The deposited version of my dissertation in Spanish Linguistics. Please cite as:
Eager, Christopher D. (2017). Contrast preservation and constraints on individual phonetic variation. Doctoral thesis. Univserity of Illinois at Urbana-Champaign.
The two datasets analyzed in the dissertation can be found in the 'data' folder. The 'ptkbdg.rds' dataset is the same as the plosives dataset in the nauf package. If you re-analyze either of these datasets in a publication, please cite the dissertation (see above).
The methodology employed in the dissertation can be replicated using the standardize and nauf packages. The nauf package does not currently support Bayesian regressions; and even when it does, it will do so through interfacing with the rstanarm package, and so exactly replicating the dissertation will require a bit of work. I will add an R-markdown file showing how to use the Bayesian models in the 'Stan' folder to replicate the analysis.
See 'typos.md' for a current list of known typos and errors in the dissertation.