Open Science Practices: A Guide for Researchers

Open science is a set of methods, tools, platforms, and practices that make scientific research more accessible, transparent, reproducible, and reliable. This includes sharing code, data, and research materials, embracing new publishing formats such as registered reports and preprints, pursuing replication studies and reanalyses, optimising statistical approaches to improve evidence assessment, and re-evaluating institutional incentives. The shift toward open science practices is partly due to mounting evidence that studies across disciplines suffer from biases, underpowered designs, and irreproducible or non-replicable results. It also stems from a general desire amongst many researchers to reduce hyper-competitivity in science and instead promote collaborative research that benefits science and society. This guide provides a practical framework for researchers at any career stage to adopt open science practices in their daily workflow, with concrete steps for implementation, common pitfalls to avoid, and criteria for when to seek professional guidance.

At a Glance

The table below summarizes the core open science practices covered in this guide, their primary purpose, the level of effort typically required to implement them, and the resources that support each practice.

Defining Open Science and Its Core Principles

Open science is a broad term that refers to a range of methods, tools, platforms, and practices that aim to make scientific research more accessible, transparent, reproducible, and reliable. This includes sharing code, data, and research materials, embracing new publishing formats such as registered reports and preprints, pursuing replication studies and reanalyses, optimising statistical approaches to improve evidence assessment, and re-evaluating institutional incentives. The ongoing shift towards open science practices is partly due to mounting evidence that studies across disciplines suffer from biases, underpowered designs, and irreproducible or non-replicable results. It also stems from a general desire amongst many researchers to reduce hyper-competitivity in science and instead promote collaborative research that benefits science and society.

Open science ensures that research is transparently reported and freely accessible for all to assess and collaboratively build on. The core principles include transparency in methods and analysis, accessibility of outputs, reproducibility of results, and collaboration across the research community. These principles apply across disciplines, from the health sciences to the social sciences, and they are implemented through a set of concrete practices that researchers can adopt incrementally.

The Current State of Open Science Adoption

Understanding the current landscape of open science adoption helps researchers set realistic expectations and identify areas where their own practices can improve. Systematic reviews across multiple fields reveal a varied picture of uptake.

In the false memory literature, a preregistered systematic review of 388 publications from 2015 to 2023 found that most studies (86.86%) adhered to at least one open science measure, with publication accessibility being the most consistently adopted practice at 73.97%. Data sharing demonstrated the most substantial growth, reaching about 75% by 2023, while preregistration and analysis script sharing lagged, with 20-25% adoption in 2023. This review highlights a promising trend towards enhanced research quality, transparency, and reproducibility, but the inconsistent implementation of open science practices may still challenge the verification, replication, and interpretation of research findings.

A scoping review of 500 gambling research studies published between 2016 and 2019 found that 54.6% of studies used at least one of nine open science practices. However, the prevalence of individual practices varied widely: 1.6% for pre-registration, 3.2% for open data, 0% for open notebook, 35.2% for open access, 7.8% for open materials, 1.4% for open code, and 15.0% for preprint posting. Only 6.4% of the studies included a power analysis and 2.4% were replication studies. Exploratory analyses showed that studies that used any open science practice, and open access in particular, had higher citation counts.

In oncology prognostic model studies using machine learning, a systematic review of 46 publications found that adoption of open science principles was poor. Only one study reported availability of a study protocol, and only one study was registered. Thirty-five studies (76%) provided data sharing statements, with 21 (46%) indicating data were available on request to the authors and seven declaring data sharing was not applicable. Two studies (4%) shared data. Only 12 studies (26%) provided code sharing statements, including 2 (4%) that indicated the code was available on request to the authors. Only 11 studies (24%) provided sufficient information to allow their model to be used in practice. The use of reporting guidelines was rare, with eight studies (18%) mentioning using a reporting guideline.

These findings demonstrate that while open science practices are gaining traction, adoption remains inconsistent across fields and individual practices. Researchers should view this as an opportunity to lead in their own disciplines by adopting practices that are still uncommon.

Open Access Publishing

Open access publishing makes research articles freely available to all readers, removing paywalls that limit access to scientific knowledge. This practice addresses the fundamental goal of open science to make research freely accessible for all to assess and collaboratively build on.

