Modeling human timing behavior

Abstract

In order to understand human motor timing, individuals are instructed to synchronize their movements with repetitive environmental events. There are cognitive models accounting for the empirical findings obtained in such task. Cognitive models are usually formalized as system of equations that receives variables as input and predict output based on the input, the mathematical expression, and the parameters [1]. Because in experiments, there are always variables that can neither be manipulated nor controlled, i.e., there is noise within and beyond the Central Nervous System (CNS), these model are often defined as parametric family of probability distributions. Schulze and Vorberg (2002) developed such a probabilistic cognitive model, called the Linear Phase Correction model (LPC). Our main goal is to provide a method of parameter estimation of the LPC built on multiple short asynchrony series that can be non-stationary, vary in size, and allow for serially correlated errors.