Motivation for this discussion paper

The benefits of citizen science are diverse and wide ranging, and can improve how we conduct research and understand our universe. Citizen science can strengthen public involvement and scientific literacy and improve collaboration between professional scientists and the public, as well as empowering communities at greater risk of exploitation, inspiring people to study and work in science, and building public trust in science through transparency and inclusivity [29–33]. It can also improve large-scale data collection (such as improving biodiversity monitoring), foster environmental stewardship, and support improved early threat detection in various disciplines.

Creating enabling conditions for inclusive citizen science can be crucial to project success. These conditions may be developed by building people’s capacity through training and development, developing institutional partnerships, and providing infrastructure, long-term funding, and accessible technologies that facilitate data collection and analysis [9,34,35]. However, initiating and sustaining citizen science projects with inclusive practices presents a range of challenges, including ensuring data quality, retaining participants, and addressing ethical concerns like consent and data ownership [36]. Additional challenges relate to inclusivity, communication, funding, and balancing scientific rigor with meaningful public involvement [37–39].

Recognising both these opportunities and barriers, this article aims to identify key themes, challenges and solutions for advancing and improving citizen science for all involved. To do so, we used a participatory action research approach to gather perspectives and experiences from experts, practitioners in the field and the wider public. This process enabled a joint exploration of how citizen science can be strengthened to address the societal and ecological challenges of the 21st century.

Methodological framework

This study uses participatory action research design and mixed-methods analysis. Participatory action research involves multiple stakeholders in the development, implementation, and evaluation of research [40–42], aligning the research process with the substantive focus of this study, namely shared governance and co-production in citizen science. We use the term ‘stakeholder’ here to include people with personal, professional or financial interests, as well as ‘interest-holders’ such as patients and people affected by enviromental pollution and climate change [43]. The study was co-created with citizen science experts from health, environmental, and education sectors across multiple countries. Participants collaboratively designed the survey, contributed to data analysis, and co-authored the resulting recommendations. The mixed-methods approach integrated inductive thematic analysis, computational text analysis and cross-case synthesis of illustrative case studies to enable triangulation and strengthen reliability and validity [44]. We transparently document the distribution of tasks among co-authors in the Methods and a Standardised Data on Initiatives (STARDIT) report, which is an open-access data-sharing system for standardising how information about initiatives is reported [45]. The Australian Citizen Science Association endorses STARDIT as a recommended reporting tool for citizen science initiatives, supporting consistent, transparent, open access and cross-disciplinary documentation of participation, process and impact.

Data collection

Collaborative process and documentation.

The co-creation process invited people to volunteer their time, with the incentive of being a co-author on a peer-reviewed publication. Everyone who responded to the survey was invited to be a co-author at multiple stages throughout the process, including data analysis and write up. The co-production process involved 20 online meetings, shared collaborative documents and text-based discussion platforms to facilitate distributed participation across time zones. A record of all drafts, coding templates, meeting notes and analytic decisions was maintained in shared documents to create a transparent record of revisions and interpretations. These collaborative processes enabled iterative refinement of survey instruments, codebooks, and recommendations, while maintaining transparency in decision-making. The various tasks of individual contributions are documented in the associated STARDIT report. Co-authors agreed during the co-creation process that some people should be paid for some tasks. An overview of the co-production process is provided below (Fig 1). Further information is provided in the Supplementary Information and the STARDIT report about this article [49].

Download:

• PNG
  larger image
• TIFF
  original image

Fig 1. Schematic overview of co-production process.

Overview of the co-production process used in the production of this paper.

https://doi.org/10.1371/journal.pone.0331161.g001

Data analysis

Case studies and data reporting

Survey respondents provided examples of citizen science initiatives that exhibited best practices to benefit society. The co-authors conducted desktop research to gather supplementary publicly accessible information and data on these best-practice examples.

