Cultivating cross-cultural collaborations: perspectives from the Developing Belief Network

Developmental science has long highlighted the limitations in generalizability arising from the fact that the sizable majority of the populations we study come from a minority of locations globally. Furthermore, team science has lauded the opportunities presented by collaborative approaches to science, while open science approaches highlight the need to ensure that methods and data adhere to principles that are FAIR (findable, accessible, interoperable, and reusable). Challenges to employing both a global sampling strategy and a team/open science approach, while centering cultural influences and diversity in early development, include challenges designing protocols that provide both standardization and cultural sensitivity, working with and interpreting data from unique samples, structural challenges focused on meeting the requirements needed for tenure and promotion, and ethical challenges with making data from minors fully accessible. Here, we present lessons we have learned from cultivating the Developing Belief Network, outlining our solutions toward mitigating these challenges. We highlight our approach to constructing and sustaining the DBN, as well as recommendations we have learned from the 5 years since its inception.

Introduction

As the PIs and Directors of the Developing Belief Network (DBN; #JTF-61542), our goal has been to advance a global, culturally-centered science of child development that seeks to understand the diversity of developmental contexts. At its core, the DBN is deeply committed to motivating diverse and inclusive research of developing children that centers and focuses on the contexts and concerns of children and families. We are not the first researchers to critique developmental science for a focus on participants that are culturally Western, higher income, and highly educated (e.g., Henrich et al., 2010; Nielsen et al., 2017). Nor are we the first to highlight the benefits of cross-cultural and international collaboration for developmental science (e.g., Broesch et al., 2023; Lansford et al., 2019; Wen et al., 2025). Indeed, our research agenda and processes have been inspired and informed by these critiques and recommendations. Topically, we aimed to highlight the importance of studying religion in development, namely, how religious beliefs are acquired and how they influence children's lives. Methodologically, we aimed to build the network in the spirit of collaboration, team science, and open science.

Building from these foundations, leading the DBN has convinced us that developmental science can only progress through genuine engagement with the diversity of children's lived experiences, accomplished through active and intentional collaborations with fellow scholars and global community members. Here, we describe forming the DBN and our approach to collaboratively sustaining and expanding its reach. Elsewhere, we have outlined our theoretical approach (Lesage et al., 2023; Richert and Corriveau, 2022; Richert et al., 2022, forthcoming), as well as our collaborative protocol (Weisman et al., 2024) and process for adapting the protocol globally and longitudinally (Williams-Gant et al., 2025). In this perspective piece, we reflect on our successes and challenges and provide recommendations we hope will benefit research teams interested in engaging in collaborative, global developmental science.

Building the developing belief network

To launch the DBN, we received a large, 5-year research grant from the John Templeton Foundation (JTF; 6/1/2020-5/31/2025). Funding was secured through an off-cycle, invited submission after several years of planning and consultation with a dedicated JTF program officer and numerous advisors. For those unfamiliar with securing funding of this nature through a private foundation, the JTF has a wide range of funding priorities, one of which broadly focuses on the topic of religion. Each of us had previously received funding for studies into the development of religious cognition over the course of a decade prior to initiating discussions about the benefit of developing and launching this type of project for the field with a program officer with whom we had previously worked.

Separately, we argued that the field of religious belief development was profoundly limited because the majority of research was focused on Abrahamic religions, which make up only about one third of the global religious population (Pew Research Center, 2025), and the majority of developmental research focuses on Westernized samples making up just about 13% of the world's population (Nielsen et al., 2017). Accordingly, it is extremely challenging to make any inferences about global generalizability and variability by sampling such a small subset of the global population over and over again. Our goal was to broaden understanding of the roles of specific cultural and religious beliefs and practices in children's lives, while maintaining validity within and across samples by employing the same types of measures throughout.

We should note our gratitude for the enormous faith JTF placed in us as directors of this endeavor. At the time discussions of this project began, we were both tenured Associate Professors; and we both advanced to Full Professor over the course of running this project. Although we had both worked with JTF separately, we had never collaborated, and so greatly benefitted from the 2-year planning grant offered by JTF to learn about each other's leadership style, to consult with various advisory boards, and to write the initial draft of the grant. This is a luxury that is not often afforded in federal funding schemes.

