Parameter estimation of the Linear Phase Correction model by hierarchical linear models

The control of human motor timing is captured by models that make assumptions about the underlying information processing mechanisms. A paradigm for its inquiry is the Sensorimotor Synchronization task, in which an individual is required to synchronize the movements of an effector, like the finger, with repetitive appearing onsets of an external event. The Linear Phase Correction model is a cognitive model that captures the asynchrony dynamics between the finger taps and the event onsets. However, when the synchronization periods are short and/or when there is variability between multiple sequences, the existing parameter estimation methods are biased. Therefore, this work is an approach of unbiased parameter estimation of the LPC model. Based on simulated data, we, first, present a method that integrates multiple sequences within a single model and estimates the model parameters of short sequences with a clear reduction of bias. Second, by relating random effects to the asynchronies sharing the same sequence, we show that parameters can also be retrieved robustly when there is between-sequence variability of their expected values. Since such variability is common in experimental and natural settings, we herewith propose a method that increases the applicability of the LPC model. This method can fit data from short and varied sequences, which may reduce parameter biases due, for example, to fatigue or attentional variation. This allows experimental control that previous methods are unable to provide.