After analysis of the emerging themes from the qualitative data, the co-authors collaboratively selected citizen science initiatives that captured the breadth and variety in the material or were illustrative of certain themes. Case studies were purposively selected to reflect variation across geographic scale, disciplinary focus, technological integration, funding model, and level of participant involvement. Selection aimed to illustrate thematic breadth rather than statistical representativeness. We have included these illustrative case studies to demonstrate with real-world examples some of the themes we identified from the survey data [25]. The initiatives were analysed in more detail by extracting and structuring data about them using STARDIT data fields [45] (including aims, outcomes, data sharing and budgets), leading to the creation of publicly available STARDIT reports for each case study.

This entire process was conducted collaboratively and transparently, with co-authors verifying data accuracy. Organisations and individuals directly involved in each initiative were contacted via email and given the opportunity to review, amend, or improve or validate the respective STARDIT reports.

Integration of participant feedback

Consistent with the participatory action research method, feedback from participants directly shaped the evolution of the study. Between Round One and Round Two, survey questions were refined to clarify terminology and expand the inquiry into payment structures and task distribution. During qualitative coding, themes initially classified under “Other” were incorporated into the formal codebook following co-author discussion, altering the final thematic structure. Draft recommendations were iteratively revised in response to co-author critique to strengthen clarity around equity, funding sustainability, and data governance. Case studies were also added or refined based on participant suggestions and validation feedback from people directly involved with the initiatives. This iterative feedback process ensured that both analytic framing and final recommendations reflected collective deliberation rather than a single-author interpretation.

Ethical considerations for a co-created paper

Ethical considerations were embedded throughout the project design and implementation to support equitable participation, appropriate data stewardship, and transparent authorship. Indigenous participation was actively sought via the direct contact networks of project co-authors, funders, and collaborators, and through the broader newsletter and social media distribution channels described earlier in the manuscript, with the aim of reducing gatekeeping and ensuring opportunities to contribute were not limited to academic or institutional networks. How this project aligns with the CARE Principles for Indigenous Data Governance (collective benefit, authority to control, responsibility, ethics) [54] is reported in the associated STARDIT report [49].

All participants consented to the use of their survey responses for the purposes of this research. To protect confidentiality and minimise re-identification risk, names and other direct identifiers were redacted prior to analysis and reporting [55]. In relation to data ownership and use, participants’ contributions were treated as shared research inputs provided under explicit consent, with analysis and synthesis undertaken in a manner consistent with the project’s co-creation intent and with care taken not to attribute specific views to identifiable individuals or communities. Finally, to safeguard equity in recognition and decision-making, co-authorship was openly available to all participants in accordance with Committee on Publication Ethics (COPE) guidance, with an emphasis on transparent criteria and an inclusive pathway for contributors to opt into authorship and review how their contributions were represented. The study received ethics approval from the Norwegian Agency for Shared Services in Education and Research (reference number 107069). All participants provided written consent to participate in the project.

Demographic data

46 respondents participated in the survey (28 in Round One, 18 in Round Two), sharing a total of 16,247 words in open text fields, with an average of approximately 350 words per respondent. Most respondents lived in Australia (80%), two (4%) of whom were Aboriginal or Torres Strait Islander. The majority of respondents considered themselves to be ‘academic’ and ‘have experience with academic writing’ (76%), including professional researchers paid to do research (65%). Most respondents also reported that they had experience as a volunteer researcher contributing to citizen science (65%). An overview of the demographics and previous expertise of participants is provided in S1 Table in the supplementary materials. All 46 respondents provided an email address and 34 answered that they wanted to contribute as a co-author. A total of 29 people completed the final feedback on the experience of being involved. A total of 18 respondents co-authored the final version of this paper (in line with the Committee on Publication Ethics Authorship guidelines [46]), and 6 asked to be acknowledged for their support. In total, people estimated that they volunteered 1,385 hours on this project.

