featured in Wave 1; see Table 2 for examples from samples in Lebanon.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 2. Sample-specific item selection for the Child as Transmission Agent task, for the five samples from Lebanon.

https://doi.org/10.1371/journal.pone.0330727.t002

As in Wave 1, children are then asked to advise the character about what to do if she/he wanted to learn more about the holiday in question. Rather than asking this in an open-ended format, in Wave 2 we instead provide three options and ask children to choose among them: “If [character] has more questions about what to do when celebrating [religious holiday], what should she/he do? Should [he/she] read, ask someone, or figure things out for [his/herself]?” Then children are then prompted to expand on their selected method of information-seeking: “What should [he/she] read?”, “Who should [he/she] ask?”, or “How should he/she figure things out?”

Finally, as in Wave 1, children are asked whether the character is a member of the target religious group or not. Depending on the sampling plans in place at a given field site, the target religious group might correspond to the participating child’s own religious group, to a religious group with which the child’s parent has indicated that they are familiar, or to a religious group that is prevalent in the participating child’s local context. (Note that the religious group named in this task is also featured in two other tasks in the protocol: the Norm Violations task and the Social Essentialism task, described in the following sections).

Adaptation for samples in Lebanon: The final set of properties featured in the Child as Transmission Agent task for the five samples from Lebanon is presented in Table 2.

Planned analyses: Coding schemes developed for analysis of Wave 1 will be applied again to the analysis of open-ended questions. These coding schemes have not yet been developed but could include coding schemes related to identifying certain content as “religious”; counting mentions of “God” or other religious agents; coding different types of explanations for why the character should or should not behave in a certain ways, as well as various manners of speaking [e.g., noting the use of generic sentences; see 1].

We hypothesize that, with increased age and more experience with religious holidays in their own lives between waves, children will be more likely to spontaneously mention religious content (e.g., rules, practices, rituals) and offer richer and more detailed responses at Wave 2, relative to Wave 1 [see 37, 38].

Social Essentialism task.

Rationale for retention and revision of task: As in Wave 1, the Wave 2 Child Protocol includes a Social Essentialism task, designed to gauge the ways in which children do or do not essentialize social groups based on religion, gender, and wealth [see 39-41].

Our goal in including such an assessment in all waves of data collection is to chart changes in essentializing at the individual level. We consider within-subjects exploration to be particularly important, because we speculate that essentializing may be driven in large part by children’s personal experiences, by events in their family lives, and by their observations of their immediate environments (including the ways in which people talk about these social groups around the child)—all of which would result in large individual differences that might be washed out by an overreliance on cross-sectional samples.

At the same time, the process of collecting Wave 1 data across the many field sites and samples included in the network has prompted substantial changes to this task. The goals of our changes are twofold: First, we aim to improve children’s comprehension of the task and the interpretability of their answers; second, we aim to address the many aspects of social (and biological) essentialism that have been assessed in previous work.

The Wave 2 version of this task omits two of the four questions included in Wave 1: one about physical insides and one about group membership at birth. With respect to the physical insides question (“Is it possible to tell whether a person is [non-reference group] or [reference group] just by looking inside their body, like by looking at their blood and bones?”), children in some samples may have less experience with tools that can be used to see people’s insides (e.g., through blood draws, X-rays) compared to children in other samples, which makes it difficult to pool responses across samples responsibly. With respect to the question about group membership at birth (“Are people born that way? Like, [non-reference group] are born [non-reference group] and [reference group] are born [reference group]?”), children might endorse what we initially intended to be an essentialist response option (that people are “born that way”) without an actual commitment to essentialist logic (e.g., children might think that people who are poor now were “born that way” because of structural factors and not because of any inherent or heritable essence).

We have replaced these two items with two new questions about the homogeneity of the social categories in question [see 42, 43]—an important aspect of social essentialism which is not assessed directly in the Wave 1 version of this task.

In addition, Wave 2 includes follow-up questions to the Wave 1 item about the possibility of transformation (a person changing group membership), to clarify the nature of children’s responses. These are intended to differentiate between essentialist (or “internalist”) explanations, which appeal to internal properties inherent and perhaps essential to membership a category (e.g., that members of Category X are genetically predisposed to excel in math), versus structuralist explanations, which instead focus on stable external constraints that systematically apply to category members (e.g., that members of Category X are systematically given opportunities to develop their math skills) [see 44–47].

The remaining item from Wave 1 about spiritual insides has been retained without modification.

