Abstract
The ability to identify a second target (T2) embedded in a rapid serial visual presentation (RSVP) is impaired when T2 is presented at shorter lags following the first target (T1) than at longer lags – an attentional blink (AB). We investigated whether the responses of spatially selective neural populations are modulated by attending to T1, and whether this modulation relates to the AB. We used an inverted encoding model to estimate location-selective neural population responses from scalp recorded EEG in the alpha-band as a measure of selective spatial attention. Participants monitored RSVP streams of alphanumeric stimuli presented in the periphery. The angular rotation of the RSVP locations encircling central fixation varied from trial-to-trial (0°:60°:360°). T1 was a color singleton arrow. Participants either attended both T1 and T2 (n = 15), or ignored T1 and attended only T2 (n = 7). T2 ('X' or 'K') only ever appeared in the center RSVP stream at either lag 3 or 9. An AB was observed when T1 was attended (p < .001), but not when T1 was ignored (p = .899). When T1 was attended we observed an increase in the slope and power of the location-selective neural responses relative to the pre-T1 period (% change) – i.e., a "boost" in selective spatial attention from ~100 to 440 ms following T1. Critically, while we observed this "boost" on trials where T1 and T2 were correctly identified, on trials where T1 was correct but T2 incorrect there was a decrease in power. This effect was modulated by lag (T2 performance x lag, p < .05). Our results indicate that attending to T1 modulates both the slope and power of location-selective neural population responses in the alpha-band. Furthermore, this post-T1 modulation of power changes with T2 performance accuracy as a function of T1-T2 lag – i.e., an AB.
Meeting abstract presented at VSS 2017