Here are a few tutorials I’ve written up over the years.
This is a walkthrough of how to make spectrograms in python that are customized for human speech research. The built-in functions
matplotlib.pyplot.specgram are both very general-purpose, so this tutorial defines a wrapper function to make the parameters more sensible for phoneticians and speech scientists.
This tutorial goes through some of the considerations around diphthong plotting (chiefly deciding how many time points to plot). There are R and python versions (both are Jupyter notebooks), which differ slightly but cover the same basic material. Originally presented at a meeting of the UW Phonetics Lab, 2017-11-03.
This tutorial uses R to go through a simple example of modeling a binary variable smapled over a period of time. Data points are assumed independent, so there is no “within-subjects” aspect; just an ordinary generalized linear model using a logit link. Code for how the data were simulated is provided at the end.
This is the tutorial vignette for the R package
phonR. It works better as a website than a PDF, and this way the package doesn’t take up extra space on CRAN’s servers.