P-hacking is a human-made research practice in which analysts exploit flexibility in data analysis — trying multiple statistical models, subgroups, outlier exclusions, or variable transformations — selectively reporting only those results that achieve statistical significance. Its parameters are: (1) the flexibility in analytical choices available to the researcher, (2) the number of alternative analyses implicitly or explicitly tried, and (3) the selective reporting mechanism that privileges significant findings. The persistence mechanism is through methodological literature exposing the practice, reproducibility initiatives, preregistration standards, and statistical education that warns against it. [formal: p-malleabilitas | substrate: behavior | horizon: a life | explicit: yes | epoch: 0.01]
Accepted ontology entry
p-hacking
P-hacking is a human-made research practice in which analysts exploit flexibility in data analysis — trying multiple statistical models, subgroups, outlier exclusions, or variable transformations — selectively reporting only those results…
Definition
Why it is in scope
A human-made research practice of manipulating data analysis — trying multiple analyses, subgroups, or variable selections until a statistically significant result emerges. Built to persist through methodological literature, reproducibility research, and statistical education reform.
Names and aliases
- p-hackingen · CANONICAL
Relations from this entry
- cmrwbs6pa00lwkyo66ry5sxjfDEPENDS_ON →
Direction tested: significance-testing existed first (Neyman-Pearson framework, 1930s). Removal test: remove significance-testing and p-hacking ceases to operate — p-hacking is the manipulation of data analysis to achieve statistical significance; without significance-testing as a practice, the concept of p-hacking has no referent.
- cmrw9hpjd00fnkyo62okra5lpDEPENDS_ON →
p-hacking literally means manipulating p-values to achieve significance. Remove p-value and p-hacking has no operational mechanism — the term, practice, and detection methods (p-curves, p-curve analysis) all center on p-value manipulation. Concrete operational collapse.
Relations to this entry
No accepted relations in this direction.
Record identity
- Created
- Jul 22, 2026, 4:32 PM UTC
- Content hash
- 2b1828b509d47cc6e0de9c810236096252bcdc478b3c970f2f4011add2ab6a06