Choosing an Open Access Route

Researchers have several options for open access publishing. The first is to publish in fully open access journals, where all articles are freely available immediately upon publication. The second is to publish in hybrid journals that offer an open access option for individual articles, often for a fee. The third is to deposit a version of the manuscript in a repository, which is often called green open access. Many funders and institutions have policies that require or encourage open access publishing, so researchers should check their funding agreements and institutional guidelines before choosing a route.

Using Literature Databases

The NCBI Literature Resources platform provides access to a range of databases for finding and accessing scientific literature. PubMed is a free search engine that provides access to citations and abstracts for biomedical literature from the National Library of Medicine. These resources help researchers find open access versions of articles and verify the accessibility of their own publications.

Practical Steps for Open Access

When planning a publication, researchers should identify the open access policies of target journals before submission. They should check whether the journal is fully open access, offers a hybrid option, or permits repository deposition. They should also verify whether their funder or institution has specific open access requirements. After publication, researchers should deposit the appropriate version of their manuscript in any required repositories and confirm that the final version is accessible through databases like PubMed.

Open Data Sharing

Open data sharing makes research data available for verification, reuse, and secondary analysis. This practice is central to reproducibility because it allows other researchers to reanalyze data and confirm findings.

Benefits of Data Sharing

Improved data sharing practices offer the potential to enhance research efficiency and impact. Opportunities include the exploration of new research questions, reduced data duplication, and more robust, conclusive, and equitable science. When data are shared, other researchers can combine datasets to address questions that no single study could answer, and they can verify the analyses reported in publications.

Barriers to Data Sharing

Despite these benefits, data sharing is constrained by structural, cultural, and methodological barriers. Challenges include institutional and geographical data fragmentation, dataset heterogeneity, time- and resource-intensive requirements, consent considerations, and publication-focused academic incentives over data stewardship. Researchers must navigate these barriers while maintaining ethical and legal obligations to research participants.

Practical Steps for Data Sharing

The Research Data Framework from the National Institute of Standards and Technology provides a structured approach to managing research data. Researchers should develop a data management plan at the start of a project that addresses data collection, storage, documentation, and sharing. They should document their data using standard metadata schemas and deposit data in recognized repositories that assign persistent identifiers. They should also include data availability statements in their publications that clearly indicate where and how data can be accessed.

Data Sharing in Specific Fields

In nutritional science, a symposium report highlighted that transitioning toward effective data sharing practices will require coordinated action across academic institutions, the research community, funders, and publishers, including clear training, incentives, policies, and an overall cultural shift toward open science. Researchers in fields where data sharing is less established should expect to encounter more barriers and should plan accordingly.

Preregistration

Preregistration involves documenting the study design, hypotheses, and analysis plan before data collection begins. This practice distinguishes confirmatory analyses, which test pre-specified hypotheses, from exploratory analyses, which generate new hypotheses from data.

Purpose of Preregistration

Preregistration addresses publication bias and questionable research practices by making the distinction between confirmatory and exploratory analyses transparent. When researchers preregister their analysis plan, they commit to reporting their results regardless of whether the findings support their hypotheses. This reduces the incentive to selectively report significant results or to adjust analyses until they produce significant findings.

Preregistration as a Flexible Tool

Open science implementation must be context-dependent instead of one-size-fits-all, and practical research realities require flexible approaches to idealized policies. Researchers should approach preregistration as an evolving tool instead of a rigid rule. During the COVID-19 pandemic, some researchers found that their preregistered plans needed to adapt as circumstances changed. The key is to document any deviations from the preregistration and explain the reasons for those deviations in the final publication.

Practical Steps for Preregistration

Researchers should preregister their study design and analysis plan before data collection begins. The Experimental Design Assistant from the NC3Rs provides guidance on experimental design and can help researchers plan their studies rigorously. When preregistering, researchers should specify their primary and secondary hypotheses, the variables they will measure, the statistical analyses they will conduct, and their criteria for interpreting results. If deviations from the preregistration become necessary, researchers should document them transparently and explain the rationale.

Registered Reports

Registered Reports are a publishing format in which the study design and analysis plan are peer reviewed before data collection begins. This format shifts the focus of peer review from the results to the quality of the methods.

How Registered Reports Work

In a Registered Report, authors submit an introduction and methods section, including their planned analyses, to a journal. The journal sends this proposal for peer review. If the proposal is accepted, the journal commits to publishing the final article regardless of the results, provided the authors follow their approved methods. This format eliminates publication bias against null results and incentivizes rigorous study design.

