Landmark study shows data sharing key to reproducibility across social and behavioural sciences
Experimental psychologist Dr Arran Reader from the University of Stirling’s Faculty of Natural Sciences was part of an international team of researchers from over 100 institutions involved in the study
/filters:format(webp)/filters:no_upscale()/prod01/cdn/media/stirling/news/news-centre/2026/apr-26/1200X630JournalArticle.jpg)
The majority of research results in social and behavioural sciences can be reproduced when data and code are provided, but most papers still do not share them – a new study has shown.
Investigating the reproducibility of the social and behavioural sciences, published on April 1 in Nature, provides the most comprehensive assessment to date of reproducibility in the social and behavioural sciences. Reproducibility is defined as whether the same results can be obtained by re-running the same analyses on the same data.
This is distinct from replicability, which tests the same question with new data, and robustness, which tests it using alternative analyses on the same data.
Experimental psychologist Dr Arran Reader from the University of Stirling’s Faculty of Natural Sciences was part of an international team of researchers from over 100 institutions involved in the study.
He explained: “It is important that results reported in research accurately reflect the analyses conducted. By working together in a large-scale collaboration, it was possible to examine whether this was the case for 600 journal articles published in a range of fields.
“That three-quarters of the evaluated results could be precisely reproduced when data and code were available highlights the importance of sharing these resources. Whilst an inability to reproduce a previous finding does not mean that it is incorrect, it could indicate that the description of the analysis is incomplete or ambiguous.
/filters:format(webp)/filters:no_upscale()/prod01/cdn/media/stirling/news/news-centre/2026/apr-26/700X600AaronReader.jpg)
Lecturer in Psychology
Our findings suggest that code and data sharing might facilitate reproducibility, thus making it easier to independently verify research findings or correct mistakes. This is clearly also important for maintaining public confidence in research. It is therefore reassuring that data sharing is increasingly becoming a requirement for journals.
“Our findings suggest that code and data sharing might facilitate reproducibility, thus making it easier to independently verify research findings or correct mistakes. This is clearly also important for maintaining public confidence in research. It is therefore reassuring that data sharing is increasingly becoming a requirement for journals.”
The new investigation from the US Defence Advanced Research Projects Agency-funded Systematizing Confidence in Open Research and Evidence (SCORE) program showed that reproducibility cannot always be assumed.
The team examined 600 quantitative research papers published between 2009 and 2018 in 62 prominent journals spanning business, economics, education, political science, psychology, sociology, and other related fields.
They assessed data and code availability, whether authors made their datasets and analytic code accessible and reproducible and whether independent analysts could reproduce the reported statistical results when data was available.
Key Findings
Only 24% of papers made their data available, and just 20% shared both data and code. Without access to the data, independent assessment of reproducibility is not possible.
Where reproduction was possible it usually succeeded, but a substantial proportion did not. Among papers where reproduction could be attempted, around 72% were reproduced at least approximately, and 53% were reproduced precisely – meaning the numbers matched exactly.
Sharing data and code was shown to make a significant difference. When both data and code were available, reproducibility was very high with 88% of papers reproduced approximately and 75% precisely.
In contrast, when analysts had to reconstruct datasets from the original source, precise reproducibility fell to just 11%.
Some fields performed better than others. Political science and economics had substantially higher data availability and reproducibility rates than other disciplines.
An exploratory investigation suggests that this may have occurred because journals in these fields are more likely to require data sharing, code sharing, and reproducibility checks.
Transparency was shown to be improving across the social and behavioural sciences with the proportion of journals requiring data sharing increasing markedly.
Between 2018 (the last year of the sample of findings reproduced in this study) and 2025, the percentage of journals in the sample requiring data sharing increased from 27% to 52%, meaning that replicating the study on a more recent sample may show higher success rates.
What the results mean
A lack of shared data doesn’t mean a finding is wrong, but it does prevent independent verification.
Reproduction can fail for a range of routine reasons – including incomplete documentation, analyst error, or differences in software and data-processing steps.
At the same time, successful reproduction simply shows that the reported numbers match the underlying analysis; it does not test whether the theory is sound, the measures are valid, or the result holds up with new data or alternative methods. Reproducibility is therefore best understood as a basic accuracy check rather than a final judgement on scientific truth.
Impact
Scientific progress depends on cumulative evidence. When results cannot be independently verified, confidence in that evidence weakens; slowing learning, complicating policy translation, and eroding trust.
The study shows that there is substantial room for improvement in fundamental research behaviours to demonstrate that findings are reported precisely and can be verified independently. Making data and code available allows errors in reporting to be detected and corrected.
Dr Gemma Learmonth, who leads the Stirling Open Research and Scholarship Network, added: “Reproducibility is about continually improving how we design, document and share our research.
“At the University of Stirling, we have recently launched a cross-disciplinary network - Stirling Open Research and Scholarship (SORS) Network - alongside a monthly, student-led, ReproducibiliTea journal club, both of which aim to support students, researchers and scholars improve their research practices.”
Lead author Olivia Miske, Project Coordinator on the SCORE project at the Center for Open Science, concluded: “Improving reproducibility is not about mistrusting researchers. It is about recognizing that even careful scientists make mistakes. Openness may be an effective tool for quality control.”