Standard t-tests assume independent data points. Neural data is autocorrelated (tomorrow’s brain state is similar to today’s). The book introduces non-parametric permutation testing and cluster-based correction for multiple comparisons (via the FieldTrip toolbox).
In the world of electrophysiology, data is messy. Neural signals are a complex mixture of neuronal activity, muscle movements, line noise, and artifacts. Most academic papers present polished results, hiding the struggle of getting there. Standard t-tests assume independent data points
This is where Cohen’s book shines. It doesn't just show you the math; it teaches you the "why" and the "how." In the world of electrophysiology, data is messy
1. The Theory: The book provides an intuitive yet rigorous explanation of the mathematical foundations. It covers Fourier transforms, wavelets, and filtering in a way that is accessible to those who aren't pure mathematicians. It forces you to ask: Does this analysis actually answer my scientific question? This is where Cohen’s book shines
2. The Practice: Unlike many theoretical textbooks, this one is deeply practical. It walks through real-world issues like:
Websites claiming to offer the "free PDF download" (often found on ResearchGate, Academia.edu, or shadow libraries) come with caveats: