Dispersion analysis

Dispersion analysis looks at how widely or evenly an item is distributed in a corpus. Indices express this information on a scale from 0 to 1, with higher values commonly denoting a more balanced distribution. Most measures rely on a division of the corpus into parts, and to ensure clarity of interpretation, these should represent meaningful contexts of language use. The increasing range of linguistic problems addressed with dispersion analysis can profit from methodological insights on the behavior of indices: Flaws have been detected in certain measures, and all of the indices reviewed in this chapter conflate dispersion and frequency.