Data on payment for expertise in citizen science projects

Round Two gathered data on tasks in citizen science projects, and 17 out of 18 respondents provided answers about whether they have completed paid or unpaid work in citizen science, as well as which tasks they think should be paid (Fig 2a). Most participants had experience contributing to academic publications, dissemination including conference presentations, and funding applications, and these tasks were predominantly performed without payment (Fig 2a). The remaining seven tasks were more commonly paid for, with the highest proportion of paid positions observed in project management, planning and design (Fig 2a). When asked about which tasks should be paid for, respondents showed the strongest support for data analysis, project management, and ethics applications (Fig 2b). While data collection was one of the most commonly performed paid and unpaid activities, there was relatively less support for making this a paid task (Fig 2b).

Download:

• PNG
  larger image
• TIFF
  original image

Fig 2. Results from Round Two survey on respondents’ experience and opinions regarding paid tasks in citizen science projects.

(a) Tasks previously performed by respondents in citizen science projects, indicating whether they were undertaken as paid or unpaid work (multiple responses allowed per task). (b) Respondents’ opinions on which tasks in citizen science projects should be financially compensated, reflecting perceptions of task value, effort, or required expertise.

https://doi.org/10.1371/journal.pone.0331161.g002

Main themes from data

Eleven main themes with at least 15 responses per theme were identified through manual coding:

• Recruitment and awareness (43 responses)
• Data collection and confidence (38 responses)
• Project support and capacity (26 responses)
• Attitudes (25 responses)
• Involvement and retention (24 responses)
• Individual capability and training (23 responses)
• Funding (23 responses)
• Inclusion and access (20 responses)
• Recognition of volunteers (20 responses)
• Technology and innovation (16 responses)
• Educational institutions (15 responses)

Overlaps and interplays in identified themes

There were considerable overlaps and interconnections between challenges and opportunities among identified themes. For example, attitudes towards citizen science was developed as its own theme, but was also indirectly related to funding, as public perception of citizen science was reported to affect its funding. Similarly, participant retention, data continuity and university engagement were all bilaterally related to funding, project capacity, and support.

‘Knowledge sharing’ was not classified as an individual theme but was prominent across many other themes. For example, there were calls for better reporting and knowledge sharing to improve citizen science awareness, volunteer retention, confidence in the data, attitudes towards citizen science, and collaboration with educational institutions. Similarly, ‘open data’ appeared in five responses directly and four others indirectly. For example, under ‘participant acknowledgement’, responses noted that participants should have access to their own data to ensure the program is ethical and to improve retention. Related, a lack of confidence in data was the second most common theme and influenced other themes including attitudes towards citizen science, funding acquisition, technology, and volunteer retention.

An overview of the most prominent themes, including ‘attitudes’, ‘data collection’ and ‘confidence’ that arose consistently in responses to most survey questions is provided in the heatmap visualisation in Fig 3 below.

Download:

• PNG
  larger image
• TIFF
  original image

Fig 3. Heatmap visualisation of the frequency of theme-related keywords in participant responses.

The colour intensity and numerical values represent the prevalence of each thematic area, revealing which themes dominated specific questions and identifying patterns in participant discourse across the survey. The full list of keywords associated with each theme is provided in S2 Table in the supplementary materials.

https://doi.org/10.1371/journal.pone.0331161.g003

An overview of the themes, challenges and opportunities with illustrative quotes identified through the analysis, as well as how they align with the 10 Principles of Citizen Science [5], is included in Table 1 below.

Download:

• PNG
  larger image
• TIFF
  original image

Table 1. Summary of themes, challenges and opportunities. Abbreviations: CS = Citizen science; EMCR = early-mid career researcher.

https://doi.org/10.1371/journal.pone.0331161.t001

Illustrative case studies

Survey participants shared 45 citizen science initiatives and an additional four were later identified by co-authors. Seventeen STARDIT reports were created from the 49 examples, which were used to help select the final 9 illustrative case studies (Table 2). Of the 9 case studies, we invited people from all the initiatives to validate the data about them in the STARDIT reports, with 7 not replying, and 2 validating the data.

