If you work on the speech signal and you are busy doing statistics on F0, formants or energy contours, then this site may be helpful for you.
But also if you do other kind of analyses on the speech signal or on other types of uni- or multi-dimensional continuous signals you may take inspiration from the material available here.
Functional Data Analysis (FDA)
is a set of modern techniques that allow you to perform statistical analysis on sets of curves.
FDA provides qualitative insight in the form of graphical display of results and
quantitative output in the well known form of p-values, percentage of explained variance, etc.
I like to describe FDA as an interface between shapes and numbers.
Analyzing curves is a common task in phonetics and phonology. For example, F0 contours are analyzed (and also modified) in intonation research. Usually a few quantitative descriptors (features) of such curves is extracted first (duration, peaks, max velocity, etc.) then ordinary statistics is applied on those descriptors. This feature extraction forces you to make hard decisions on what to retain and what to discard from the complex information encoded in the curve shapes. FDA allows you to bypass entirely the intermediate step of feature extraction and operate directly on you set(s) of curves.
Here you can download both published papers and unpublished tutorials on FDA for speech research written by me and my colleagues from different institutes. Even though the general FDA literature is of excellent quality and software tools are publicly available and well documented (see References), those resources are not easily accessible for scientists who do not have a solid background in math (linear algebra, calculus, advanced stats) as well as some programming skills (R or MATLAB). Moreover, so far FDA has been rarely used in speech research. The material provided here aims at filling those gaps to the best of our effort.
Do not hesitate to contact me for questions, suggestions, requests. I am happy to share all the code I write, and I am eager to learn from other FDA users within and outside the speech research community!