Once the funding was approved, our project went into high gear through the circulation of a Request for Proposals to participate in the DBN on varying listservs, rather than limiting the opportunity to researchers we already knew might be interested in such an endeavor. Reaching out only to our professional and personal connections would have constrained the scope of the research questions and methods, not to mention the participant populations. We asked interested teams to be willing to engage in protocol development and commit to using that protocol, demonstrate the ability to secure IRB approval and interview 4- to 10-year-old children and their caregivers at a research field site with a diversity of religious groups, and commit to interviewing the same children and their caregivers at three time-points. Teams also committed to being actively engaged in developing practices that codified our collective team science and open science approach.

Our Advisory Board helped with selection; and as of August 2025, the DBN incorporates 2 Lead PIs/Directors, 32 additional Fieldsite PIs and co-PIs, 19 current or former postdoctoral scholars, 25 graduate students, 7 Advisory Board members, and over 100 additional researchers (e.g., data collectors, translators; more than 200 scientists). Accordingly, a key challenge has been to document, track, and acknowledge the contributions of all researchers. To that end, all individuals involved at varying stages of our collaboration are documented in a project-specific version of the CRediT system (see https://osf.io/xu6ma).

Our research examines how children develop concepts of religious agents and what they understand about religious group membership (e.g., the nature of groups, norms) with a focus on the socialization and transmission of religious beliefs. As of August 2025, we have worked with 47 unique religious groups. Figures 1, 2 present the sample breakdowns for over 2,500 caregiver-child dyads for all Wave 1 samples plus 5 samples that joined at Wave 2 (Christian and Unaffiliated in Australia, Buddhist in Indonesia, Buddhist and Malay Christian in Malaysia). Samples vary in the extent to which children have contact with members of other religious groups (and whether that contact is highly polarized, e.g., Northern Ireland, or coexists peacefully, e.g., Singapore), whether religious groups are locally in the minority or majority, and whether there is violence associated with group membership. Research fieldsites additionally vary across other socio-cultural factors (e.g., urban vs. rural, race/ethnicity, high vs. low cultural diversity, exposure to political conflict, socioeconomic and educational dimensions).

Figure 1

Figure 1. Overview of developing belief network participants broken down by superordinate religious group and UN Geoscheme subregion.

Figure 2

Figure 2. Overview of developing belief network participants broken down by sample.

A key success is that we have developed a rich protocol and resulting datasets that can be used to answer innumerable questions focusing on how religious cognition is acquired, transmitted, and impacts social group dynamics. Our interviews and surveys have been translated into 15 written languages (Arabic, Traditional and Simplified Chinese, English, Greek, Hebrew, Hindi, Indonesian, Malaysian, Rutooro, Sepedi, Spanish, Tsimané, Xitsonga, and Yucatec Mayan). For each child-caregiver dyad, we have a substantial amount of raw data, including video or audio recordings of data collection and task performance, video or audio recordings of caregiver-child conversations, interview transcripts (in most cases, in more than one language), quantitative and qualitative survey responses, and field-specific notes and contextual information. Considering language, fieldsite, wave, and measure, we are currently working to make over 300 datasets (and their associated protocol versions) clean, findable, and usable to the global research community.

Recommendations based on our experience with the network

To support scholars interested in engaging in similar endeavors, we describe (a) how we approached the development of culturally and developmentally relevant research questions and protocols, which required (b) fostering trust in a genuinely collaborative team of researchers, (c) approaching leadership as collaborators (not dictators) in the process, and (d) committing to open and transparent research practices. We conclude by reflecting on the impact of our approach on junior scholars embarking on this kind of research.

Developing culturally relevant and developmentally appropriate protocols

Our primary research goal was to center cultural influences on early development. For the project to be ecologically valid and findings to be meaningful, we needed to consider if the research questions, as well as the methods, were relevant for children within each fieldsite. When we submitted our original proposal to JTF, we focused the research questions around our primary areas of relevant expertise: religious concept development and social transmission. Reflecting the team science goals of our project, JTF was open to the likelihood that our research questions would be updated and adapted once the members of the DBN had been determined. As a team, we developed research questions that broadly address three categories:

Religious agents

How do folk theories (psychology, physics, biology, sociology) support and constrain concepts of and beliefs in religious agents? How does exposure to religious and supernatural beliefs influence the development of folk theories?

