The base‑rate problem refers to the difficulty of determining how common or rare a linguistic feature is within the relevant population of writers, a gap that can lead to “methodological subjectivity and bias” when an examiner relies on intuition alone. In forensic authorship attribution it is addressed by obtaining “Base Rate Knowledge” – data from specially designed corpora that reveal the population‑level distinctiveness of particular style markers or idiolectal features. Researchers use such corpora (e.g., the Enron email corpus) to calculate how frequently specific markers occur, thereby grounding the significance of those markers in empirical base‑rate information. The problem remains a cautionary point: without appropriate population data, the significance of a feature cannot be reliably assessed, limiting the objectivity of stylistic analysis.
Definition drawn from An Introduction to Forensic Linguistics: Language in Evidence (Coulthard, Johnson & Wright, 2nd ed.). Extracted text: archive:books/an_introduction_to_forensic_linguistics_language_in_evidence/an_introduction_to_forensic_linguistics_language_in_evidence.txt.
Bears on: authorship & stylometry · the letters · sources