Benefits and Challenges

Registered Reports offer significant benefits for research quality. They reduce publication bias, encourage rigorous methods, and provide authors with feedback on their study design before data collection. However, they also pose considerable time and other resource burdens. Researchers should weigh these costs against the benefits when deciding whether to pursue this format.

Practical Steps for Registered Reports

Researchers interested in Registered Reports should identify journals that offer this format. The EQUATOR Network provides resources on reporting guidelines and can help researchers identify appropriate journals and reporting standards. Researchers should prepare their Registered Report proposal according to the journal's instructions and be prepared to revise their methods based on reviewer feedback.

Open Materials and Code

Sharing research materials and analysis code allows other researchers to understand exactly how a study was conducted and how the data were analyzed. This practice is essential for computational reproducibility.

What to Share

Researchers should share the materials they used to conduct their study, including questionnaires, stimuli, protocols, and analysis scripts. They should also share the code they used to process and analyze their data. The level of sharing can vary from providing materials on request to depositing everything in a public repository.

Current Adoption Rates

Adoption of open materials and code sharing varies widely across fields. In the gambling research scoping review, 7.8% of studies shared open materials and 1.4% shared open code. In the oncology machine learning review, only 12 studies (26%) provided code sharing statements, including 2 (4%) that indicated the code was available on request to the authors. These low rates suggest substantial room for improvement.

Practical Steps for Sharing Materials and Code

Researchers should organize their materials and code in a way that others can understand. They should include comments in their code, provide a readme file that explains the structure of their files, and deposit everything in a repository that assigns persistent identifiers. They should also cite their materials and code in their publications so that other researchers can find them.

Reporting Guidelines

Reporting guidelines provide checklists of the information that should be included in a publication to ensure complete and transparent reporting. These guidelines help authors report their methods and results in a way that allows readers to understand and evaluate the study.

Available Guidelines

The EQUATOR Network is an international initiative that provides resources on reporting guidelines for health research. It maintains a comprehensive database of reporting guidelines for different study types, including randomized trials, observational studies, diagnostic accuracy studies, and prognostic model studies. Researchers should select the guideline that matches their study design.

Current Usage

The use of reporting guidelines is often rare. In the oncology machine learning review, only eight studies (18%) mentioned using a reporting guideline. This low adoption rate means that many publications lack the information needed to evaluate or replicate the research.

Practical Steps for Using Reporting Guidelines

Researchers should identify the appropriate reporting guideline for their study design at the planning stage. They should use the guideline checklist to ensure that their manuscript includes all required information. They should also mention in their manuscript that they followed the guideline and provide the completed checklist as supplementary material.

Practical Implementation Workflow

Adopting open science practices requires a systematic approach that integrates these practices into the research workflow from project inception to publication.

Step 1: Plan for Open Science at Project Inception

At the start of a project, researchers should develop a data management plan that addresses data collection, storage, documentation, and sharing. They should also decide whether to preregister the study and identify the appropriate reporting guideline. The Research Data Framework provides guidance on data management planning.

Step 2: Design the Study with Reproducibility in Mind

Researchers should design their studies to be replicable and reproducible. This includes documenting all decisions about sample size, variable measurement, and statistical analysis. The Experimental Design Assistant can help researchers plan their experimental designs rigorously.

Step 3: Preregister the Study Before Data Collection

Before collecting data, researchers should preregister their study design and analysis plan. This documentation should be detailed enough that another researcher could replicate the study based on the preregistration alone.

Step 4: Document Everything During Data Collection

During data collection, researchers should maintain detailed records of their procedures, including any deviations from their preregistered plan. They should also document their data processing and analysis steps so that they can share these details with others.

Step 5: Share Materials, Data, and Code

After data collection and analysis, researchers should prepare their materials, data, and code for sharing. This includes cleaning and documenting the data, adding comments to the code, and depositing everything in an appropriate repository.

Step 6: Publish with Transparency

When writing the manuscript, researchers should follow the appropriate reporting guideline and include data availability statements. They should also consider posting a preprint to make their findings available before formal publication.

Step 7: Verify Accessibility After Publication

After publication, researchers should verify that their article is accessible through databases like PubMed and that their data and code are accessible through their chosen repositories.

Records and Measurements

Maintaining accurate records of open science practices is essential for demonstrating compliance with funder and institutional policies and for tracking the impact of these practices.

What to Record

Researchers should maintain records of their preregistrations, including the date of registration and any subsequent amendments. They should also record where their data, materials, and code are deposited and the persistent identifiers assigned to these deposits. For publications, they should record the journal's open access policy and the route they used to achieve open access.

