You wrote the introduction and methods, defended them in peer review, and got a promise of publication before you had collected a single data point. Months later the data are in, and the main hypothesis was not supported.
That is exactly the situation a Registered Report was designed for, and it changes how you should talk about the paper. The credibility comes from the plan, not the result, so the story you tell has to carry the plan with it.
What Makes a Registered Report Different
At Nature, a Registered Report is described as a format for empirical articles in which the hypotheses, methods and planned analyses undergo peer review before the research is conducted. If the Stage 1 proposal passes review, it can be granted in-principle acceptance: the study will be publishable provided the authors register the protocol, follow it, meet the quality checks and offer a sound interpretation.
At Stage 2, Nature says that after completing the study exactly as approved, the article is published regardless of outcome or statistical significance, pending checks on protocol adherence, any exploratory analyses and the interpretation. The Center for Open Science makes the same point from the editor's side: editorial decisions are made independently of whether the results support the pre-registered hypotheses.
That design has a measurable footprint. In psychology, Scheel, Schijen and Lakens compared 71 Registered Reports published by November 2018 with a random sample of 152 standard hypothesis-testing studies. Looking at the first hypothesis of each article, they found positive results in 96% of standard reports but only 44% of Registered Reports, and suggest that a plausible factor is reduced publication bias and/or Type I error inflation. If your main hypothesis was not supported, you are in large company.
Two Moments Worth Announcing
A standard paper has one launch. A Registered Report has two, and they need different messages. The table below uses Nature's process as described at the time of writing (November 2026); other journals run their own versions, so check yours.
| Moment | What exists publicly | What to share | What not to claim |
|---|---|---|---|
| Stage 1 in-principle acceptance | Your registered protocol, if you registered it publicly rather than under embargo | The question, why it matters, the design and the sample you plan to collect | Any expected result, or that the paper is "published" |
| Stage 2 publication | The final article, with a Protocol Registration section linking the approved protocol | The answer to the registered question, then any exploratory findings, clearly labelled | That an exploratory finding was the planned test |
Nature notes that Stage 1 protocols are not published in the journal; authors register them in a recognised repository, either publicly or under embargo. If yours is under embargo, the Stage 1 announcement can wait.
If you do announce at Stage 1, it is a good moment for early career researchers in particular. The Center for Open Science points out that a PhD student can gain in-principle acceptance for several papers before submitting their thesis. A short post about the question and the design shows your work long before results exist.
Keep the Confirmatory and the Exploratory Apart
The single most useful habit when promoting a Registered Report is keeping the two kinds of result separate. Nature asks authors to report the outcome of all registered analyses, and to put any additional analyses in a separate Results section titled "Exploratory analyses", justified and caveated. It also warns authors not to base their conclusions entirely on statistically significant post hoc analyses.
Promotion is where that line tends to blur. The exploratory result is often the more surprising one, so it drifts into the headline, and readers come away thinking it was the tested prediction. The Center for Open Science is clear that the format places no restriction on reporting exploratory analyses; it simply requires that readers can tell which analyses were confirmatory and which were post hoc. Your summaries should make the same distinction.
A Worked Example
Say you ran a Registered Report in psychology testing whether a short gratitude-writing exercise reduces exam anxiety in first-year students. The registered test found no meaningful difference. An exploratory analysis suggested a possible effect among students who reported high anxiety at baseline.
- Misleading: "Gratitude writing reduces exam anxiety in highly anxious students." The planned test is gone, and an exploratory subgroup reads as the finding.
- Accurate but lifeless: "The pre-registered hypothesis was not supported; exploratory analyses were conducted." True, and nobody will read past it.
- Accurate and readable: "In a study whose design was peer reviewed before we collected any data, a short gratitude exercise did not reduce exam anxiety in first-year students. An unplanned analysis hints at an effect in the most anxious students, which would need its own test."
The third version leads with the registered answer, explains why it can be trusted, and labels the subgroup for what it is. When you share numbers, use the effect sizes and confidence intervals from the paper. Nature, for example, requires exact p values, effect sizes and confidence intervals for all inferential analyses when authors report significance tests, so in a paper like that they are already there. For more on wording, the guide to communicating null results goes deeper.
A Checklist Before You Share the Final Paper
- Open with the registered question and the answer, in that order.
- Say in one clause that the design was reviewed before data collection.
- Give the main effect with its confidence interval, not a significance verdict.
- Label any exploratory finding as unplanned and in need of its own test.
- Link the paper of record and the registered protocol.
- If you shared data and code, link them too; the guide to promoting a research dataset covers how.
Where Amplify Fits
A Registered Report depends on the order of its claims: registered question first, answer second, exploration last. Amplify generates every format from the PDF you upload and nothing else, with no open-web filler, so the video abstract, vertical short, audio brief and infographic start from the paper where that order is already set.
Each format arrives as three variants you can edit line by line, so if an exploratory result creeps into the opening line, you move it back. A subject-matter reviewer reads every draft before it can leave your workspace, every claim is traced to a page in the PDF, and nothing publishes until you approve it. Every artifact carries the title, the authors and a link back to the paper of record.
The embeddable widget puts every format beside the paper on your lab site or faculty profile, where readers can read, listen and chat with the paper. The analytics then show reads, listens, questions asked and which format moved people, which is useful when you report on the work later.
Registered Reports ask researchers to commit before they know the answer. Promoting one well means keeping that commitment visible, so readers can trust the result whichever way it went.
