The Nature Portfolio style places a relatively short Methods section as the final section of our manuscripts. For Nature Climate Change, this section is published online only, after the main text, and allows approximately 3,000 words. However, this position and suggested length should not be interpreted as a judgement on the importance of the Methods section.

Credit: pablohart / E+ / Getty Images

Getting the methods right is a key part of open, transparent and reliable science. As a first step, sufficiently detailed and clearly understandable methods are a requirement for our reviewers to both understand and assess the quality of the work. It is not uncommon that reviewers comment on missing or unclear methodological information in the first round of review, or insufficient access to data and code. These issues can effectively block the review process — preventing comprehensive assessment at the first round and adding multiple rounds of review to the timeline of a paper, while increasing author and reviewer burden.

We want to acknowledge that this may arise in part because our formatting guidelines seem to suggest that ‘shorter is better’. But while we do ask for succinctness, this should not come at the cost of information loss. As methods get more complex, and science more interdisciplinary, we urge our authors to take the time and space to ensure that their methods are fit for purpose. If the space in the Methods section is not sufficient, we strongly encourage more details to be provided in the Supplementary Information section.

The Methods section should explain in detail what was done, how it was done and why it was done (see our reporting standards). As a general guideline, a reader should be able to reproduce the results of the study if they have a basic understanding of the methods of the field. This involves the inclusion of information on data collection and location, and data processing, for example. Certain information — such as ethics approval for research with human participants and animals — is mandatory.

Writing methods also goes beyond just the Methods section itself and can require a bit of repetition. For example, to ease reading and to give the reader confidence in the approach and results, we request that a brief methods description (what was done and why) be included in the main text, and that caveats and limitations also be given some space. Moreover, independently of whether the information is already provided in the Methods section, all relevant manuscripts must include a separate data availability statement and a code availability statement that act as a one-stop shop for reviewers and readers to access all data and code used to produce the results of the paper. To facilitate the review process, data and code should also be made available to reviewers at the time of review. Note that there are some exceptions possible (for example, for sensitive data), which should be discussed with the editor, while deposition of some data types is mandatory. Before publication, data and code should be made publicly available in a persistent repository (for example, Figshare, github or Code Ocean) and assigned a doi that is cited in the Methods reference list (see the Springer Nature research data policy).

Outside the paper itself we generally ask that manuscripts reporting original data include a full reporting summary, and that a software and code checklist be completed for those with original code development. These are completed before the paper is sent to review, and updated before final publication of the work. Any metanalyses should also include ROSES (RepOrting standards for Systematic Evidence Syntheses) or PRISMA (Preferred Reporting Items for Systematic reviews and Meta-Analyses) workflows. Finally, given the recent and rapid rise of machine learning and its proliferation into different fields, the Nature Portfolio has developed a machine learning checklist, covering reproducibility, datasets, model and training, evaluation and computational resources used that authors may consider useful to ensure sufficient details for readers. Ultimately, these checklists and guidelines are founded on the understanding that good research should follow best practices, including in the reporting step.

Overall, completing the Methods section and associated documents and populating repositories is an investment of authors’ time. But we hope that providing these details in full is rewarded by facilitating a smoother assessment of the manuscript and ultimately increases the accessibility and reproducibility of the research.