p-hacking

English

Blend of p-value +‎ hacking, from "p" for probability. Coined by researchers Uri Simonsohn, Leif D. Nelson, and Joseph P. Simmons in 2014.^[1]

p-hacking (uncountable)

(sciences, statistics) The practice of reanalysis of data until a desired result is obtained which is statistically significant
Synonym: data dredging
- [2016, Ronald L. Wasserstein, Nicole A. Lazar, “The ASA Statement on p-Values: Context, Process, and Purpose”, in The American Statistician, volume 70, number 2, →DOI:
  Cherry-picking promising findings, also known by such terms as data dredging, significance chasing, significance questing, selective inference, and “p-hacking,” leads to a spurious excess of statistically significant results in the published literature and should be vigorously avoided.]

^ Uri Simonsohn, Leif D. Nelson, Joseph P. Simmons (2013 April 24) “P-Curve: A Key to the File Drawer”, in Journal of Experimental Psychology: General, Forthcoming, volume 143, number 2, →DOI, page 534:
While collecting and analyzing data, researchers have many decisions to make, including whether to collect more data, which outliers to exclude, which measure(s) to analyze, which covariates to use, and so on. If these decisions are not made in advance but rather are made as the data are being analyzed, then researchers may make them in ways that self-servingly increase their odds of publishing (Kunda, 1990). Thus, rather than placing entire studies in the file-drawer, researchers may file merely the subsets of analyses that produce nonsignificant results. We refer to such behavior as p-hacking.