Download:

• PNG
  larger image
• TIFF
  original image

Table 2. Illustrative case studies.

https://doi.org/10.1371/journal.pone.0331161.t002

Cross-case analysis

A cross-case analysis of the nine illustrative case studies identified several key patterns in citizen science initiatives and provides examples of technologically integrated and collaborative approaches to enhance sustainability, data quality and public involvement. Projects varied in accessibility and citizen scientist commitment. For example, three projects focused on coral reefs but used different approaches. CoralWatch enables broad, low-barrier in-water monitoring; Reef Check requires extensive training to support data reliability; while the Great Reef Census enables involvement in reef conservation from anywhere globally.

The integration of artificial intelligence (AI), as demonstrated in the Great Reef Census and NOBURN, reduced reliance on expert training, and enabled broader involvement. The use of AI to analyse photos uploaded by citizen scientists also reduced manual data processing, minimising human error and improving data accuracy.

Many initiatives foster deeper community stewardship by emphasising individual involvement and education, such as WomSAT and Citizens for Refuge Ecology (C4RE) Camps. CoralWatch, WomSAT and the Marine Debris Initiative are examples of how interactive online platforms can be used to involve and educate volunteers and give them access to data they have collected. These platforms are presented in user-friendly and accessible formats, such as interactive maps.

Collaboration between paid scientists, unpaid volunteers, and partner organisations is highlighted as a key factor in project sustainability. However, funding models can differ. Corporate partnerships helped support the longevity and digital capabilities of the Great Reef Census, whereas C4RE Camps and other projects showcase the immediate benefit of short-term government grants. These variations highlight the evolving nature of citizen science, where balancing accessibility, data reliability, technological innovation, and funding stability remains a central challenge for long-term impact.

Participation experience data

We collected data about people’s experience of being involved in this project, with 55% (15/27) responding ‘I felt I could be involved in all tasks of the project’, 45% stating ‘I felt I could be involved in some tasks’ and no one responding ‘I felt there was no opportunity to be involved in any tasks of the project’. When asked ‘Would you recommend using STARDIT to report citizen science or other initiatives?’, 10 people responded to the question, and 90% (9/10) responded yes, with one ‘maybe’.

Recommendations

• 1. Strengthen institutional and government support, and develop publicly led ‘hubs’: Developing infrastructure, funding mechanisms, and policy alignment is important for sustaining long-term citizen science programs [3,6,25]. These challenges could be addressed if governments provide long-term funding for ‘super hubs’ that support institutions to host multiple citizen science projects. These ‘super hubs’ would use centralised funding and infrastructure for administration, ethics, project management, data hosting and sharing, communications, reporting and evaluation, and address the lack of centralised support. For example, a ‘super hub’ at a university could build partnerships with schools, communities, specialist clubs (such as dive clubs and natural history societies) and beyond, while sharing control of the individual citizen science projects with the public. The proposed ‘super hubs’ model is not widely formalised in the literature, but it aligns with calls for shared infrastructure, coordination, and public governance [3,25,26].
• 2. Improve partnerships with schools and educational institutions: Findings under the emergent themes of ‘Educational institutions’ and ‘Recruitment and awareness’ highlighted low awareness of citizen science benefits and difficulties in sustaining student involvement. The WomSAT and CoralWatch case studies suggest that integrated education leads to higher engagement. Strengthening partnerships between schools and higher education institutes can increase the number of people involved and the exchange of skills and knowledge among students, teachers and researchers. This could be achieved by integrating citizen science into curricula and providing appropriate funding and resourcing, such as training for early- to mid-career researchers and education professionals on developing partnerships. Citizen science can be used as a tool for expanding science knowledge and scientific literacy [36], and enhancing public understanding of science [29].
• 3. Improve communication about citizen science: Evidence-informed communication methods should be used to inform the public and raise awareness about citizen science opportunities and their benefits. Increasing research on knowledge and attitudes toward citizen science is important to identify gaps, involve more people and work in ways that align with their needs and preferences. This addresses the finding in the ‘Attitudes’ theme that science is often perceived as ‘unengaging’. Targeted communication can address misperceptions that currently undermine recruitment, funding, and institutional legitimacy [3,65]. This recommendation draws on findings from ’Attitudes’, ‘Recruitment and awareness’, and ‘Involvement and retention’ themes.
• 4. Improve citizen scientist support and recognition: Multi-platform tools, training, and project management should be used to support inclusive ways of working and ensure data quality. Improved mechanisms to recognise citizen scientists are important and may include providing remuneration, clearer guidelines on acknowledgement and authorship in research articles, and better reporting on the contributions of citizen scientists using tools such as STARDIT [45]. This recommendation is informed by findings related to the themes ‘Recognition of volunteers’, ‘Funding’, and ‘Data collection and confidence’, which highlighted recurrent under-recognition of volunteer contributions. These findings identified links between recognition, participant retention, and the continuity and reliability of collected data. They are consistent with established principles of citizen science that emphasise participant acknowledgement, feedback and ethical practice as core elements of effective participation [5,52]. These findings support existing literature showing that participant motivation and sustained engagement are influenced by how volunteer contributions are recognised and supported within projects [7,39].
• 5. Improve transparency about methodology and processes: Developing standardised frameworks for citizen science projects is key, including planning, reporting and evaluation. Quality assurance processes should be linked to reporting standards. Information from these frameworks can improve data validation, quality control, and communication about data reliability. Case studies using such frameworks (e.g., Reef Check’s training-intensive model) can help demonstrate the value and effectiveness of citizen science methods, and any associated participatory methods.