To address the need to reduce the overall length of the protocol as a whole, research teams can opt to include two instead of three blocks of this task (i.e., to assess children’s tendency to essentialize religious groups plus either gender-based or wealth-based groups, rather than all three social categories).

The Lebanon research team extended this task to include a block about language (e.g., assessing the essentialization of French speakers vs. English speakers). This was a sample-specific addition to the procedure and is not considered part of the network-wide standard Child Protocol.

Task design and procedure: The Wave 2 Social Essentialism task consists of two or three blocks, presented in random order, focusing on at least two of the following three social categories: religion, gender, and wealth. All research teams must include the religion block in their adaptation of the protocol; beyond this, research teams may opt to include one or both of the remaining social categories.

As in Wave 1, at the beginning of each block, children are shown silhouettes of two groups of people side by side; the researcher identifies these groups as members of contrasting social categories. One group is considered the “reference group” (generally either the child’s ingroup, a majority group in the local setting, or a group of higher social status in the local setting) and one the “non-reference group” (the child’s outgroup, a minority group, or a group of lower social status).

For the religion block, the reference and non-reference groups vary across samples, and are matched to the groups used for each sample in Wave 1. As in Wave 1, this yields a clear ingroup versus outgroup contrast for participating children in some samples, while in other samples, the contrast is between a majority versus minority religious group, or between a higher- versus lower-status religious group.

As in Wave 1, the reference group in the gender block is the participating child’s own gender (e.g., “girls”), and the non-reference group is the contrasting binary gender (e.g., “boys”). If the participating child does not identify as a boy or a girl, or if the participating child’s gender is unknown, the child is shown the “girl” version of the task.

As in Wave 1, the reference group in the wealth block is always “rich people,” and the non-reference group is always “poor people.”

The introduction to the task and each block is identical to Wave 1.

The four questions for each block are then presented in a fixed order. The first two questions are new additions for this wave of data collection: “Do you think all [reference group] are basically the same – for example, they behave in the same way and like the same things?” (reference group homogeneity); and “Do you think all [non-reference group] are basically the same – for example, they behave in the same way and like the same things?” (non-reference group homogeneity).

The next item is repeated from the Wave 1 version of this task: “Is it possible for [non-reference group] to become [reference group]?” (transformation), If a child says “yes” or “maybe”, they are asked the following two follow-up questions (which were not included in Wave 1): “How?” and “Would something in the person need to change or would something in the world need to change?” If a child says “no” to this question, they are instead asked, “Why not?” and “Is it because of something in the person themselves or is it because of something in the world?” For both of these follow-up questions, the two response options (“something in the person” versus “something in the world”) are presented in a random order across children, but in the same order across blocks of the task for each child.

The final item is a repeated question from the Wave 1 Social Essentialism task: “Are [non-reference group] people’s souls different from [reference group] people’s souls?” (spiritual insides).

When the child has answered all four questions in that block, the researcher proceeds directly to the next block of the task, beginning with the introduction of the next two social groups.

Adaptation for samples in Lebanon: The final set of reference and non-reference religious groups featured in the Social Essentialism task for the five samples from Lebanon is presented in Table 3.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 3. Sample-specific item selection for the Social Essentialism task, for the six religions included in samples from Lebanon.

https://doi.org/10.1371/journal.pone.0330727.t003

Planned analyses: For single-time point (cross-sectional) analyses, see the planned analyses in [1]. We note that, just as item-level analyses will be critical for interpreting response patterns in Wave 1, item-level analyses of Wave 2 will be particularly important for understanding both how children reason about social categories and how children interpret the particular questions posed to them in this version of the protocol.

Additional single-time point analyses will focus on children’s open-ended and forced-choice responses to follow-up questions regarding the transformation item (“Is it possible for [non-reference group] to become [reference group]?”). In particular, we plan to code children’s open-ended responses for spontaneous intrinsic versus extrinsic explanations related to why it was, or was not, possible to change one’s social category membership [44–47]. See [48] for a possible coding scheme.

Across the network, there is widespread interest in social essentialism as both a predictor of beliefs and behaviors and as an outcome of interest in its own right. Multi-time point analyses might include examinations of relationships between responses to the Wave 1 versus Wave 2 versions of the Social Essentialism task, as well as examinations of the degree to which individual-level changes in social essentialism across timepoints might explain changes in responses to other parts of our protocols. We note, however, that within-subjects analyses of individual items will only be possible for the subset of social groups included at both waves for any given cultural-religious sample, and to the two questions included in both versions of this task (spiritual insides and transformation). Likewise, any within-subjects analyses of composite social essentialism scores (combining responses across items) will need to take into account the differences in tasks across waves.

