/PCAphon

PCA for phonetics workshop folder

Welcome to the PCA workshop for phonetic research

Dear all, this is Maelle Amand. I'm a lecturer at the University of Paris Sorbonne Nouvelle, professeure agrégée and currently doing and interdisciplinary PhD in sociophonetics & statistics.

Aims of the workshop

At the end of the workshop you will:

  • have a general understanding of PCA
  • know the correspondence of PCA key words from one school of PCA to the next to understand results in publications
  • know how to organise your data table to use it for PCA
  • know the main functions of the R package {FactoMineR}
  • know how to interpret PCA graphs and correlation circles
  • find the relevant information in the loadings (PCA output) to answer your research questions
  • know how to perform a hierarchical cluster analysis based on PCA

Examples of your achievements at the end of the workshop:

PCA in language variation

PCA analysis performed on the percentage of FACE, GOAT and PRICE variants in the 1994 PVC corpus of Newcastle English (Amand, Ballier & Corrigan, 2018).

PCA in language variation_2

Cluster analysis based on the above PCA analysis (Amand, Ballier & Corrigan, 2018).

Get prepared before the workshop!

Before the workshop, please install the package FactoMineR in RStudio or R on your computer. It would be advisable to bring your own dataset(s).

During the workshop

  • We will go through the files as we proceed through the workshop.
  • A PCA analysis will be performed on several datasets related to the field of phonetics so that you may improve your interpretation of PCA results and graphs.
  • I will also add published articles using PCA analysis for further illustration.

Other workshop coming soon

  • 11th & 18th June: ggplot2 & multivariate analysis (Paris Sorbonne Nouvelle University)

Previous workshops I have given

  • RStudio for postgraduate linguistic research (Paris Sorbonne Nouvelle University)
    • Weekly course for one semester
    • Regular two-day workshops
  • Data analysis with the coding scheme of the PAC-project database on English varieties (Paris Diderot University)
  • Mixed effects models applied to phonetic data (Paris Diderot University)
  • Statistics for the humanities: learn to build statistical models for qualitative data (Paris Diderot University)
  • Regular one-on-one courses in statistics with RStudio.
  • Statistics in RStudio for linguistics: clinic session for PhD students (Newcastle University)