A novel approach is suggested for detecting the presence or absence of sync
hronization between two or three interacting processes with different time
scales in univariate data. It is based on an angle-of-return-time map. A mo
del is derived to describe analytically the behavior of angles for a period
ic oscillator under weak periodic and quasiperiodic forcing. An explicit co
nnection is demonstrated between the return angle and the phase of the: ext
ernal periodic forcing. The technique is tested on simulated nonstationary
data and applied to human heart rate variability data.