Religious identity

How do children develop an understanding of religion as a social category? When and under what conditions do children essentialize religious group membership? When and under what conditions do children treat religious norms similarly or different from other kinds of norms (moral, conventional)?

Social transmission

How are religious and supernatural beliefs (including both formal religions and more traditional beliefs/practices) transmitted? What are the ways in which caregivers communicate (directly or indirectly) with their children about religious beliefs, behaviors, and expectations?

To address these research questions, DBN members worked together to develop a collaborative protocol that would be used with all samples. In order to ensure each protocol version was culturally-appropriate and relevant for each caregiver-child sample, all versions of the protocols were developed and finalized using the following steps:

1. Research team members who are experts in each topic divided into working groups to identify existing tasks that could address the relevant research question.

2. Existing tasks were evaluated and updated; new tasks were designed if existing tasks did not meet project needs or criteria.

3. The working group proposed tasks to the full network of researchers, who provided feedback.

4. The full network worked together iteratively until all network members gave preliminary approval for the tasks.

5. The tasks were presented through Semi-Structured Interviews to 2–3 educators, leaders, or caregivers for each fieldsite sample. Community members were asked (1) if the topics/questions were appropriate and relevant for children in their community, and (2) to provide examples of sample-specific religious beliefs and practices (e.g., how to refer to “God”, what religious norms are most familiar to children).

6. Feedback from the Semi-Structured Interviews was compiled; full task structures were revised when relevant and individual protocols were developed with sample-specific stimuli for each fieldsite sample.

7. Protocols were translated and back-translated.

8. Protocols were piloted with children spanning the planned age-range within each fieldsite sample.

9. Minor revisions to protocols were coordinated across network teams as necessary.

10. Finalized protocols were submitted to Institutional Review Boards (IRB) for approval.

11. Once IRB approval was secured, a version of each fieldsite sample protocol for both children and caregivers was created in Qualtrics by post-doctoral researchers funded as core team staff.

12. Once Qualtrics versions were ready for deployment, data collection began in that fieldsite or with the sample.

This process required substantial time investment from all DBN researchers. We provide an overview of our timeline of activities in Supplementary Figure 1.

In Fall 2020, we hosted a series of virtual workshops in which researchers brainstormed updates and edits to the research questions, with a final set of research questions approved by all team members by the end of February 2021. Subsequently, we hosted workshops to finalize and design tasks; and then teams spent an additional 2 to 6 months creating sample-specific versions of the protocol. Weisman et al. (2024) provides a thorough description of the Wave 1 protocol used in the United States (along with preliminary hypotheses and planned analyses); and Williams-Gant et al. (2025) describes the processes for sample-specific adaptations and revisions made for Waves 2 and 3 (again with preliminary hypotheses and planned analyses).

The year and a half devoted to this process was invaluable for ensuring our research protocols were appropriate and relevant across the varied contexts while also meeting our shared research goals and providing data that could be combined across fieldsites and samples. The interviews with local religious leaders and experts in each fieldsite also served an invaluable function of maintaining ongoing partnerships between the researchers and the local community. These interviews helped research teams establish community trust by listening to the ideas and concerns raised and demonstrating our openness to adapting and updating our tasks to ensure they conformed to community-specific values and minimize the fears and concerns community members may have about researchers asking their children about highly personal religious beliefs and experiences. For example, one task that where we had thought would not pose problems—but did pose many—was a backwards digit span task (Wechsler, 1974). This type of task is used to test verbal working memory, and in it, children hear a series of digits and then are invited to repeat those digits but in the reverse order than they had heard them. Because of differences in the developmental trajectory of the acquisition of numbers across our fieldsites, we did this task with animals and numbers. However, what we had not anticipated may have been obvious to the reader: when these items are translated to different languages: their lexical difficulty is not maintained, nor are the number of their syllables. Accordingly, when children are asked to hold these items in working memory, this introduces a confound across fieldsites! Therefore, every task required careful attention to all of these details while being developed.