Measuring Impact

Researchers can measure the impact of their open science practices by tracking citation counts, data downloads, and code reuse. Studies have shown that open access publications tend to have higher citation counts. Researchers can also track the number of times their data and code are accessed or cited.

Institutional Monitoring

A Delphi study involving research administrators, researchers, specialists in dedicated open science roles, and librarians reached consensus on 19 open science practices that should be monitored at biomedical research institutions. This core set of open science practices forms the foundation for institutional dashboards and may also be of value for the development of policy, education, and interventions. Researchers should be aware that their institutions may be tracking their open science practices.

Common Failure Patterns

Understanding common failure patterns helps researchers anticipate challenges and avoid mistakes when implementing open science practices.

Incomplete Data Sharing

A common failure is providing data sharing statements that indicate data are available on request but not actually sharing the data. In the oncology machine learning review, 21 studies (46%) indicated data were available on request to the authors, but only 2 studies (4%) actually shared data. Researchers should ensure that their data sharing statements accurately reflect what is available.

Preregistration After Data Collection

Another common failure is preregistering a study after data collection has already begun. This defeats the purpose of preregistration, which is to document the analysis plan before seeing the data. Researchers should be honest about the timing of their preregistration and should not present post hoc analyses as preregistered.

Inadequate Documentation

Sharing data or code without adequate documentation is a common failure. Other researchers cannot use data or code that is poorly documented. Researchers should include readme files, code comments, and metadata that explain their files.

Ignoring Reporting Guidelines

Failing to use reporting guidelines is a common failure that reduces the transparency and completeness of publications. Researchers should identify the appropriate guideline for their study design and use it throughout the writing process.

Treating Open Science as Rigid Rules

Open science implementation must be context-dependent instead of one-size-fits-all, and practical research realities require flexible approaches to idealized policies. Researchers who treat open science practices as rigid rules may find them unworkable and abandon them entirely. Instead, researchers should approach these practices as evolving tools that can be adapted to their specific circumstances.

Limitations and Professional Escalation Criteria

Open science practices have limitations, and researchers should know when to seek professional guidance.

Limitations of Open Science Practices

Open science practices pose considerable time and other resource burdens. Researchers need to balance these burdens against the benefits of transparency and reproducibility. Research is needed to help determine the value of these added burdens and to identify efficient strategies for implementing open science practices.

Data sharing may be constrained by legal and ethical obligations, including consent requirements and data protection regulations. Researchers must navigate these constraints while maximizing the transparency of their research.

When to Seek Professional Guidance

Researchers should seek professional guidance when they encounter situations that exceed their expertise. This includes complex data sharing agreements, sensitive data that require special handling, and legal or ethical questions about data sharing.

The Research Data Framework provides guidance on data management that can help researchers address common challenges. Institutional data librarians and research support staff can also provide assistance with data management planning, repository selection, and compliance with funder policies.

Researchers should escalate to professional guidance when they are unsure about the legal or ethical implications of data sharing, when they need to negotiate data sharing agreements, or when they are dealing with particularly sensitive data.

Welfare and Safety Context

Open science practices have implications for research ethics and participant welfare that researchers must consider.

Protecting Participant Privacy

When sharing data, researchers must protect the privacy of research participants. This includes de-identifying data, implementing access controls for sensitive data, and ensuring that data sharing agreements include appropriate safeguards. In psychiatric genetics, open science implementation guidelines must be specific to data, privacy, and research conduct challenges.

Balancing Transparency and Privacy

Researchers must balance the goal of transparency with the obligation to protect participant privacy. This balance requires careful consideration of the risks and benefits of data sharing in each specific context. Researchers should consult with their institutional review boards or ethics committees when they have questions about data sharing.

Equitable Collaboration

Open science practices should promote equitable collaboration. This includes equitable authorship and citation practices, consideration of the interests of nonacademic collaborators, and attention to equity considerations in implementation. Researchers should consider how their open science practices affect all stakeholders, including research participants, collaborators, and communities.

Frequently Asked Questions

What is the difference between open access and open science?

Open access is one component of open science. Open access specifically refers to making research publications freely available to readers. Open science is a broader term that encompasses open access, open data, open materials, open code, preregistration, and other practices that make research more transparent and reproducible.

How much time does preregistration take?

