Subtopic Deep Dive

Research Reproducibility and Reproducibility Crisis
Research Guide

What is Research Reproducibility and Reproducibility Crisis?

Research reproducibility crisis refers to the widespread failure to replicate biomedical study findings due to p-hacking, HARKing, publication bias, and poor reporting standards.

Low replication rates plague health research, with meta-research revealing systemic flaws in methods and incentives (Ioannidis, 2018, 277 citations). Checklists and pre-registration improve reporting quality (Han et al., 2017, 129 citations). Over 1,000 papers address reproducibility in preclinical studies since 2015.

15
Curated Papers
3
Key Challenges

Why It Matters

Reproducibility crisis undermines evidence-based medicine, as non-replicable findings lead to wasted resources and flawed treatments. Ioannidis (2018) shows meta-research exposes biases inflating false positives in clinical trials. Bero and Grundy (2016) argue nonfinancial interests bias results as much as funding, affecting drug approvals. Karp and Reavey (2018) highlight sex bias in preclinical work skewing human trial translations, with reforms like BCPT policies (Tveden-Nyborg et al., 2023) boosting reliability.

Key Research Challenges

P-hacking and HARKing

Researchers manipulate data post-hoc (p-hacking) or hypothesize after results (HARKing), inflating false positives. Ioannidis (2018) quantifies how incentives drive this in biomedicine. Pre-registration counters it but adoption lags.

Publication Bias

Journals favor positive results, burying null findings and distorting meta-analyses. McDowell et al. (2015) note tight funding exacerbates selective reporting in health research. Registered Reports mitigate this by accepting methods pre-results.

Poor Reporting Standards

Incomplete methods and data omission hinder replication. Han et al. (2017) found checklists raise reporting quality in preclinical papers. Michel et al. (2019) push statistical guidelines, yet compliance varies.

Essential Papers

1.

BCPT 2023 policy for experimental and clinical studies

Pernille Tveden‐Nyborg, Troels K. Bergmann, Niels Jessen et al. · 2023 · Basic & Clinical Pharmacology & Toxicology · 289 citations

The manuscript is an updated version of the previously published "BCPT policy for experimental and clinical studies" from 2021.

2.

Meta-research: Why research on research matters

John P. A. Ioannidis · 2018 · PLoS Biology · 277 citations

Meta-research is the study of research itself: its methods, reporting, reproducibility, evaluation, and incentives. Given that science is the key driver of human progress, improving the efficiency ...

3.

Sex bias in preclinical research and an exploration of how to change the status quo

Natasha A. Karp, Neil Reavey · 2018 · British Journal of Pharmacology · 157 citations

There has been a revolution within clinical trials to include females in the research pipeline. However, there has been limited change in the preclinical arena; yet the research here lays the groun...

4.

Co-existing Notions of Research Quality: A Framework to Study Context-specific Understandings of Good Research

Liv Langfeldt, María Nedeva, Sverker Sörlin et al. · 2019 · Minerva · 134 citations

5.

A checklist is associated with increased quality of reporting preclinical biomedical research: A systematic review

SeungHye Han, Tolani F. Olonisakin, John P. Pribis et al. · 2017 · PLoS ONE · 129 citations

Irreproducibility of preclinical biomedical research has gained recent attention. It is suggested that requiring authors to complete a checklist at the time of manuscript submission would improve t...

6.

Why Having a (Nonfinancial) Interest Is Not a Conflict of Interest

Lisa Bero, Quinn Grundy · 2016 · PLoS Biology · 125 citations

A current debate about conflicts of interest related to biomedical research is to question whether the focus on financial conflicts of interest overshadows "nonfinancial" interests that could put s...

7.

New Author Guidelines for Displaying Data and Reporting Data Analysis and Statistical Methods in Experimental Biology

Martin C. Michel, Teri J. Murphy, Harvey Motulsky · 2019 · Journal of Pharmacology and Experimental Therapeutics · 96 citations

Reading Guide

Foundational Papers

Start with Ioannidis (2018) for meta-research overview and McDowell et al. (2015) for junior scientist perspectives on funding pressures driving irreproducibility.

Recent Advances

Study Tveden-Nyborg et al. (2023) BCPT policy updates and Horbach (2020) on pandemic peer-review changes impacting reproducibility.

Core Methods

Core techniques include pre-registration, reporting checklists (Han et al., 2017), statistical guidelines (Michel et al., 2019), and Registered Reports to curb biases.

How PapersFlow Helps You Research Research Reproducibility and Reproducibility Crisis

Discover & Search

Research Agent uses searchPapers and exaSearch to find reproducibility papers like 'Meta-research: Why research on research matters' by Ioannidis (2018); citationGraph maps influence of BCPT policy (Tveden-Nyborg et al., 2023); findSimilarPapers uncovers related bias studies.

Analyze & Verify

Analysis Agent applies readPaperContent to extract methods from Han et al. (2017) checklist paper; verifyResponse with CoVe checks replication claims; runPythonAnalysis runs meta-analysis simulations on p-hacking effects using GRADE for evidence grading in biomedical reproducibility studies.

Synthesize & Write

Synthesis Agent detects gaps like low pre-registration in sex bias papers (Karp and Reavey, 2018); Writing Agent uses latexEditText, latexSyncCitations for reform proposals, latexCompile for reports, exportMermaid for bias flowcharts.

Use Cases

"Analyze p-hacking prevalence in recent preclinical trials"

Research Agent → searchPapers → Analysis Agent → runPythonAnalysis (pandas simulation of p-values) → statistical verification output with effect sizes.

"Draft Registered Report on sex bias reproducibility"

Synthesis Agent → gap detection → Writing Agent → latexEditText + latexSyncCitations (Ioannidis 2018) + latexCompile → formatted LaTeX manuscript.

"Find code for reproducibility checklists in biomedicine"

Research Agent → paperExtractUrls → Code Discovery → paperFindGithubRepo → githubRepoInspect → executable checklist validation scripts.

Automated Workflows

Deep Research workflow conducts systematic reviews of 50+ reproducibility papers, chaining searchPapers → citationGraph → GRADE grading for structured crisis reports. DeepScan's 7-step analysis verifies claims in Ioannidis (2018) with CoVe checkpoints and runPythonAnalysis on bias data. Theorizer generates reform theories from McDowell et al. (2015) and Tveden-Nyborg et al. (2023).

Frequently Asked Questions

What defines the reproducibility crisis?

It is the low replication rate of biomedical findings due to p-hacking, publication bias, and poor reporting (Ioannidis, 2018).

What methods address it?

Pre-registration, Registered Reports, and checklists like those in Han et al. (2017) and BCPT policy (Tveden-Nyborg et al., 2023).

What are key papers?

Ioannidis (2018, 277 citations) on meta-research; Han et al. (2017, 129 citations) on checklists; Karp and Reavey (2018, 157 citations) on sex bias.

What open problems remain?

Low adoption of guidelines (Michel et al., 2019) and nonfinancial bias (Bero and Grundy, 2016) persist despite tools.

Research Health and Medical Research Impacts with AI

PapersFlow provides specialized AI tools for your field researchers. Here are the most relevant for this topic:

Start Researching Research Reproducibility and Reproducibility Crisis with AI

Search 474M+ papers, run AI-powered literature reviews, and write with integrated citations — all in one workspace.