Further analyses might include assessments of the validity of each version of the Social Essentialism task in a given cultural-religious setting; comparisons across cultural-religious samples; and correlations between Social Essentialism and other tasks in the Child Protocol.

Norm Violations task.

Rationale for retention and revision of task: The Wave 2 Child Protocol features a slightly modified version of the Norms Violation task from Wave 1. To reduce the overall length, the Wave 2 Child Protocol does not include the familiarity assessment portion of the Wave 1 Norm Violations task. The Wave 2 Child Protocol additionally no longer includes severity ratings for the questions about alterability and outgroup members, and the conventional norm used across samples is changed from the norm used in Wave 1. Including this task with only minor changes in both waves allows us to examine the development of children’s reasoning about violations of religious, moral, and conventional norms from a within-subjects perspective.

Task design and procedure: The researcher provides the same introduction to this task as used in Wave 1 [see 1] and provides the child with two practice trials to familiarize children with the task; one practice trial focuses on an action that is clearly impermissible (insulting someone), and the other focuses on an action that is clearly permissible (giving someone a present). As in Wave 1, the gender of the characters featured in this task are matched to the participating child’s gender; if the participating child does not identify as a boy or a girl, or if the participating child’s gender is unknown, the child is shown the “woman” version of the task.

The task consists of four blocks of questions presented in a random order, each block centered on the violation of a particular norm (two religious norms, one moral norm, and one conventional norm; see Table 4 for norms selected for each sample in Lebanon).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 4. Sample-specific item selection for the Norm Violations task, for the five samples from Lebanon.

https://doi.org/10.1371/journal.pone.0330727.t004

At the beginning of each block, the child is presented with a grayscale silhouette of character, described as an adult member of the target religious group who violates the norm. For religiously affiliated children, this adult is described as a member of their religious group; see further details about target religious groups under “Child as Transmission task,” above.

For each sample, the Wave 2 Child Protocol features the same two religious norms and the same moral norm used in Wave 1.

The conventional norm employed in Wave 1—wearing socks on one’s hands—was flagged by network members as having direct negative consequences for the person who engages in the norm violation (e.g., making it harder for them to grasp objects). In Wave 2, we replace this with a new, standardized conventional norm that better aligns with standard definitions of “conventions” in the psychological literature [49,50]: either “wears pajamas [to the office] to work” or “wears a swimsuit [to the office] to work” (the phrase “to the office” is included in some but not all adaptations of this task across the network). This decision emerges from an iterative process of development within and across research teams, with particular attention to the idea that following or breaking the norm should not have intrinsic or direct consequences for the welfare or goals of the person who violates the norm or the people around him or her; additional considerations include choosing a conventional norm that is widely followed by members of the local community, with which children are familiar, and which is not rooted in religion.

Within each block of questions, Wave 2 features the same fixed order questions used in Wave 1. The child first judges and explains the permissibility of a norm violation by a religious group member (e.g., “This is a woman named Fatima. Fatima is Shia. Fatima prays without doing wudu’ (ablution). Is that okay or not okay? Why is that okay/not okay?”). Then the child rates the severity of the norm violation (“And how good/bad is it that...? Not good/bad, just a little bit good/bad, or very good/bad?”) and assesses the alterability of the norm by a religious authority (e.g., “What if God said that it was okay to [violate norm] – then would it be okay...?”). Finally, the child is asked to judge the permissibility of the norm violation if it were committed by someone who is not a member of the religious group.

These questions are identical to Wave 1; the only modification is the omission of two severity ratings questions within each block (after the question about alterability by a religious authority, and after the question about an non-group member).

Adaptation for samples in Lebanon: The final set of properties featured in the Norm Violations task for the five samples from Lebanon is presented in Table 4.

Planned analyses: For single-time point (cross-sectional) analyses, see the planned analyses in [1]. Note that because the Wave 2 Child Protocol no longer includes severity ratings for the questions about alterability and outgroup members, analyses including these questions cannot be conducted with Wave 2 data.

