ABSTRACT
Time-series analysis methods are applied to data obtained from psychological experiments. Perhaps an analysis of response time (RT) variability from one trial to the next can uncover the hidden dynamic psychological processes underlying such a simple paradigm. A therapist notices how a previously anxious client learns to become more confident and productive over a course of behavioral therapy. A large statistical literature has emerged to analyze temporal changes in data that occur over repeated observations. Data pooling can even generate an average curve that does not represent any individual participant's performance. In particular, techniques are required that can detect changes in the autoregression parameters over time, as was apparent in the nonstationary typing data. The power spectrum is used in many physiological and psychological applications and plays an important role in the diagnosis of nonlinear psychological systems.
