CoChat

Initial evidence of research quality of registered reports compared with the standard publishing model

Open access

Published version of a preprint

The DOI matches its Crossref record; 47 of the 47 references it cites were found in a registry or an index; 18 of the 54 sources its claims cite were checked for support.

Abstract

In registered reports (RRs), initial peer review and in-principle acceptance occur before knowing the research outcomes. This combats publication bias and distinguishes planned from unplanned research. How RRs could improve the credibility of research findings is straightforward, but there is little empirical evidence. Also, there could be unintended costs such as reducing novelty. Here, 353 researchers peer reviewed a pair of papers from 29 published RRs from psychology and neuroscience and 57 non-RR comparison papers. RRs numerically outperformed comparison papers on all 19 criteria (mean difference 0.46, scale range -4 to +4) with effects ranging from RRs being statistically indistinguishable from comparison papers in novelty (0.13, 95% credible interval [-0.24, 0.49]) and creativity (0.22, [-0.14, 0.58]) to sizeable improvements in rigour of methodology (0.99, [0.62, 1.35]) and analysis (0.97, [0.60, 1.34]) and overall paper quality (0.66, [0.30, 1.02]). RRs could improve research quality while reducing publication bias and ultimately improve the credibility of the published literature.

Overview

CoChat Summary

TL;DR

In a naturalistic comparison, 353 researchers peer reviewed paired Registered Reports (RRs) and standard-model articles, rating RRs higher on all 19 evaluated criteria. RRs showed the largest advantages in methodological and analysis rigor, with little evidence of costs to novelty or creativity.

Key findings

  • Across 29 RRs and 57 matched comparison articles, RRs outperformed comparison papers on all 19 criteria, with an average difference of 0.46 on a −4 to +4 scale and a range of 0.13 to 0.96; 12 of 19 criteria had 95% credible intervals excluding zero.
  • Largest advantages were for rigor of methodology (0.99, 95% CrI [0.62, 1.35]) and analysis (0.97, [0.60, 1.34]), and overall paper quality (0.66, [0.30, 1.02]); smallest differences were novelty (0.13, [−0.24, 0.49]) and creativity (0.22, [−0.14, 0.58]).
  • Objective coding showed more preregistration (59% of RRs vs 7% author-based and 0% journal-based comparisons), open data (86% vs 18% and 14%), open materials (59% vs 11% and 7%), and larger median sample sizes (220 vs 118 and 78).
  • Between-subjects analysis using only the first reviewed article showed a similar pattern, with an average effect size of 0.43 and range 0.04 to 0.93.

Methods

  • Observational quasi-experimental study using 29 published RRs in psychology and neuroscience reporting novel research (replication RRs excluded) and 57 comparison articles matched by journal/date or lead author/date, totaling 86 articles.
  • 353 recruited researchers, matched to articles by self-reported expertise, each peer reviewed one RR and one comparison article in randomized order; they rated 19 criteria on bipolar 9-point scales from −4 to +4.
  • Preregistered Bayesian multilevel models estimated within-subjects and between-subjects effects; article coding assessed sample size, sample justification, difficult-to-reach populations, study design, open practices, and statcheck-detected reporting errors.

Limitations

  • The paper states that the naturalistic design lacked random assignment of research to RR versus standard publication, and authors/editors self-select into RRs; matching on topic, journal, and lead author was intended to reduce but not eliminate confounding.
  • The paper also states that RRs are a new format, so early-adoption behavior may not generalize; quality was assessed via peer-review ratings rather than replicability or generalizability, and reviewer representativeness is unknown.
  • Evident from the text: reviewers were not fully blinded to article type, with 54% correctly identifying the RR and 54% correctly identifying the comparison paper as not an RR; exploratory analyses showed larger perceived RR advantages among reviewers more familiar with or more favorable toward RRs, though every subgroup estimate remained directionally favorable.

Model-written summary. Check it against the paper before relying on it.

Integrity signals

Registry

Crossref · 10.1038/s41562-021-01142-4

Cites

47 of 47 references verified0 citation errors · 0 not checkable

Statistics

Nothing to recompute.

