School of Psychology, University of Newcastle, Callaghan, New South Wales, Australia.
Priority Research Centre for Translational Neuroscience and Mental Health Research, University of Newcastle, Callaghan, New South Wales, Australia.
Psychophysiology. 2018 Apr;55(4). doi: 10.1111/psyp.13013. Epub 2017 Oct 3.
Internal models of regularities in the world serve to facilitate perception as redundant input can be predicted and neural resources conserved for that which is new or unexpected. In the auditory system, this is reflected in an evoked potential component known as mismatch negativity (MMN). MMN is elicited by the violation of an established regularity to signal the inaccuracy of the current model and direct resources to the unexpected event. Prevailing accounts suggest that MMN amplitude will increase with stability in regularity; however, observations of first-impression bias contradict stability effects. If tones rotate probabilities as a rare deviant (p = .125) and common standard (p = .875), MMN elicited to the initial deviant tone reaches maximal amplitude faster than MMN to the first standard when later encountered as deviant-a differential pattern that persists throughout rotations. Sensory inference is therefore biased by longer-term contextual information beyond local probability statistics. Using the same multicontext sequence structure, we examined whether this bias generalizes to MMN elicited by spatial sound cues using monaural sounds (n = 19, right first deviant and n = 22, left first deviant) and binaural sounds (n = 19, right first deviant). The characteristic differential modulation of MMN to the two tones was observed in two of three groups, providing partial support for the generalization of first-impression bias to spatially deviant sounds. We discuss possible explanations for its absence when the initial deviant was delivered monaurally to the right ear.
世界规律的内部模型有助于感知，因为可以预测冗余输入，并为新的或意外的输入保留神经资源。在听觉系统中，这反映在一种称为失匹配负波（MMN）的诱发电位成分中。MMN 是由违反既定规则引起的，用于指示当前模型的不准确，并将资源引导至意外事件。流行的解释表明，MMN 幅度会随着规则的稳定性增加；然而，第一印象偏差的观察结果与稳定性效应相矛盾。如果音调以罕见的偏差（p = .125）和常见的标准（p = .875）旋转概率，如果最初的偏差音调引起的 MMN 比后来遇到的第一个标准的 MMN 更快达到最大幅度，则会出现差异模式，这种模式会持续整个旋转过程。因此，感觉推断会受到长期的上下文信息的影响，而不仅仅是局部概率统计。使用相同的多语境序列结构，我们研究了这种偏差是否会推广到使用单耳声音（n = 19，右第一个偏差和 n = 22，左第一个偏差）和双耳声音（n = 19，右第一个偏差）诱发的 MMN。在三个组中的两个组中观察到了 MMN 对两个音调的特征性差异调制，为第一印象偏差向空间偏差声音的推广提供了部分支持。我们讨论了当初始偏差仅传递到右耳时其缺失的可能解释。
