unsupervised adaptation analysis

Description Usage Format

These are based on two sources: Kronrod, Copress, and Feldman (2012, CogSci) used a combination of categorization and discrimination data to estimate underlying mean and variance for /b/ and /p/, assuming a mean of 60ms for /p/ (which is based on Lisker & Abramson, 1964). Goldrick, Vaughn, and Murphy (2013, JASA) analyzed VOTs produced by 12 talkers producing 80 monosyllabic words with word-initial voiced stops (only /b/ analyzed here). Many talkers prevoiced some or most of their /b/s, and these are included separately here.

1	prior_stats

A data frame with 4 observations of 5 variables:

prevoiced: Whether or not stats are for prevoiced stops (only for goldrick et al. data
mean: Mean VOT (in ms) for this category
sd: VOT standard deviation (in ms; inferred or sample)
category: 'b' or 'p'
source: Source of this estimate ('goldrick2013' or 'kronrod2012')