Measurement of ERP latency differences: A comparison of single‐participant and jackknife‐based scoring methods

Abstract

We used computer simulations to evaluate different procedures for measuring changes in the onset latency of a representative range of event-related components (the auditory and visual N1, P3, N2pc, and the frequency-related P3 difference wave). These procedures included several techniques to determine onset latencies combined with approaches using both single-participant average waveforms and jackknife-subsample average waveforms. In general, the jackknife-based approach combined with the relative criterion technique or combined with the fractional area technique (J.C. Hansen & S.A. Hillyard, 1980; S.J. Luck, 2005) provided the most accurate method and the greatest statistical power, with no inflation of Type I error rate.Descriptors: ERP latency, Jackknife, N1, P3, N2pc, Frequency-related P3Research using event-related potentials (ERPs) often focuses on differences in amplitudes as well as differences in latencies of ERP components. However, for both measurements there are several scoring methods, and quite often researchers have to decide rather arbitrarily which method might be most appropriate. In this article we compare several methods for determining ERP latency differences. To evaluate the methods we ran computer simulations based on data of five ERP components: the visual N1, the auditory N1, the P3 (hereafter used to mean the P3b), the N2pc, and the frequency-related P3 component (infrequent minus frequent difference wave).These five components were chosen because they are very different and representative of a broad range of components, as is apparent from the following brief review of these components. The N1 components (visual, auditory) are relatively early components, strongly influenced by physical properties of the stimulus. They are characterized by a clear onset and a sharply increasing amplitude, but clearly reflect different underlying neural generators. In contrast, the P3 is a late component that is relatively insensitive to the physical properties of the stimulus (with the exception of tone intensity; see Covington & Polich, 1996), but is influenced by probabilities, expectations, and resource allocation (see Johnson, 1986). Quite often the P3 does not show a clear onset, and its peak latency is difficult to determine because the component has a wide temporal extension without a sharp peak. The N2pc and frequency-related P3, for their part, are measured from difference waves that are obtained by subtracting ERP waveforms computed at different electrode sites or at same electrode sites but in different experimental conditions. The N2pc is an index of the allocation of visualspatial attention and is isolated by subtracting the ERP at posterior electrode sites ipsilateral to an attended item from the ERP at the corresponding contralateral electrode site, whereas the frequency-related P3 is isolated by subtracting the ERP elicited by frequent targets from the ERP elicited by infrequent targets.For all components, researchers are often interested in estimating the latency differences a...

Lateralized electrical brain activity reveals covert attention allocation during speaking

“…Because the N2pc was expected to show both a shift of attention to Object 1 and a shift to Object 2, the jackknifed fractional peak latency was measured for the most negative peak and the most positive peak in the epoch. The onset of the negativity was defined as the last sample before the negative peak that was equal to or less negative than 50% of the peak amplitude (a setting known to yield high power; Kiesel et al, 2008). The onset of the positivity was defined as the last sample before the positive peak that was equal to or less positive than 50% of the peak amplitude.…”

Sound offset-related brain potentials show retained sensory processing, but increased cognitive control activity in older adults

“…ANOVA of the offset-related P1 amplitudes (measured in 20 ms windows centered at the peak amplitudes at Cz in the group-average short-minus-long difference waveforms, Table 2 The Group × Condition (150, 300, or 450 ms vs. 750 ms) ANOVAs of the jackknife peak latencies (measured at Cz in the short-minus-long difference waveforms; Figure 6) as described by Kiesel et al (2008) showed no significant effects for P1 or P3b. For N1/N2, however, a significant Group main effect was present: F(1,39) = 18.814, η G 2 = 0.990, p < 0.001, indicating longer peak latency in the older adult group.…”

“…Using the jackknife procedure described by Kiesel et al (2008), the offset-related P1 (measured at T7), N1/N2 (measured at Cz), and P3b…”

“…Using the jackknife procedure described by Kiesel et al (2008), the offset-related P1 (measured at T7), N1/N2 (measured at Cz), and P3b (measured at Pz) latencies were submitted to mixed Group × Partition (faster-, or slower-than-median) × Condition (150, 300, or 450 ms vs. 750 ms)…”

“…To assess whether between-group differences were present in the latencies of the ERPs observable in the difference waveforms, their peak latencies measured at Cz were analyzed by applying the jackknife procedure as described by Kiesel, et al, (2008) To assess whether the latency of the offset-related waveforms varied with reaction time, which may indicate that reaction-time sensitive components, like the N2 or P3b contribute to the waveforms, short tones followed by correct responses were sorted according to reaction time, and ERP epochs corresponding to short tones with faster-than-, and slowerthan-median reaction times were averaged separately (see Horváth, 2016). Using the jackknife procedure described by Kiesel et al (2008), the offset-related P1 (measured at T7), N1/N2 (measured at Cz), and P3b (measured at Pz) latencies were submitted to mixed Group × Partition (faster-, or slower-than-median) × Condition (150, 300, or 450 ms vs. 750 ms)…”