Statistical methods in phonetic sciences

Organisers

Timo B. Roettger (University of Cologne, timo.roettger@uni-koeln.de)
Bodo Winter (University of Birmingham, bodo@bodowinter.com)

Workshop time and place

The workshop will be held in Cologne on June 11, 2017, as part of the Phonetics and Phonology in Europe (PaPE) conference.

Workshop description

Phonetics as a field has its own set of analysis methods and methodological assumptions that have grown in response to the demands of speech as a data source. As new statistical methods and experimental design ideas come in from other fields (data science, psycholinguistics, cognitive science), it is important to have an open discourse about the methodological basis of phonetics as a field. What insights from other fields can help improve methodological practice? In what way is phonetics unique and requires its own set of methods? In what ways can the practice of data analysis in phonetics be improved? In what way does statistics education in phonetics need to be altered to keep up with new developments in the field?

Workshop Schedule

9:30 – 9:40 Introduction
– Timo B. Roettger & Bodo Winter

9:40 – 10:05 A simulation-based comparison of approaches to significance testing using generalised additive mixed models
.
– Márton Sóskuthy

10:05 – 10:30 Statistics of vowel formants: Normality of distributions across the syllable.
– Doug Whalen

10:30 – 10:55 Fundamentals and applications of resampling methods for the analysis of speech production and perception data.
– Olivier Crouzet

10:55 – 11:25 Coffee Break

11:25 – 11:50 Bayesian Generalized Linear Mixed Effects Models: Accessible Bayesian inference with the rstanarm package
– Bradley Rentz

11:50 – 12:15 Gibbs Sampling as an Experimental Design: Prospects and Opportunities in Phonetics.
– Bill Thompson & Bart de Boer

12:15 – 12:40 Model selection and statistical argumentation.
– Morgan Sonderegger & James Kirby

12:40 – 13:00 Panel Discussion