1.5 The Aims of the Present Study
Most studies looking at single-trial data analyzed one or two
components, typically the P300 and N200. By looking at multiple
components, including baseline, researchers can determine if the
variability is random or systematic (Gaspar, Rousselet, & Pernet,
2011). If within-subject variability is systematic, it would be helpful
to find a reliable method to analyze single-trial data that preserves
individual differences and can demonstrate systematic changes during
task performance within and across individuals. Therefore, the purpose
of this study was to determine the feasibility and validity of measuring
multiple ERP components (e.g., the N1, P2, N2, and P3 peaks) using a
simple computational strategy to obtain ST measurements. The feasibility
of these ST measures was considered by computing the SNR for the
traditional averaged ERP, the STP approach, and the Single Trial Average
Time Window (STW) approach. Reliability was determined through the
comparison of split-half and test-retest reliability measures of
amplitude and latency computed for single trials to those computed for
measures of amplitude for the STW approach and amplitude and latencies
obtained from the traditional averaged ERP. We expect that both the SNR
and correlations for reliability measures will demonstrate similarities
in the different ST approaches. Validity was considered by comparing the
standard measurement error (SME) of the STP and STW approaches. The
validity of the ST measures was further determined by computing how well
an individual’s mean ST amplitude, ST amplitude variability, ST latency
variability, and mean ST noise predict the individual’s peak amplitude
measures of the averaged ERP. Given issues of SNR, we expect measures of
reliability and validity to improve across the phases of cognitive
processing, reflected by successive ERP components. Furthermore, we used
curve fitting to look for the presence of systematic changes across
trials. If the process of identifying ERP components within ST data can
be shown to be reliable and valid, the methodology can then be used to
demonstrate that systematic changes do occur within a paradigm from one
trial to the next.