Multi-time point analyses might include examinations of relationships between responses to the Wave 1 versus Wave 2 versions of the Norm Violations task, as well as examinations of the degree to which individual-level changes across timepoints might be related to responses to other parts of our protocols. We note that exact within-subjects comparisons will only be possible for the subset of norms included at both waves for any given cultural-religious sample. We acknowledge that, although this is not the ideal approach for formal longitudinal analyses, if the Wave 2 conventional norm functions similarly to that in Wave 1–a reasonable assumption in the absence of theoretical reasons to expect otherwise– it may still be conceptually informative to include both waves of conventional norms in analyses when examining patterns of norm violation judgments. Likewise, any within-subjects analyses of composite scores derived from this task (i.e., attempts to combine responses across questions) will need to take into account the difference in conventional norm across waves.

Further analyses might include comparisons across cultural-religious samples, focusing on how religious affiliation versus other social-cultural factors might shed light on any observed similarities and differences across samples for other tasks in the Child Protocol.

Direct questions about religion.

Rationale for retention and revision of task: The final element of the Wave 2 Child Protocol is a set of questions designed to directly assess children’s explicit understanding of religion and their own religious identity [see, e.g., 51–53]. Some of these questions are retained without changes from Wave 1; the purpose of including similar questions in both waves of data collection is to observe individual differences in children’s understanding of religion and their own identity. To further explore the development of religious identity, two new sets of questions are included in the Wave 2 Child Protocol; one focuses on the child’s definition and identity of the child’s religious ingroup, and the other focuses on the child’s awareness of and participation in religious rituals.

Task design and procedure: The child is presented with the following set of questions in a fixed order.

First, the researcher asks three open-ended questions about religion, taken directly from Wave 1: 1) “When I say the word ‘religion,’ what do you think of? Can you explain what ‘religion’ means?”; 2) “There are lots of different religions in the world. Can you tell me all the religions you’ve heard of?”; and 3) “What about your family: Does your family belong to a certain religion?” [or, in some cases, “Does your family believe in a certain religion?”; see 1].

Then the researcher proceeds to a series of new questions, not included in Wave 1. The first set of questions all focus on a religious “ingroup” relevant to the child (e.g., “Maronite” for Maronite Catholic children). For sites that include religiously unaffiliated children (e.g., England, the United States), their “ingroup” for this question is “not religious.” The researcher begins by saying, “Okay. I know some people around here are [religious ingroup, e.g., ‘Maronite’]” and asks the following three questions [based on 51]: “What does it mean to be [religious ingroup]?”; “How do you know if someone is [religious ingroup]?”; and “How does a person become [religious ingroup]?”

The second set of questions focus on one important or salient religious ritual relevant to either the child’s own religious ingroup or to a prominent religious group in the child’s local setting. The researcher introduces this ritual by saying, “I’ve heard that one thing that [religious ingroup, e.g., ‘Maronite’] people do is [perform religious ritual, e.g., ‘take first communion’]. What does it mean to [perform religious ritual]?” If the child is unfamiliar with the ritual, the researcher does not ask any further questions. If the child provides a response, regardless of accuracy, they are asked four additional questions about the ritual: 1) “What kinds of things do people do when they [perform religious ritual]?”; 2) “Why do they [perform religious ritual] that way?”; 3) “Could someone [perform religious ritual] in a different way?”; and 4) “Have you ever [performed religious ritual]?” Religious rituals vary across sites and samples. Research teams are instructed to select a ritual that would be familiar to children, and, where possible, to select a ritual that is in some way related to growing up or becoming a more full-fledged member of the religious group (e.g., a coming of age ritual or initiation). Priority is given to rituals that children participate in themselves or observe at a certain age.

As in Wave 1, at the end of the interview the child is asked, “Is there anything else you want me to know about you and your family?”

Adaptation for samples in Lebanon: The final set of religious rituals featured in the Wave 2 Direct Questions about Religion task for the five samples from Lebanon is presented in Table 5.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 5. Sample-specific item selection for the Direct Questions about Religion task, for the five samples from Lebanon. See main text for descriptions of network-wide guidelines for item selection.

https://doi.org/10.1371/journal.pone.0330727.t005

Planned analyses: Analyses of the open-ended questions retained from Wave 1 (e.g., “When I say the word ‘religion,’ what do you think of? Can you explain what ‘religion’ means?”) will be qualitative, using the themes and codes developed from children’s responses in Wave 1 (not yet developed as of the submission of this manuscript).

We anticipate researchers will apply both cross-sectional and within-subjects analysis techniques to the data from the questions that have been retained across waves. Cross-sectional analyses include comparisons of samples from different cultural-religious groups, to further our understanding how children define religion and come to self-identify as members of religious groups. Single-timepoint (cross-sectional) analyses include individual differences in children’s understanding of what religion is, their awareness of specific religions, and their own sense of religious identity.