Figures

Figure 1 Posterior probability distributions for parameter estimates from within-subjects analysis using partial-pooling across the 19 outcomes comparing the difference of Registered Reports (RRs) and comparison articles. Black horizontal lines indicate 80% (thick) and 95% (thin) credible intervals and dots show the mean of the posteriors. Positive values indicate a performance advantage for RRs, negative values indicate a performance advantage for comparison articles. Outcomes are sorted from strongest to weakest performance advantage for RRs in the three evaluation groups: Before knowing study outcomes, after knowing study outcomes, and after finishing the paper.
Figure 1 Posterior probability distributions for parameter estimates from within-subjects analysis using partial-pooling across the 19 outcomes comparing the difference of Registered Reports (RRs) and comparison articles. Black horizontal lines indicate 80% (thick) and 95% (thin) credible intervals and dots show the mean of the posteriors. Positive values indicate a performance advantage for RRs, negative values indicate a performance advantage for comparison articles. Outcomes are sorted from strongest to weakest performance advantage for RRs in the three evaluation groups: Before knowing study outcomes, after knowing study outcomes, and after finishing the paper.From the PDF
Image 2
From the PDF
Image 3
From the PDF
Image 4
From the PDF
Image 5
From the PDF
Image 6
From the PDF
Image 7
From the PDF
Image 8
From the PDF
Image 9
From the PDF
Image 10
From the PDF
Image 11
From the PDF

Verification

Claims the paper makes

8 Valid · 1 Not valid · 12 Not checked · 165 Own results and interpretations

46 more claims

Create a free account to see all 186 claims and every source verdict.

See all claims — free account
  • Probably not a substantial claim. 116
  • It may state two separate claims as one. 42
  • The quote only partly states this claim. 10

Create a free account to see these claims.

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

Each contradiction between two passages listed here was confirmed by an independent check.

2 of 2

Citations

Cites

Show all 47 references

Cited by

See all 295

Metadata

Authors
Soderberg, C. K., Errington, T. M., Schiavone, S. R., Bottesini, J., Thorn, F. S., Vazire, S., Esterling, K. M., & Nosek, B. A.
Affiliations
  • Center for Open Science (US)
  • University of California, Davis (US)
  • The University of Melbourne (AU)
  • University of California, Riverside (US)
Corresponding author
Nosek, Brian A.
Venue
Nature Human Behaviour
Published
2021-06-24
Volume
5
Issue
8
Pages
990-997
ISSN
2397-3374 (ISSN-L)
OpenAlex
W3174993952
Cited by
295
References
41
Citations by year
2026: 27 · 2025: 47 · 2024: 43 · 2023: 61 · 2022: 85 · 2021: 28
FWCI
39.83
Type
journal article
Language
en
Open access
Licence
Locations
  • Nature Human Behaviour (publishedVersion)
  • MetArXiv (OSF Preprints) (submittedVersion, public-domain)
  • PubMed (publishedVersion)
  • RePEc: Research Papers in Economics (acceptedVersion)
  • MetArXiv (OSF Preprints) (acceptedVersion)
Funders
  • James S. McDonnell Foundation (220020498)
  • Arnold Ventures, no reference number
Topics
  • Social Sciences › Psychology › Experimental and Cognitive Psychology › Psychology Research and Bibliometrics
  • Social Sciences › Library and Information Sciences › Academic Publishing and Open Access
  • Social Sciences › Decision Sciences › Statistics, Probability and Uncertainty › Meta-analysis and systematic reviews
Keywords
research quality, publication bias, research credibility, novelty, creativity, methodological quality, open science
MeSH terms
Data Analysis, Humans, Neurosciences, Psychology, Registries, Research, Research Design, Peer Review, Research, Research Report

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

The claim verdicts this report counts are experimental: their accuracy is not yet measured.

A language model drew the claims from the paper on 2026-10-10, under version 3 of the claim definition (model deepseek-v4.1-flash, prompt extract-v3).

A second model labelled each claim on 2026-10-10 (model jev-1.13-20260917, definition revision gold-guideline-v0.2).

Cited sources are checked with a language model’s help (gpt-6-sol).

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review

How to cite

Cite CoChat for the checks on this page, and the paper for its own findings.

This page
CoChat. (2026). Citation check: Initial evidence of research quality of registered reports compared with the standard publishing model. https://app.cochat.ai/papers/10.1038/s41562-021-01142-4
The paper
Soderberg, C. K., Errington, T. M., Schiavone, S. R., Bottesini, J., Thorn, F. S., Vazire, S., Esterling, K. M., & Nosek, B. A. (2021). Initial evidence of research quality of registered reports compared with the standard publishing model. Nature Human Behaviour. https://doi.org/10.1038/s41562-021-01142-4

Data sources

Bibliographic data: Semantic Scholar, OpenAlex, Crossref.

Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).

Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.