Lindley, D. V. (1956). On a measure of the information provided by an experiment. The Annals of Mathematical Statistics, 27(4), 986–1005. doi:10.1214/aoms/1177728069.
Not read: a lawful open copy exists but could not be fetched on 2026-10-27 (see the status note). The summary is drawn from what the work is standardly cited for and from the exchange’s description of it. A reading should check it first.
Why this is here
A16 §7 cited it as the foundation of expected-information-gain design, and drew the limit that matters here: information gain is relative to the hypothesis space already specified, so “The inquiry can become increasingly precise without becoming correct.” The record holds the later review (LIT-1525). It bears on CLAIM-470.