Data about citizen science, or collected and analysed through it, also needs to be publicly accessible to facilitate validation of its accuracy and/or reliability, as a perception of low-quality data can hinder data usage, future funding, and participant retention. Project design should include funding for scientific expertise to address concerns about data quality, help develop robust protocols, and to validate data. This means involving scientists from the outset, rather than having organisations design citizen science projects in isolation. Improved transparency about how data was collected and analysed can help improve trust in citizen science and improve data reliability and usability. This recommendation results from findings under the theme ‘Data collection and confidence’ and ‘Attitudes’, where concerns about perceived data quality repeatedly influenced funding, uptake and legitimacy. Transparent methodological reporting addresses both real and perceived limitations. This strengthens the evidentiary standard of citizen science in line with FAIR and open science principles [56].

• 6. Follow FAIR data principles: Clear reporting standards enabling FAIR data (findable, accessible, interoperable, and reusable) remain critical for impactful citizen science [56]. FAIR-aligned data practices mitigate ethical risks, improve re-use, and maximise long-term impact beyond the lifespan of individual projects [56]. Following FAIR data principles requires building, hosting and maintaining digital infrastructure for collaboration, including working with trusted partner organisations such as the Wikimedia Foundation and affiliated national chapters such as Wikimedia Australia. Large open data aggregators for citizen science should be utilised wherever possible, such as the Global Biodiversity Information Facility. All government-funded research should mandate FAIR data sharing.
• Such data sharing should be mindful of the United Nations Declaration on the Rights of Indigenous Peoples (UNDRIP) which declares the rights of Indigenous peoples ‘to maintain, control, protect and develop their intellectual property’. The framing of knowledge as “data” to be owned, shared, or made “open” can itself reflect colonial logics that extract and commodify Indigenous knowledge outside of its cultural and relational context. The CARE Principles for Indigenous Data Governance [54] provide a complementary framework recognising Indigenous peoples’ rights to govern data about their communities, territories, and cultural practices — rights that may directly conflict with calls for open or centralised data access. Where citizen science projects involve Indigenous participants or knowledge, project designers should co-develop data governance agreements with relevant communities early, which may include tiered access controls or community-held repositories. Our recommendations referencing FAIR principles should be read with these considerations in mind.
• This recommendation responds to the findings from ‘Data collection and confidence’, ‘Recognition of volunteers’, and ‘Inclusion and access’ themes. There were many learnings about data collection, sharing and analysis. For data collection and sharing, online tools for citizen science need to work on multiple platforms and be inclusive and accessible, as exemplified by the ‘Sweaty and Stressed’ and ‘Marine Debris Initiative’ datasets. While four of the case studies we presented shared data according to the FAIR principles, five did not. This aligns with findings from other research that warns against projects where people donate their ‘unpaid work’ for data that may then be privately owned [9]. The use and promotion of large open data aggregators can ensure that data can continue to be accessible and used even once the citizen science initiative itself is no longer sustained. Important considerations for these data aggregators include clear quality controls, centralisation of the database, and the ability to flag incorrect records and filter data.

