What's special about the P300 response to deviance?
The late positive complex (LPC; also referred to as the P300) of the event-related potential (ERP), is elicited by stimuli of any modality, and has been associated with orienting, attention, stimulus evaluation and memory processes. A large number of studies examining the LPC have employed variants of the “oddball” paradigm, inwhich subjects may be asked to attend (e.g., count or button press) or may ignore deviant stimuli randomly presentedwithin a series of homogeneous stimuli. The inverse relationship between LPC amplitude and deviant probability has been examined extensively and is well established. A number of studies have since used principal components analysis (PCA) to elucidate the late ERP components elicited by deviant stimuli under ignore and attend conditions. They have consistently shown that all deviant stimuli (attend-deviants, ignore-deviants and rare novel stimuli) elicit the P3a and P3b components, and that the relative amplitudes of each depend on the relationship between the particular deviant stimuli and task requirements. For the current study,we utilized PCA to examine the late ERP components to both the deviant and the standard stimuli. ERPs were recorded from 48 participants during a short-train auditory oddball paradigm, consisting of 120 auditory tone pips with a 15% deviant probability. Acoustic stimuli consisted of 1000 Hz tones at 50 (Soft) or 80 (Loud) dB SPL intensity; ISI was 1.1–1.3 s. Participants were randomly assigned to alternate oddball conditions: soft standard/loud deviant or loud standard/soft deviant. Half of each group were instructed to button press when they heard a tone of a different intensity, giving four groups in total. For each stimulus type (i.e. deviant and standard) four factorswere extracted, that corresponded to the P3a, the P3b, and two late slow-waves (SWs), recently reported by Barry, MacDonald & Rushby (2010) in a long ISI paradigm(N50 s). The P3a had a central maximum peaking at 300 ms, and was larger to deviants than standards (pb0.001), and overall (i.e. to both standards and deviants) was larger to loud than soft tones (pb0.001). The P3b had a parietal maximumpeaking at 420 ms, andwas larger to deviants than standards (pb0.001), larger to press than no-press (pb0.01), and larger to soft than loud tones (pb0.05). Both SWs showed a negative frontal and a parietal positive topography, with the first (SW1) showing a positive peak at 520 ms, and the second (SW2) at 650 ms. There were no significant differences shown for condition or stimulus type for SW1. SW2 showed a stimulus effect (deviantsNstandards, pb0.01), and was larger for press than no-press (pb0.01). The P300 has been extensively explored in clinical populations to determine its utility as a marker or endophenotype of psychopathology. However, its usage as an endophenotype is problematically heterogeneous, i.e. both amplitude and latency measures of the P300 have been found to be impaired in schizophrenia, affective disorders,AD/HD,Autism, violent offenders, etc. Themajor flawwith the bulk of clinical literature, and indeed the bulk of basic research on the P300 component, is a failure to determine which of, and to what extent, the individual subcomponents of the P300 are contributing to the response being examined. The results in this study show that in addition to examination of the underlying P300 subcomponents to the deviant response, comparison with those found to underlie the standard stimulus will provide further clarification as to ‘when and where’ this component is indexing psychopathology.