Allowing time to build trust in the team

This kind of endeavor requires extensive trust between team members. If feasible, we recommend building in time for the team to collaboratively determine shared goals and processes for achieving them. As described above, DBN members began working together in the Fall of 2020. Fieldsite PIs spent the first 18 months revising and finalizing research questions and designing a shared protocol. We believe this time was crucial for setting up a healthy trust in each other and success for the next years of the grant. Indeed, all fieldsite PIs have indicated an eagerness to continue to collaborate—even after the grant officially ended! Moreover, this established trust was critical, given the varied sociopolitical situations we faced throughout the 5 years of our grant. For example, COVID affected all fieldsites starting in 2020, which required some fieldsites to pivot to remote protocol development and other fieldsites (where remote data collection was unfeasible due to limited Wifi access or technological fluency) to fully pause in piloting and data collection. Moreover, some fieldsites faced major economic (e.g., Lebanon), political (e.g., Peru, US), and natural disasters (e.g., UK, India, Indonesia, Mexico, Uganda), and geopolitical conflict (e.g., India, Israel, Lebanon, US, Mexico). Having taken the time to establish trust in each other was invaluable to supporting each other through these challenges and to keep the project moving forward.

A key area in which trust in each other has been critical is in determining processes for sharing and disseminating our findings. We operate in a career structure and field that primarily rewards first and lead author contributions to projects, which can be in direct conflict with the more supportive contributions needed to keep large-scale, collaborative projects moving. Our approach to resolving these tensions was to discuss from the beginning how to determine authorship and crediting for various kind of contributions and outputs. We created internal agreements in a set of Authorship Guidelines (https://doi.org/10.31219/osf.io/rwhjd_v2), and we created our own categories for crediting contributions based on the CrediT system (https://osf.io/xu6ma/files/83psw). Finally, in order to facilitate data sharing, each fieldsite committed to making their data available (to the extent that their participants agreed) on OSF or Databrary.

Collaborative leadership and decision-making

As leaders of the DBN, we aimed to foster genuine collaboration and shared ownership through employing the principles of team science. From the beginning, as PIs, our grant submission included our vision for research questions and methods, but also indicated an explicit understanding that all research questions and methods would be finalized in collaboration with the Fieldsite PIs and co-PIs who joined the DBN. This intentionally collaborative approach is in contrast to a more traditional top-down approach in which the finalized protocol is developed by the grant PIs, and fieldsite PIs are given a fully-baked protocol to translate, backtranslate, and deploy. We approached decisions throughout the grant in similar ways, including authorship practices (including CRediT designations) and protocol updates for future waves.

Commitment to open science

Although our initial grant submission highlighted our commitment to open science, it was clear early on that we had not fully considered the varying challenges of engaging in open science in this endeavor. We continue to grapple with ethical considerations about balancing the privacy and safety of our participants with the open science and transparency goals of data sharing. One such challenge has involved variability in the willingness of caregivers and children to share their data, with different options for internal data sharing within the DBN vs. public data sharing. This has meant that we are simultaneously developing datasets with different participants (e.g., some who agreed to Databrary but not OSF), which has made data cleaning extremely labor intensive. Furthermore, we continue to consider the ethical challenges in publicly sharing data including minors—even if they are willing to do so. One solution is that because it cannot be deidentified, all video data will only be available on Databrary (which incorporates layers of privacy protections) if a participant is willing. Future work should explore the extent to which the missingness in a fully open dataset represents systematic differences in families who are willing to share their data publicly vs. those who are not. Moreover, we are sensitive to the fact that many of these datasets include cultural groups, environments, and unique historical contexts that may not be front-of-mind to the future researchers who mine them. We aim to ensure that scholars, both those in the DBN and those with whom data were shared, understand the communities well enough to appropriately interpret the data once it is made publicly available. As such, we are curating PI-developed fieldnotes that will be published along with all datasets when they are made publicly available.

In considering the transparency of our protocol design and planned analyses, we additionally faced challenges balancing the primarily exploratory nature of our cross-cultural research with the desire to be open and transparent about our hypotheses and methodological approaches (e.g., through preregistrations). We resolved this tension by publishing two manuscripts that include a description of our reasoning and approach to task design and the process of engaging in cross-cultural updating and validation (Weisman et al., 2024; Williams-Gant et al., 2025). These manuscripts additionally describe preliminary planned analyses that will be conducted with each task. We then have individual teams of researchers publish specific preregistrations for planned analyses prior to initiating analysis of data for conference presentations or manuscripts.