• 7. Improve public involvement in all stages of research: The public should be involved in all stages of science, from topic selection to project implementation and decision-making. This involvement should be reported transparently. This can be achieved by implementing inclusive involvement strategies with sufficient resourcing, including funding and staff time. All grant funding for citizen science by the government should require demonstration of how the public can be involved at each stage of the project. To ensure citizen science is inclusive, grant funding should provide specific funding to pay people for their time for being involved and ensure taxation practices do not inhibit involvement. This recommendation arises from the finding that 45% of participants felt only partially involved in project tasks. From these themes we identified exclusionary power dynamics and limited agency for participants.
• 8. Expand training opportunities: University-led and community-driven training, including formal credentials, could be offered to support both professional researchers and citizen scientists. Training and support networks should be established, including creating ‘communities of practice’ for early- to mid-career researchers who are initiating, leading or working on citizen science projects. This recommendation is informed by findings from the themes ‘Individual capability and training’ and ‘Data collection and confidence’, which indicate that insufficient training can constrain both the quality of engagement and the robustness of citizen science data. The C4RE Camps demonstrate how expert-led training passes on knowledge and builds wider community capacity. As citizen science increasingly incorporates new technologies, including data platforms and AI, training requirements expand. Training needs to extend beyond technical skills to include data interpretation, ethical practice, and responsible data management. In this way, confidence in data quality and trust in research processes can be maintained. Addressing these training needs can help build capacity on both sides of the research relationship and support sustainable participation and career pathways for participants [9,39].
• 9. Encourage transparent and ethical research partnerships: There is a need to improve ways for the public to be involved in setting priorities for both non-government and government funded research. Findings across the themes ‘Inclusion and access’, ‘Funding’, and ‘Project support and capacity’ indicate that opportunities for public involvement in decision-making are often constrained by structural factors. These include limited governance capacity and short-term or inflexible funding arrangements. These constraints reflect systemic barriers rather than an absence of public interest in participation. Transparent collaborations between industry, government, and academia can help address these constraints by clarifying roles, responsibilities, and decision-making processes. Such collaborations can also improve the availability of information about how well stakeholders align with the values and principles of citizen science and ethical research conduct. This type of partnership, which emphasises transparency, aligns with established guidance on opening science to society and strengthening ethical, participatory research governance [9,22].
• 10. Improve consistency and terminology: Relevant governments and international organisations (such as the United Nations, World Health Organization and the Citizen Science Global Partnership) should lead a process to ensure there is consistent terminology describing citizen science involvement and its underlying values and principles across all disciplines (e.g., health, environment and education) in multiple languages. Terminological consistency supports cross-sector recognition and integration of citizen science into policy and funding frameworks [3,6]. Calls in the literature for improved consistency and terminology [8,40] are supported by this recommendation. This recommendation came from the qualitative data across the themes ‘Attitudes’ and ‘Educational institutions’ that highlighted that inconsistent language can hinder funding eligibility and methodological legitimacy.