The time required for preregistration varies depending on the complexity of the study and the level of detail in the preregistration. Researchers should expect to spend several hours preparing a preregistration document. This investment can save time later by reducing ambiguity about analysis decisions and by providing a clear record of the study plan.

Do I have to share all of my data?

No. Researchers should share data to the extent possible while protecting participant privacy and respecting legal and ethical obligations. Some data cannot be shared because of consent restrictions, data protection regulations, or other constraints. In these cases, researchers should provide a clear data availability statement that explains the restrictions.

What if my preregistered analysis plan needs to change?

Deviations from a preregistered analysis plan are sometimes necessary. Researchers should document any deviations and explain the reasons for them in their publications. The key is to be transparent about what changed and why, instead of presenting post hoc analyses as preregistered.

How do I choose a repository for my data?

Researchers should choose a repository that is appropriate for their field, assigns persistent identifiers, and meets funder and institutional requirements. The Research Data Framework provides guidance on data management that can help researchers select appropriate repositories.

What are reporting guidelines and why should I use them?

Reporting guidelines are checklists that specify the information that should be included in a publication to ensure complete and transparent reporting. The EQUATOR Network provides a comprehensive database of reporting guidelines. Using reporting guidelines improves the quality of publications and makes it easier for other researchers to evaluate and replicate the research.

How can I find open access versions of articles?

PubMed provides access to citations and abstracts for biomedical literature and links to freely available full-text versions when they exist. The NCBI Literature Resources platform provides access to a range of databases for finding scientific literature.

What is a Registered Report and how is it different from preregistration?

A Registered Report is a publishing format in which the study design and analysis plan are peer reviewed before data collection begins. Preregistration is the act of documenting the study plan before data collection, which can be done independently of any journal. Registered Reports incorporate preregistration into the peer review process and provide a commitment to publish regardless of the results.

Related Articles

• Biomedical Science Research Building
• Scientific Data Provenance For Reproducible Research
• Scientific Data Provenance For Reproducible Research
• Scientific Data Provenance For Reproducible Research
• Research Assistant Careers in Life Science

References and Further Reading

• Research Data Framework. National Institute of Standards and Technology.
• EQUATOR Network. EQUATOR Network.
• Experimental Design Assistant. NC3Rs.
• NCBI Literature Resources. National Center for Biotechnology Information.
• PubMed. National Library of Medicine.
• Open Science Practices in Psychiatric Genetics: A Primer.. Biological psychiatry global open science, 2024.
• Open science practices in the false memory literature.. Memory (Hove, England), 2024.
• Open Science Practices in Gambling Research Publications (2016-2019): A Scoping Review.. Journal of gambling studies, 2023.
• Open and reproducible science practices in psychoneuroendocrinology: Opportunities to foster scientific progress.. Comprehensive psychoneuroendocrinology, 2022.
• Open science practices need substantial improvement in prognostic model studies in oncology using machine learning.. Journal of clinical epidemiology, 2024.
• Open science practices for eating disorders research.. The International journal of eating disorders, 2021.
• Community consensus on core open science practices to monitor in biomedicine.. PLoS biology, 2023.
• Open science.. Current biology : CB, 2023.
• EVALUATION OF OPEN SCIENCE IN THE PEER REVIEW PROCESS. 2026.
• Research Technical Professionals as Catalysts in Advancing Open Science. 2026.
• Extent of Open Science Practices in the Reporting of Real World Evidence Research
• Scoring employment interviews with large language models: Evaluation design components, validity investigations, and best practice recommendations.. 2026.
• Improving data sharing in nutritional science: a symposium report of challenges, opportunities, and best practice recommendations.. 2026.
• Preregistration: Open Science and the Public's Trust in Science. 2020.
• Open science: My insights into data sharing, preregistration, and replication. Canadian journal of experimental psychology = Revue canadienne de psychologie experimentale, 2025.
• Towards Transparency and Open Science (A Principled Perspective on Computational Reproducibility and Preregistration). 2023.
• Yes! We’re open. Open science and the future of academic practices in translation and interpreting studies. Translation and Interpreting, 2021.
• The Practice Progress and Future Exploration of Open Science at Home and Abroad. Library and Information Service, 2021.
• The dawn of an open exploration era: Emergent principles and practices of open science and innovation of university research teams in a digital world. Technological Forecasting and Social Change, 2020.
• Advantages and challenges to open science practices. Terra Economicus, 2023.

This article is educational and does not replace institutional policy, professional advice, or applicable safety and regulatory requirements.