Policy on Data Sharing and Research Reproducibility
1. Purpose and scope
The Andean Journal of Education considers the data, materials and code underlying a published article to be an integral part of the scholarly record. Making them available strengthens the transparency, verifiability and reuse of educational research, and it is consistent with the journal's diamond open-access model: if readers pay nothing to read our articles, they should also be able to examine the evidence on which those articles rest.
This policy applies to all submissions reporting original empirical research (quantitative, qualitative, mixed-methods and design-based), as well as to systematic reviews and meta-analyses. Essays, editorials and book reviews are exempt, although authors of such contributions are encouraged to share supporting materials where relevant.
The journal follows the principle "as open as possible, as closed as necessary." Openness is the default; restrictions are legitimate but must be justified.
2. Guiding principles
The Andean Journal of Education aligns its practice with:
- the FAIR principles (Findable, Accessible, Interoperable, Reusable);
- the CARE Principles for Indigenous Data Governance (Collective Benefit, Authority to Control, Responsibility, Ethics), which are especially relevant to research conducted with Indigenous, Afro-descendant, rural and other historically marginalised communities in the Andean region;
- the Joint Declaration of Data Citation Principles, which establishes that datasets are legitimate, citable products of research;
- the ethical standards described in the journal's Publication Ethics Policy and in COPE guidance.
Openness is never a justification for breaching the confidentiality, dignity or safety of research participants, particularly minors, teachers, school communities and Indigenous peoples.
3. Data availability statement (mandatory)
Every research article must include a Data Availability Statement placed immediately before the reference list. The statement is published with the article and must indicate where the data can be found, under what conditions they may be accessed, and — where access is restricted — why.
Model statements (authors should adapt one of the following):
- Openly available data: "The data supporting the findings of this study are openly available in [repository] at [DOI/URL], under a [licence] licence."
- Data available on request due to ethical restrictions: "The data supporting the findings of this study contain information that could compromise the privacy of participants. A de-identified subset is available from the corresponding author on reasonable request and subject to approval by [ethics committee]."
- Data available under controlled access: "The data are deposited in [repository] at [DOI]. Access is restricted and may be requested through the procedure described in the repository record."
- Data derived from third-party sources: "The data analysed in this study were obtained from [source]. Restrictions apply to their redistribution; they were used under licence for the present study and are available from [source] at [URL]."
- No new data: "No new data were generated or analysed in this study."
Statements such as "data are available on request" without further specification are not acceptable, as they cannot be verified and are known to be unreliable in practice.
4. What should be shared
Authors are expected to make available all materials required for an independent researcher to understand, verify and, where appropriate, reproduce the reported results. Depending on the design, this typically includes:
- de-identified datasets underlying tables, figures and statistical results;
- data collection instruments: questionnaires, scales (where copyright permits), interview and focus-group guides, observation protocols, rubrics;
- codebooks, variable dictionaries, coding schemes and category systems;
- for qualitative studies: anonymised transcripts or extracts, analytical memos, coding trees, and a description of the analytical procedure sufficient to make the interpretive path traceable;
- for systematic reviews: search strategies for each database with dates, inclusion and exclusion criteria, screening decisions, the list of excluded full texts with reasons, and extraction and risk-of-bias tables;
- for design-based and intervention research: the pedagogical or technological design, learning materials, implementation logs and fidelity records;
- syntax, scripts and analytical code (see section 8).
Where raw data cannot be shared, authors should share whatever can be released without risk—for example, aggregated data, synthetic datasets, metadata-only records, or the analytical code and instruments.
5. Data repositories
A data repository is a storage space in which researchers deposit the datasets associated with their research and obtain a persistent identifier for them. Authors seeking to comply with this policy must identify a repository suited to their data.
The Andean Journal of Education requires that the chosen repository:
- assign a persistent identifier (preferably a DOI);
- allow the licence and access conditions of the deposit to be specified;
- provide structured metadata and long-term preservation;
- be publicly searchable, even when the data themselves are under controlled access.
Generalist repositories: Zenodo, Open Science Framework (OSF), Harvard Dataverse, Dryad, Figshare, Mendeley Data.
Regional and institutional infrastructure: the institutional repository of the Universidad Andina Simón Bolívar, the repositories of the authors' own institutions, and aggregators such as LA Referencia and RedCLARA. RAE encourages using Latin American infrastructure whenever it meets the criteria above, as a matter of scholarly sovereignty and long-term regional capacity.
Discipline-specific repositories: where the field has established options — for example ICPSR and the Qualitative Data Repository (QDR) — these are preferred, since their metadata and curation practices are better adapted to educational and social science data.
