| dc.description.abstract | In the conventional attention blink paradigm, researchers have been focusing on the deteriorated performance of the second target (T2) caused by the preceded identification of the first target (T1). Many behavioral and neuroimaging studies have revealed that the main cause of the attentional blink phenomenon is a post-perceptual deficit, namely, working memory updating limitation. Recently, Potter and colleagues (2002; 2005) have demonstrated that T1 can be outperformed by T2 when they are presented temporally close to each other (i.e.: the SOA between T1 and T2 is within 53 ms). Their findings have challenged a prevailed assumption in the AB literature, that is, the T1 always dominates in the access to the central processing system. Irrespective of the earlier entry of T1 to the system at short lags, subjects’ performance is better on T2. Potter et al (2002) has proposed that T1 and T2 intensely compete for attentional resource during the early stage of processing and T2 could win over the competition if it is presented right after the T1 without a distracter between them. Davenport and Potter (2005) have indicated that this competition is reliant on the relationship of T1 and T2. They used semantic related words as stimuli and showed that only T2 can prime T1 but not the reverse. Their demonstration successfully illustrates some features of competition. The competition two-stage model open up the possibility that the competition between T1 and T2 may be the main course of the AB and the lag1 sparing effect. Nonetheless, the semantic stimuli can only result in an all-or-none behavioral index. By using a more quantitatively measures, namely, the numerical distance effect, I attempt to clarify the characteristics of the competition in the attentional blink. With the manipulation of the number distance between T1 and T2, I aim to investigate whether the competition between T1 and T2 is a gradual competing process and the AB will be modulated by the number distance.
In the behavioral experiment, I use a rapid serial visual presentation (RSVP) task to probe the fate of the number distance effect at lag 1 and in the attentional blink period. In the RSVP task, subjects were required to identify a double-digits number (served as T1) then compared whether the second double-digits number (served as T2) was larger or smaller than T1. The numerical distance between T1 and T2 were systematically manipulated in the task. The behavioral data indicated that the performance of T2 surpassed over that of T1 at lag1 and the reverse pattern of the result was observed in the longer lags (i.e. AB effect). In the longer lags condition, subjects’ performance was deteriorated with the decrement of the numerical distance. This pattern of results is consistent with the predication that the competition between T1 and T2 is a gradual effect and the numerical distance between T1 and T2 can modulate AB. However, in the lag 1 condition, the numerical distance effect was also found but the pattern of it is reverse from the one observed in the condition of the longer lags condition. Apart from the distance effect, the error pattern indicated that subjects were inclined to mistaken T2 as T1 when they made errors. These intrusion errors also intensify with the decreased distance between T1 and T2. It seems that the numerical information is processed in the lag 1 condition.
In the ERPs experiment, I investigated the relationship between AB and P300 component (e.g.: Kranczioch et al., 2003) and probed whether the number distance effect can be reflected by ERPs waveforms (e.g.: Dehaene, 1996) The pattern of the result is consistent with the previous findings showing that the amplitude of P300 of correct T2 is significantly higher than incorrect T2 during the attentional blink period. Furthermore, the results also demonstrate that when T2 was correctly reported, the numerical distance information can also be reflected on ERPs component N1~ P2p during the attentional blink period. | en_US |