Lee Jussim argues that if a claim appears only as a peer‑reviewed paper, chapter, or conference presentation in psychology, you should provisionally disbelieve it until independent replications accumulate. He assembles an equation that adds unreplicable findings (~50% by his account) plus overclaiming, citation of bad work, censorship and fabrication to justify an approximate 75% false‑claim rate.
— If true, the claim forces media, policymakers, clinicians, and funders to change how they treat single psychology studies — privileging replication, preregistration, and evidence‑synthesis before action.
Arnold Kling
2026.05.04
45% relevant
The article points to a recent obedience 'obedience lite' study (N=80) that reproduces Milgram‑style high compliance rates and reports experimenter‑gender null effects (88% vs 90%), which qualifies as empirical pushback to broad claims that classic social‑psychology findings do not replicate.
Paul Bloom
2026.04.08
78% relevant
Bloom documents a repetitive, formulaic class of age‑comparison studies that produce incremental, often trivial findings—exactly the sort of work that raises concerns about low evidential value and contributes to psychology's replication and reliability problems; actor: Paul Bloom (Cognitive Development Society conference observations).
2026.04.04
100% relevant
The essay’s empirical pivot is a cited run‑rate of ~50% unreplicable findings and the author’s derived estimate that roughly 75% of psychology claims are false.
← Back to all ideas