Modeling Issues in Speech and Hearing
A seminar on the theory and practice of scientific modeling, using hearing and speech research to ask how models are built, tested and judged.
What makes a scientific model good, and how would you tell? This seminar explores that question using research on hearing and speech biophysics. Students read groups of papers twice, first to master what was done and how, then to ask why, examining how models get built and tested and how theory relates to experiment. A few lectures add special topics on method.
What You Will Achieve
- Critique a scientific model on its assumptions and aims
- Compare analytic and computational modeling approaches
- Apply dimensional analysis to modeling problems
- Lead a discussion of a research paper
Program Overview
The Process of Model Building
The first weeks pose core contrasts between different ways of modeling.
- The process of model building
- Analytic vs. computational
- Detailed vs. heuristic
Dimensional Analysis and Resampling
Formal lectures introduce methods that apply across many sciences.
- Dimensional analysis and scaling
- Curve fitting, resampling, and Monte Carlo
- Chaos and the limits to prediction
- Molecular mathematical biology
Theory and Experiment
The term ends with further contrasts, students’ own selections and a final look at how theory meets data.
- Forward vs. inverse
- Student selections
- The relation between theory and experiment
Course Completion
Job Outlook
This way of thinking is useful in auditory science, speech research, biomedical modeling and any research field that relies on mathematical models.
Good for: Graduate students in hearing, speech or biophysics, and scientists in other fields who want to think harder about how they use models.
Keep in mind: It depends on discussion, and the reading list itself is not shown here, so it works best with a small group. Only assignments and notes on the seminar format are published.
Bottom line: An unusually reflective class on the craft of modeling, with a seminar format worth borrowing.
Frequently Asked Questions
What is a preparation sheet?
It is a page or so of comments, criticisms, observations and questions about a paper, written before class. Its purpose is to stimulate thought and improve discussion, not to create busywork.
How do the paired seminars work?
Two consecutive sessions usually cover each group of papers. The first aims at understanding what was done and how, and the second at integration and critique, focused on modeling issues.
What are the formal lectures on?
Four lectures cover dimensional analysis, resampling and Monte Carlo methods, spooky actions at a distance and quantum statistics, and the historical background of Hodgkin and Huxley's work.



There are no reviews yet.