For the new questions added to the Wave 2 protocol, we predict that children who identify themselves as part of a religious ingroup will provide richer and more “accurate” (i.e., more aligned to the cultural setting) responses to questions about the religious ingroup and the group-relevant religious ritual. Analyses of these hypotheses will be cross-sectional.

Demographics and Child Religion.

Task design and procedure: In Wave 2, parents/caregivers are asked to indicate their relationship to the child (as in Wave 1). Then they are asked a short series of questions to update the demographic and religious information collected about their child during Wave 1, including the child’s current age and current grade in school, as well as the child’s educational experiences to date; parents/caregivers indicate whether the child has ever been homeschooled (religiously or secularly), has ever attended a secular or religious school, or has ever participated in other forms of religious or spiritual education. Parents/caregivers are also asked to indicate the frequency of their child’s involvement in religious and spiritual activities and practices over the past year, as well as the frequency with which their child attended religious services over the past year.

Adaptation for samples in Lebanon: This task does not require any sample-specific customization, beyond translation.

Planned analyses: As our focus in this paper is to lay out the methods and analyses for the Wave 2 Child Protocol, we do not detail planned analyses for the Parent/Caregiver Survey here.

Parental Ethnotheories Measures (optional).

Task design and procedure: Research teams are given the option to include a Parental Ethnotheories measure in the Wave 2 Parent/Caregiver Survey, adapted from the measure developed in Wave 1, which was designed to document parents’/caregivers’ beliefs about how, where, and from whom children should learn about religion and social issues [1]. The Wave 2 adaptation uses the same questions, adapted to document beliefs about how, where, and from whom children should learn about medical and health-related issues. Given the timing of the administration of Wave 2 in relation to the COVID-19 pandemic, parents/caregivers are explicitly asked to respond to the questions as though there were no restrictions related to COVID in place.

Parents/caregivers are first asked to rate their agreement with 14 statements reflecting beliefs about how, where, and from whom children should learn about medical and health-related issues (e.g., “It is important for my child to read and understand books about medical and health-related issues”; “My child should ask questions about medical and health-related issues.”). Questions are presented in a random order and parents/caregivers respond on an 11-point scale, from “I do not agree with this at all” to “I completely agree with this.”

Some of these questions feature follow-up questions about ethnotheories of development in the medical and health-related domain; for example, after rating their agreement with the statement “My child should only learn about medical- and health-related issues in our home,” parents/caregivers are then asked to select any other locations where it would be okay for children to learn about these issues (e.g., at school, at their friend’s house, at a religious site, through a TV show).

After completing this set of questions, parents/caregivers are asked to select the age at which they believe children should know about medical and health-related issues (with ages provided in two-year increments from 0–2 years to 18 years and older; “never” is also included as an option); to report the age at which they believe children tend to start asking questions about medical and health related issues (with the same response options); and to describe the specific medical or health-related issues they had in mind when answering these questions (open-ended). Parents/caregivers are also given the opportunity to provide any additional information about their views.

Adaptation for samples in Lebanon: This task does not require any sample-specific customization, beyond translation.

Planned analyses: As our focus in this paper is to lay out the methods and analyses for the Wave 2 Child Protocol, we do not detail the planned analyses for the Parent/Caregiver Survey here.

Open-ended parent questions.

The following section of the Wave 2 Parent/Caregiver Survey include a set of open-ended questions to obtain additional information about the parent/caregiver’s and child’s experiences with religion since the first time they participated.

Task design and procedure: The Wave 2 Parent/Caregiver Survey includes a set of questions designed to give parents/caregivers an opportunity to reflect more freely on their child’s religious development (and, if included in the sample, on their own ethnotheories about religion and medical and health-related issues). Research teams are given the option to include these open-ended questions at the end of their parent/caregiver survey for parents to complete on their own or as a qualitative oral interview done with a researcher where responses are recorded and transcribed.

The first three questions ask parents/caregivers to reflect on what might have changed since the previous wave of data collection: 1) “Some time has passed since you first participated in this study. Has anything changed in how your child learns about religion?”; 2) “Now let’s talk about religious activities your child participates in. Has anything changed about the activities related to religion that your child is involved in in school, at home or in the community, more generally?”; and 3) “Now let’s talk about you. Has anything changed in the ways in which you talk, teach, or answer questions about religion with your child?” In all cases, parents are asked to describe these changes.