Fig 4 below provides an overview of these recommendations, with each recommendation mapped to the most relevant stakeholders who should act on them: funders, citizen science initiative operators, academic institutions, policy makers, and local communities. A detailed overview of the recommendations along with key information about the paper is provided in the supplementary material as S1 Fig.

Download:

• PNG
  larger image
• TIFF
  original image

Fig 4. The recommendations that are most relevant and actionable to each major stakeholder group involved in citizen science initiatives.

https://doi.org/10.1371/journal.pone.0331161.g004

Considerations for broader societal context

Study Limitations

The survey responses were analysed qualitatively because the sample size of 46 was not large enough to make statistical inferences. Most respondents were Australian, and although the results primarily reflect the Australian context and perspectives, the findings and learnings are considered broadly generalisable. There was also an unequal representation of fields. For example, ecology was heavily represented while some fields were underrepresented, such as health (despite its significant share of governmental budget spending), or absent, such as climate and weather projects. The survey likely targeted individuals already engaged in citizen science, which introduced a bias toward those who view it positively. Perspectives from stakeholders with competing interests who might be less supportive of citizen science may be underrepresented. While responses included a range of stakeholders, the sample was skewed toward academia. This overrepresentation may have narrowed the range of perspectives, which in turn may have affected emerging themes and potentially omitted key issues important to the public. While the sample size is relatively small, the co-authors agreed that the survey responses, including 16,247 words in open text fields, as well as the analysis of identified citizen science projects enabled the identification of key barriers and opportunities for citizen science and added valuable contributions to the field.

The generalisability of the 10 recommendations presented in this article should be interpreted with caution in contexts that differ substantially from the study sample. Given the predominantly Australian, academic, and ecology-oriented composition of respondents, several important contextual limitations warrant acknowledgement.

In Low- and Middle-Income Countries (LMICs), the structural preconditions assumed by several recommendations (including long-term government funding, centralised university infrastructure, and access to digital platforms) may not be readily available. Additionally, power dynamics between international funders or institutions and local communities can complicate the co-production principles central to our recommendations. Recommendations related to institutional ‘super hubs’, formal credentialing, and mandated FAIR data sharing may require significant adaptation to be feasible and equitable in lower-resource contexts. Future research should explore how these recommendations translate to citizen science initiatives in LMICs, where community-led models and existing local knowledge systems may offer alternative pathways to the ones we have identified. Nonetheless, citizen science is inherently a global endeavour, and these recommendations should not be read as exclusively relevant to high-income, English-speaking contexts. Higher-income countries, institutions and funders have both the opportunity and responsibility to support citizen science in LMICs.

In clinical and health research contexts, the degree of public involvement and data openness our recommendations advocate may require adaptation to ensure appropriate ethical oversight. Further research is required to ensure a balance between public interest and commercial interests (including those of pharmaceutical sponsors or private research funders) which encourages transparency and the co-production methods that are central to this articles’ recommendations.

Knowledge gaps and opportunities for further research

As part of this research, we have identified two key knowledge gaps and opportunities for further research. Firstly, while we suggest that many of the themes and challenges identified as part of this research are reflective of wider patterns across the world, there are also likely large differences across social, economic, technological, political and cultural contexts. Secondly, while the citizen science experts surveyed as part of this research hold key insights into how the impacts of citizen science can be improved, we recognize that other stakeholder groupings – including policymakers, people working for industry, and people from marginalised communities – hold important perspectives for strengthening their own involvement in citizen science projects and improving the overall outcomes of citizen science initiatives.

Further research could explore both gaps by extending the approach applied in this research to new geographic contexts and stakeholder groupings, including policymakers, industry actors, and marginalised communities. A key element of this work could involve identifying best practices that can be scaled, replicated, and used to strengthen international collaboration and better leverage the transformative potential of citizen science. Future work could also adopt more structured and transparent approaches to planning and evaluation, including assessing impact, feasibility, costs, and time horizons, alongside developing shared indicators and monitoring frameworks aligned with FAIR data principles and using STARDIT reporting to support comparability and accountability across initiatives.