Informally, a p-value gives the probability, under a specified statistical model, that a statistical summary of the data would be at least as extreme as the value actually observed. [grounded] corpus source (“Informally, a p-value is the probability under a specified statistical model that a statistical summary of the data (for example, the sample mean difference between two compared groups) would be equal to or more extreme than its observed value.”)
Against a specified statistical model, a p-value can indicate how incompatible the data are with it. [grounded] corpus source (“P-values can indicate how incompatible the data are with a specified statistical model.”)
[Agent synthesis — not in the cited work] A p-value measures neither the probability that the studied hypothesis is true nor the probability that the data arose by random chance alone. [grounded] corpus source (“P-values do not measure the probability that the studied hypothesis is true, or the probability that the data were produced by random chance alone.”)
Scientific conclusions and business or policy decisions should not be based only on whether a p-value passes a specific threshold. [grounded] corpus source (“Scientific conclusions and business or policy decisions should not be based only on whether a p-value passes a specific threshold.”)
Inference is proper only with full reporting and transparency. [grounded] corpus source (“Proper inference requires full reporting and transparency.”)
[Agent synthesis — not in the cited work] Neither a p-value nor statistical significance measures an effect's size or a result's importance. [grounded] corpus source (“A p-value, or statistical significance, does not measure the size of an effect or the importance of a result.”)
A result can clear a threshold and still be trivial; a result can miss it and still matter. The threshold is a reporting convention, not a property of the world. [author assertion]
[Agent synthesis — not in the cited work] Read together, the ASA's principles describe a discipline: report the estimate and its uncertainty, show the whole analysis rather than the surviving fragment, and never let a single number carry a verdict it cannot support. [synthesis]
This publication intends to keep that discipline. Every quantitative claim we publish carries its provenance for exactly this reason: a reader who can re-check the chain from source to sentence does not have to take our word — or a p-value's — for anything. [author assertion]
