Postersession 2
Poster #: 59
Topic: Error signals
Thursday, Sep 10, 2015
14:30-16:00
1st floor

Pre-stimulus ERP-correlates of predictive auditory processing

Martin Reiche1, Andreas Widmann2, & Alexandra Bendixen1

1Department of Physics, Chemnitz University of Technology, Chemnitz, Germany
2Institute of Psychology, University of Leipzig, Leipzig, Germany
martin.reiche@physik.tu-chemnitz.de

The Mismatch Negativity (MMN) has been proposed to be associated with a predictive mechanism forming a model of the environment and testing incoming sounds for their compatibility with this model. The MMN is thought to reflect the outcome of that comparison or a model update. Most studies investigate the underlying processes in situations with almost perfect predictability. Yet predictability in natural signals is often compromised by uncertainty; hence the proposed predictive mechanism is only plausible if it can be shown to dynamically adapt to the certainty provided by the sensory context. We investigated this issue by experimental manipulations of certainty in an event-related potential (ERP) paradigm in which predictability was implemented via frequency repetition during passive listening. In two independent experiments, we systematically manipulated the certainty of predictive relations between tones across conditions using different parameters of (un)certainty, such as the overall proportion of valid predictions or the accuracy with which predictions can be made. Results show an ERP correlate of predictive processing size-dependent on predictive certainty in the sequence. In contrast to the MMN, which occurs after the onset of an event, we find that prediction-related ERP-effects can be observed immediately before the onset of a sensory event. This suggests that the proposed predictive mechanism rapidly and flexibly adapts to the degree of certainty in the auditory sensory context. To exclude alternative explanations, we are currently assessing the replicability of this effect with higher and lower tone presentation rates.