Growing and diversifying the field

As noted above, we contend that engaging in responsible cross-cultural research requires an investment in truly collaborative, team science practices, which takes time. As mentors who aim to grow and diversify the field of researchers engaging in these kinds of research programs, we intentionally included and supported many emerging and junior scholars in this project. Along the way, it became apparent that our more time-consuming collaborative process, as well as the longitudinal aspect of our design, are challenging for junior scholars because it will take several years to produce the kinds of empirical outputs that lead to academic jobs and promotions. As such, we have considered ways to support junior scholars individually, and to use our collective voice to support the field. Within our network, we strongly encourage junior scholars to serve as lead authors on manuscripts, which they indicate to other network members by using an “Intention to Publish” form that is circulated to all network authors. This opt-in method makes all DBN members aware of the publication effort, and provides an open invitation for them to participate and support the lead author(s). We also have encouraged junior scholars to write manuscripts about our process, to document our methods, and to highlight our global approach to protocol development, leading to multiple manuscripts that are either published or under review.

Outside of what we can do in the network itself, we encourage our field to consider ways to support junior scholars engaging in this kind of critical scholarship. First, many journals in developmental science still focus on including “comparison groups” in order for cross-cultural research to be published. We strongly encourage reviewers and editors to consider that this approach may unwittingly lead to deficit interpretations of findings, while also limiting the opportunities junior members of our field have to submit fieldsite data as standalone, and just as informative, scientific contributions. Second, we encourage merit and promotion processes within universities to find ways to incorporate the time investment this kind of critical research takes as a product in and of itself, even before it results in published, peer-reviewed manuscripts with citable impact factors.

Conclusion

These 5 years leading the DBN have been personally and professionally fulfilling for the both of us, and we are forever grateful for the JTF staff members, our Advisory Board, our DBN colleagues, our community partners, and our many participants who trusted us and believed that our work could make an impact. We are exceptionally proud of what the DBN has accomplished, and we look forward to sharing our findings with the developmental science community as our data become cleaned and ready for analysis. We share these reflections with the intention of supporting continued investment in global, collaborative developmental science that centers the diversity of children's lived experiences. As such, we would be remiss if we did not acknowledge that as we write this piece, research centering diversity in science is under threat and is likely to be underfunded for some significant period of time. By making our processes, as well as protocols and data, transparent, findable and usable to the community, we hope future generations of scholars can use the work we have been able to produce to continue to advance and center diversity in developmental science during these more lean periods.

Data availability statement

The original contributions presented in the study are included in the article/Supplementary material, further inquiries can be directed to the corresponding author.

Author contributions

RR: Data curation, Investigation, Methodology, Writing – original draft, Conceptualization, Resources, Writing – review & editing, Project administration, Funding acquisition, Validation, Formal analysis, Supervision. KC: Writing – review & editing, Funding acquisition, Data curation, Writing – original draft, Investigation, Resources, Conceptualization, Formal analysis, Validation, Project administration, Supervision, Methodology.

Funding

The author(s) declare that financial support was received for the research and/or publication of this article. This publication was made possible through the support of a grant (#JTF61542) from the John Templeton Foundation to Rebekah A. Richert and Kathleen H. Corriveau.

Acknowledgments

We gratefully acknowledge the contributions and commitment of all members of the Developing Belief Network (see https://osf.io/xu6ma/wiki/home/), as well as the children, caregivers, and community members who have facilitated this research. We especially thank Dr. Jenny Nissel for her feedback on this manuscript and for creating the participant figures (Figures 1, 2) and Carole Meyer-Rieth for help in creating the timeline (Supplementary Figure 1).

Conflict of interest

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Generative AI statement

The author(s) declare that no Gen AI was used in the creation of this manuscript.

Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.

Publisher's note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

Author disclaimer

The opinions expressed in this publication are those of the author(s) and do not necessarily reflect the views of the John Templeton Foundation.

