John P. A. Ioannidis

The paper maps scientific cosmopolitanism by analyzing Scopus data, finding 20.4% of top-cited scientists have substantial publication records in multiple countries.

This study examines gender differences in retraction rates among highly cited and all authors, finding no major gender disparities but field-specific variations.

This paper examines the uptake and implementation of multiverse-style analyses across 613 studies, finding widespread use primarily in psychology, social sciences, and medicine.

This study analyzes the scientific output of AI unicorn startups, finding minimal participation in formal scientific communication.

The paper maps and ranks institutions based on authors' impact and research integrity proxies, revealing disparities in rankings due to self-citation and retractions.

The paper analyzes global scientific publication trends, finding that most research originates from countries with low democracy and freedom of press ratings.

The paper evaluates the DrugAge database, finding poor reporting quality and inconsistent effect sizes in anti-aging preclinical studies across species.

This study reviews RCTs on polygenic risk score communication, finding no significant effects on health outcomes despite frequent positive claims.

The paper analyzes precocious citation impact, finding that top-cited authors in short timeframes often exhibit indicators of potential fraud or manipulation.

The paper estimates how much fMRI connectivity can be reliably detected given within-subject time course reliability and finds that less than a quarter of observed connectivity is reliably detectable.

This study examines the use of genetic and other omics information in highly cited clinical trials, finding that 22% of trials used such data, mostly in oncology, with significant industry involvement and limited transparency.

This study analyzes the involvement of high-impact researchers in leadership roles at top research universities and finds a significant underrepresentation of such researchers in administrative positions.

Linking 39,468 eligible Retraction Watch retractions to Scopus shows 3.3% of 217,097 career-long and 4.0% of 223,152 single-year (2023) top-2% cited scientists have at least one retraction.

The paper analyzes the rise of extreme publishing behavior across scientific fields and countries from 2000-2022, finding increasing prevalence and concerns about authorship standards.

The paper assesses the replication of genome-wide significant SNPs from earlier GWAS in the UK Biobank, finding an 85% overall replication rate with differences by phenotype type.

The paper evaluates sample size trends in neuroimaging studies from 1990-2012 and 2017-2018, finding increasing sample sizes and low use of pre-study power calculations.

This paper identifies differentially expressed microRNAs in Parkinson’s disease through meta-analyses of multiple datasets, finding significant associations in brain and blood samples.

The paper critiques null hypothesis significance testing (NHST) and advocates for alternative statistical methods in biomedical and psychological research.