Parents/caregivers are then asked to describe whether they attribute changes in their approach to their child’s religious development to a number of specific factors, including: “My own views about what to discuss about religion with my child have changed”; “My own religious beliefs have changed”; “Participating in this study last year had an impact on how we approach religion in our home (e.g., child has been asking more questions about religion; my child has been more interested in participating in religious activities)”; “My child is older now so new approaches are needed”; “The COVID-19 situation has changed since last year and this has had an impact on my family’s approach to religion.” Parents/caregivers are also given the option to report some other reason for changes in their approach, to report no changes in their approach, or to decline to respond to this question.

The next two questions ask parents/caregivers to reflect on the specific material and social sources that their child might use to learn about religion.

First, parents/caregivers are asked, “In many places, children learn about religion through prayers, books, stories, songs, videos, or other activities. Thinking about your child, please tell us about one or two examples that your child has been exposed to at home or at school that they spend a lot of time with or engage with regularly. Is there a particular book, story, song, video, or activity that you like for helping teach your child about religion in general or the beliefs and practices that are valued in your family? What makes you like it? How do you think this has taught your child about these ideas and values?” In addition to eliciting reflections on the specific circumstances of children in this study, this question is designed to lay the foundation for compiling a corpus of religious materials for future analysis.

Second, parents/caregivers are asked, “Thinking about your child, is there a particular person or group of people who have impacted your child’s learning about religion?”

Finally, for any samples that opts to include the Parental Ethnotheories measure (described in the previous section), a final open-ended question asks parents/caregivers to describe their feelings about when and how children should learn about medical- and health-related issues.

Planned analyses: As our focus in this paper is to lay out the methods and analyses for the Wave 2 Child Protocol, we do not detail the planned analyses for the Parent/Caregiver Survey here.

Data management plans.

Data from the Child Protocol and Parent/Caregiver Survey are collected through Qualtrics survey software. When authorized, audio and/or video recordings of the child interview are collected either using a portable recording device or using tele-conferencing software (e.g., Zoom, Microsoft Teams). These recordings are used to help transcribe responses to open-ended questions in the child interview.

Once datasets have been collected, data cleaning will be performed by each team. Data from participants who authorize that their de-identified responses be shared publicly will be shared on platforms for open- science, including the Open Science Framework (https://osf.io/). Likewise, audio/video recordings from participants who authorize sharing these recordings will be shared with the research community on Databrary (https://www.databrary.org/).

Once all datasets from the second wave of data collection have been cleaned and de-identified, the current members of the DBN will have exclusive access to the full dataset for a finite interval of time. Following that interval, the data will then be shared publicly as described above.

Ethical considerations and declarations.

This study has been approved by the Institutional Review Board for Socio-Behavioral research (IRB-SB) at the University of California, Riverside, under protocol #HS-21–124;, and by the Institutional Review Board (Charles River IRB Office) at Boston University under protocol #4631E. Together, these IRBs cover data collection for all US samples as well as the samples in Mexico and Taiwan R.O.C. Data collection in Lebanon has been ethically reviewed and approved by the Institutional Review Board for Social and Behavioral Sciences at the American University of Beirut, Beirut, Lebanon under protocol #SBS-2021–0343. Parents/caregivers give informed consent either in writing (on paper or via an online survey) or verbally via a live conversation with the researcher (in person or over video chat); and children give oral assent. The mode of consent varies across these samples depending on the modality of participation (online vs. in person) and on the degree to which parents/caregivers are comfortable reading written text. Additional information regarding the ethical, cultural, and scientific considerations specific to inclusivity in global research is included in the Supporting Information.

All other planned samples are covered by ethics boards at the home institutions of their respective research team leaders (who are independent subaward PIs). Research team leaders are also responsible for securing approval from any additional community organizations, school boards, or governmental bodies overseeing research in their field sites and samples.

Status and timeline.

As this is the second wave of a multi-time point study, recruitment for the study is complete in all of the samples that completed Wave 1 of this study, although we anticipate continuing to add samples to this data collection effort in the future. All participants are contacted nine to 15 months after their first participation at Wave 1 to participate again at Wave 2. At the time of this submission, data collection is underway in 40 cultural-religious samples, and is complete in 21 of these samples. One additional sample is set to begin data collection in Summer 2025. We anticipate data collection for all of the 41 samples described in the current manuscript to be complete by January 31, 2026. We expect to share results from this study approximately 18 months after data collection is complete.