Supplementary material

The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fdpys.2025.1699323/full#supplementary-material

References

Broesch, T., Lew-Levy, S., Kärtner, J., Kanngiesser, P., and Kline, M. (2023). A roadmap to doing culturally grounded developmental science. Rev. Philos. Psychol. 14, 587–609. doi: 10.1007/s13164-022-00636-y

Crossref Full Text | Google Scholar

Henrich, J., Heine, S. J., and Norenzayan, A. (2010). The weirdest people in the world? Behav. Brain Sci. 33, 61–83. doi: 10.1017/S0140525X0999152X

PubMed Abstract | Crossref Full Text | Google Scholar

Lansford, J. E., Gauvain, M., Koller, S. H., Daiute, C., Hyson, M., Motti-Stefanidi, F., et al. (2019). The importance of international collaborative research for advancing understanding of child and youth development. Int. Perspect. Psychol. Res. Pract. Consult. 8, 1–13. doi: 10.1037/ipp0000102

Crossref Full Text | Google Scholar

Lesage, K. A., Ghossainy, M. E., Richert, R. A., and Corriveau, K. H. (2023). “Cultural transmission and acquisition in the development of religious thinking,” in Oxford Handbook on Cultural Evolution, eds. R. Kendal, J. Kendal, and J. Tehrani (Oxford, UK: Oxford University Press), C50P1–C50S15.

Google Scholar

Nielsen, M., Haun, D., Kärtner, J., and Legare, C. H. (2017). The persistent sampling bias in developmental psychology: a call to action. J. Exp. Child Psychol. 162, 31–38. doi: 10.1016/j.jecp.2017.04.017

PubMed Abstract | Crossref Full Text | Google Scholar

Pew Research Center (2025). How the Global Religious Landscape Changed From 2010-2020. Available online at: https://www.pewresearch.org/religion/2025/06/09/how-the-global-religious-landscape-changed-from-2010-to-2020/ (Accessed November 12, 2025).

Google Scholar

Richert, R., Weisman, K., Lesage, K., Ghossainy, M., Reyes-Jaquez, B., and Corriveau, K. (2022). Belief, culture, development: insights from studying the development of religious beliefs and behaviors. Adv. Child Dev. Behav. 62, 127–158. doi: 10.1016/bs.acdb.2021.11.002

Crossref Full Text | Google Scholar

Richert, R. A., and Corriveau, K. (2022). Development of religious cognition. Annu. Rev. Dev. Psychol. 4, 185–206. doi: 10.1146/annurev-devpsych-120920-041303

Crossref Full Text | Google Scholar

Richert, R. A., and Corriveau, K. H. (forthcoming). “Belief in the religious supernatural: social mechanisms of belief formation transmission,” in The Oxford Handbook of the Cognitive Science of Belief, eds. N. Van Leeuwen, T. Lombrozo (Oxford, England: Oxford University Press).

Google Scholar

Wechsler, D. (1974). Wechsler Intelligence Scale for Children-Revised. New York, NY: Psychological Corporation.

Google Scholar

Weisman, K., Ghossainy, M. E., Williams, A. J., Payir, A., Lesage, K. A., Reyes-Jaquez, B., et al. (2024). The development and diversity of religious cognition and behavior: protocol for Wave 1 data collection with children and parents by the Developing Belief Network. PLOS One 19:e0292755. doi: 10.1371/journal.pone.0292755

PubMed Abstract | Crossref Full Text | Google Scholar

Wen, N. J., Amir, D., Clegg, J. M., Davis, H. E., Dutra, N. B., Kline, M. A., et al. (2025). Construct validity in cross-cultural, developmental research: challenges and strategies for improvement. Evol. Hum. Sci. 7:e17. doi: 10.1017/ehs.2025.3

PubMed Abstract | Crossref Full Text | Google Scholar

Williams-Gant, A. J., Weisman, K., Amin, T. G., Ghossainy, M. E., Soueidan, G., Nissel, J., et al. (2025). Working across religions, cultures, settings, and development: protocol for Wave 2 data collection with children and parents by the developing belief network. PLOS One 20:e0330727. doi: 10.1371/journal.pone.0330727

PubMed Abstract | Crossref Full Text | Google Scholar