Watch for
Initial exciting findings that later studies fail to replicate at the same magnitude.
Outcome & SelfBias #135
Decline Effect
Scientific effect sizes shrink over time.
The phenomenon where the effect size of a scientific discovery diminishes over time upon repeated replication attempts often because initial studies overestimate effects.
Why it matters: Central to the replication crisis in psychology medicine and other sciences.
Try this
Treat single studies as provisional and look at meta-analytic averages.
Real-world example
A groundbreaking psychology finding with a large effect size gradually shows smaller effects in subsequent replications.
Key researchers
John Ioannidis, Richard Lehrer
First described in 2005
Psychological mechanism
Multiple factors including publication bias p-hacking regression to the mean and genuine moderating effects that reduce effect sizes in broader samples.
Seminal research
John P. A. Ioannidis (2005), "Why most published research findings are false."
Related Biases
Debrief
Biases are not character flaws. They are recurring patterns in how minds compress uncertainty, save energy, and narrate reality. Once you recognise the pattern, you can slow the decision down, test the assumption, and make space for a better explanation.
Sources
Original ResearchIoannidis (2005) — Why most findings are false
EncyclopediaWikipedia overview