Phase-Locking Value (PLV)
The phase-locking value quantifies how consistent the phase difference between two narrowband signals is across trials (or across a time window), independent of their amplitudes. After extracting the instantaneous phase of each signal (via the Hilbert transform or a wavelet), one forms the phase difference at each sample, represents it as a unit-length complex phasor, and averages; the magnitude of that average is the PLV, ranging from 0 for a uniformly scattered phase relationship to 1 for a perfectly fixed one.
PLV isolates phase coupling from power co-fluctuation, which is exactly what a coherence estimate blends together, making it a favoured index of putative oscillatory communication. Its central weakness is that a genuine zero phase difference is indistinguishable from the zero-lag mixing produced by volume conduction and a shared reference, so a high PLV, on its own, is not evidence of interaction between distinct sources.
Frontal-midline theta phase-locking between prefrontal and parietal sensors that rises during a working-memory delay is often reported as PLV, but the claim only holds after confirming the coupling survives a zero-lag-insensitive metric.
PLV needs a volume-conduction control to mean anything.