A methods section explains what a paper says was done. A prior protocol or registration can also show what researchers intended to do before the relevant results were known. Reading these documents together helps identify planned analyses, justified changes and questions that emerged during the work.
Separate the protocol, registration and final report
A protocol can describe a study in considerable detail, but the word protocol does not establish when that version was written. Look for a version, date and stable record, together with information about what data had already been collected or accessed.
OSF describes registrations as time-stamped, read-only versions of a project. This gives readers a fixed record to compare with later work; the content still needs to specify the decisions relevant to the claim being evaluated.Center for Open Science — Open Science Framework (opens in a new tab)
| Record | Question it helps answer |
|---|---|
| Prior plan | What was intended before the relevant results were known? |
| Amendment | What changed, when and why? |
| Final report | What was actually done and reported? |
| Analysis files | How were those reported results produced? |
A short registration containing only a broad research question may leave outcomes, exclusions and analyses open. Its existence is useful information, but it cannot answer details that it never recorded.
Audit the decisions that shape the main claim
Start with the primary question and outcome. Then compare the planned comparison, sample-size approach, inclusion and exclusion rules, and analysis with the corresponding sections of the paper. Keep the comparison focused on decisions that affect interpretation.
In an original hypothetical example, a plan identifies cellular readout X as primary, while the paper headlines readout Y. The first task is to locate the explanation: was Y an added exploratory analysis, a documented replacement after a measurement problem, or an unexplained change?
Those possibilities should not be collapsed into one judgement. A transparent change can be scientifically reasonable, while presenting an outcome selected after inspection as the original primary outcome conceals a consequential choice.
Interpret the date in relation to data access
A plan dated before publication is not necessarily a plan made before analysis. Publication often follows data collection and inspection by a considerable interval. The date needs to be connected to the investigators’ access to the relevant information.
Existing datasets make this particularly visible. Data can predate a new research question, while the analysts may or may not have examined the outcomes of interest. A useful record explains what was already known and what choices were still being committed in advance.
An amendment should retain its place in the sequence. Replacing an old file with a revised document without preserving the earlier version makes the history harder to assess, even if the final methods are described clearly.
If a link is inaccessible, avoid inferring its contents from the paper’s statement that registration occurred. Report the registration claim and the access limitation separately until the actual record can be inspected.
Distinguish a record from a reviewed design
Registered Reports add prospective assessment of a study’s question and methods through a journal process. The Center for Open Science describes initial review before the results stage, followed by assessment of the completed work; acceptance remains subject to the journal’s conditions.Center for Open Science — Registered Reports (opens in a new tab)
Ordinary registration does not necessarily include that review. Conversely, a reviewed protocol does not guarantee successful execution, complete reporting or a particular result. Read the actual process rather than treating similar labels as interchangeable.
An evidence note can finish with three items: where the plan is recorded, which main decisions match, and which changes affect interpretation. That gives readers an inspectable account without turning registration into a quality score.
Sources and further detail
- Center for Open Science — Open Science Framework (opens in a new tab)
Official description of time-stamped, read-only registrations read. Document comparison and the X/Y example are original.
- Center for Open Science — Registered Reports (opens in a new tab)
Stage-one/stage-two process and acceptance conditions read. Described as a journal format, not a universal requirement or guarantee of publication.
Sources checked 19 September 2026. Worked examples are illustrative unless a supplied report is explicitly identified. This article has not undergone independent scientific peer review.