Personal or institutional web pages, cloud drives (Google Drive, Dropbox, OneDrive), academic social networks (ResearchGate, Academia.edu), and email attachments are not acceptable data-sharing mechanisms, since they offer neither persistence nor citability. Supplementary files hosted by the journal may complement a deposit but do not replace it.
6. Community-endorsed public repositories
Where community-endorsed mandates require submitting data to public repositories, authors must deposit their datasets in the appropriate repositories and provide the accession numbers or identifiers (where available) in the article. Where a funder, national agency or research consortium imposes its own data management requirements, authors are responsible for ensuring that their deposit satisfies both those requirements and this policy.
7. Sensitive data, consent and ethical restrictions
Educational research frequently involves minors, teachers, families and school communities. RAE therefore recognises that full openness is not always possible or desirable, and asks authors to plan for sharing from the outset rather than to treat it as an afterthought:
- Whenever feasible, informed consent (and assent, in the case of minors) should explicitly cover depositing and reusing anonymised data. Authors should state in the Methods section what participants were told about data sharing.
- De-identification. Remove direct identifiers and assess indirect identifiers for re-identification risk, paying particular attention to small populations—a single school, a rural community, a specific cohort—where combinations of variables can be identifying even after names are removed.
- Controlled access. Where anonymisation would destroy the data's analytical value or cannot guarantee protection, authors should deposit the data under controlled access, publish the metadata record openly, and specify the conditions and procedure for requesting access.
- Collective governance. For research with Indigenous or Afro-descendant communities, decisions about deposit, licensing and reuse must respect the community's authority over its own data, and the article should state how that authority was exercised.
- Ethical approval. Where the study required approval from an ethics committee, the corresponding reference must appear in the manuscript.
An ethically justified restriction, clearly explained, is fully compatible with this policy. An unexplained restriction is not.
8. Custom computer code, software tools and mathematical algorithms
To enable a full assessment of submissions, authors must make available to the Editors and/or reviewers, on request, any custom computer code, software tool or algorithm used to generate the results and conclusions reported in the manuscript.
- In addition, the Andean Journal of Education strongly encourages authors to:
- deposit analytical code (R, Python, Stata, SPSS syntax, NVivo or Atlas.ti project files, etc.) in a repository or a version-controlled platform, with an archived release that has a DOI;
- release code under an OSI-approved open-source licence;
- document the computational environment (software versions, packages and dependencies) so that results can be re-executed;
- provide a brief README explaining how the files relate to the tables and figures in the article.
For manuscripts whose main contribution is a piece of software, a model or an algorithm, deposit and open licensing are a condition of publication, not a recommendation.
9. Use of artificial intelligence in data generation and analysis
Where generative AI systems or machine-learning models have been used to generate, transform, code or analyse data, authors must report this in the Methods section, specifying the tool and version, the purpose for which it was used, and the procedures used to validate its output. Where prompts are part of the method (for example, in automated coding of open responses), include them in the shared materials. Responsibility for the accuracy and integrity of the results always rests with the authors.
10. Licensing and citation of data
The Andean Journal of Education recommends CC0 or CC BY 4.0 for datasets, and an OSI-approved licence for code, consistent with the journal's open-access licensing.
Datasets and code deposited by the authors, as well as any third-party datasets analysed, must be cited in the reference list in APA format, with author, year, title, version, repository and DOI, and cited in the text like any other source. Data citation is the mechanism by which those who produce data receive academic credit for doing so.
11. Access during peer review
Where data or code have been deposited under embargo or restricted access, authors must provide the Editors with a means of access for review purposes (for example an OSF view-only anonymous link or a private Zenodo/Dataverse link) that does not compromise the anonymity of the review process. Reviewers may consult these materials but may not use, distribute or cite them prior to publication.
12. Compliance, exceptions and non-compliance
- Compliance with this policy is verified during editorial screening; manuscripts without a Data Availability Statement will be returned to the authors before peer review.
- Requests for exception must be submitted in the cover letter and are assessed by the editorial team on a case-by-case basis.
- Where, after publication, the journal receives a substantiated report that shared materials are unavailable, incomplete or inconsistent with the article, the Editors will contact the corresponding author and may, depending on the outcome, publish a correction, an editorial expression of concern or, in serious cases, a retraction, following COPE flowcharts.
13. Support for authors
Authors with questions about repository selection, anonymisation or data documentation may contact the editorial team at revista.andinaeducacion@uasb.edu.ec. The Andean Journal of Education is committed to ensuring that this policy does not become a barrier for researchers working with limited institutional infrastructure and will provide guidance where needed.
