Individual differences in dual-target RSVP task performance relate to entrainment but not to individual alpha frequency

Abstract

The attentional blink (AB) paradigm is widely used to study visual temporal attention. An important feature of the standard AB paradigm is repetitive visual stimulation, more precisely the rapid serial visual presentation (RSVP) of numerous distracters interspersed with two targets. The RSVP stream is likely to result in entrainment of visual cortex, which has been suggested to negatively affect target identification in the AB paradigm. The present EEG study tested this idea with an inter-individual differences approach. AB task performance and measures of entrainment were derived from 51 participants. Other than predicted, moderate positive correlations were observed for inter-trial coherence and performance, but only for targets not immediately preceded by other targets. A positive correlation with power was evident for targets presented in the critical AB time window. In a second step, it was tested whether the distance between individual alpha frequency and RSVP frequency mediated correlations with inter-trial coherence, as entrainment of the visual cortex through repetitive visual stimulation is particularly effective when the frequency of the stimulation matches the individual alpha frequency. However, no evidence was found supporting such link. While compatible with a number of findings related to the AB and to visual entrainment, the findings of the present study do not provide evidence for the notion that entrainment to the RSVP stream creates a neural environment unfavourable for detecting targets an RSVP stream.

Fig 1. Schematic depiction of the trial layout for the three different trial types.

Targets were digits, distracters were letters. A trial could contain either one or two targets. If it contained only one target the target appeared either in an early position or a late position in the RSVP stream. Trials with a single target in a late position (bottom) were used to extract the ssVEP. For statistical analysis data were extracted from the time range corresponding to the presentation of the 11^th to 15^th distracter in the RSVP stream. The baseline interval was extracted from the pre-RSVP stream fixation period.

Fig 2. IAF histogram and topography.

(A) Histogram of all IAF values for the first resting session. (B) Topographical distribution of resting power at the IAF across the group for the first resting session. Black circles indicate the position of the electrodes selected for statistical analysis.

Fig 3. Power and ITC at 10 Hz as measures of entrainment.

(A) Top: Total power at 10 Hz during the RSVP stream relative to the pre-RSVP baseline. The plot depicts the mean total power from 900 to 1900 ms post RSVP stream onset. Middle: Mean ERP power at 10 Hz from 900 to 1900 ms post RSVP stream onset relative to pre-RSVP baseline. Bottom: ITC at 10 Hz, mean value from 900 to 1900 ms post RSVP stream onset. (B) Scatter plots and regression lines of 10 Hz ITC and 10 Hz total power (top) and 10 Hz ITC and 10 Hz ERP power (bottom).

Fig 4. Significant correlations between 10 Hz ITC and task performance.

(A) T1 identification rate as a function of 10 Hz ITC. The black line corresponds to the linear regression line. (B) Identification rate for T2|T1 lag 2 as a function of 10 Hz ERP power. The black line corresponds to the linear regression line.