Cognitive Behavioural Therapy

Phonological Similarity

and the

Attentional Blink

Name: Aleena Bablanian

Student ID: 40731316

Abstract

Previous research has shown that during Rapid Serial Visual Presentation (RSVP) tasks where participants are required to identify two targets (T1 and T2) within a series of distractors, presented at fast rates, a marked deficit exists in the detection of the second target when the first target has been correctly identified, particularly if the targets appear within a Stimulus Onset Asynchrony (SOA) between 200 and 500ms. This is known as the Attentional Blink deficit. Raymond, Shapiro & Arnell’s (1992) gating model and Chun and Potter’s (1995) Two-stage model are two of the many accounts that attempt to explicate this phenomenon. The following experiment investigated the existence of the AB as a function of lag type and examined whether phonological similarity of distractors interfered with dual target report accuracy. It was hypothesized that the AB deficit would be greatest at lags 2 and 3, gradually improving at subsequent lags, with evidence of lag one sparing and that conditional report of T2 would be poorer in the presence of phonologically similar distractors. Significant results were produced and both hypotheses were supported. Implications of these results are discussed in terms of temporal attention limitations and how this may effect us in real-life situations.

On a daily basis, humans are faced with the challenge of distributing their attention simultaneously to various sources. Attention is a complex mechanism of the brain and in instances that require visual search and detection of objects in visual space, for instance in driving situations and airport baggage security checks, competing sources of information can cause a limit to the neural processing of this highly functional system (Marois, Chun & Gore, 2000; Wright, 1998; Shapiro, Arnell & Raymond, 1997) Braun, Koch & Davis (2001) highlight spatial limits in attention; whether we can concurrently attend to more than one visual event in space, whilst Shapiro (2001) draws attention to temporal limits; how quickly we can attend to many events presented in succession. This is the basis of the following study.

Rapid Serial Visual Presentation (RSVP) methodologies have allowed researchers (Coltheart & Langdon, 1998; Chun & Potter, 1995) to investigate the temporal limits on attention in studies which require participants to identify two targets (T1 and T2) within a series of distractors, presented at a rate of about 10 per second. Previous research (Olson, Chun & Anderson, 2001; Chun & Potter, 2001; Shapiro et al., 1997) has shown that there is a marked deficit in the detection of the second target when the first target has been correctly identified, particularly if the targets appear within a Stimulus Onset Asynchrony (SOA) range; time between onset of T1 and onset of T2, between 200 and 500 ms (Chun & Potter, 1995; Raymond, Shapiro & Arnell, 1992; Broadbent & Broadbent, 1987). This is known as the Attentional Blink (AB) deficit (Raymond et al., 1992).

Over the years, various accounts for this phenomenon have originated. Raymond et al., (1992) proposed the gating model which suggests that the second target suffers attentional suppression, as the incomplete processing of T1 occupies attentional resources required to process T2 and so it often it is missed. Their findings were consistent with the most distinctive result of AB experiments which follows that the percentage of correct T2 responses (given correct report of T1) drops rapidly in the first half second after T1 presentation and subsequently improves- Figure 1 (Shapiro, Arnell & Raymond, 1997; Chun & Potter, 1995).

However, there appears to be an exception, in that if T1 and T2 appear adjacent to one another, performance is not considerably affected. Hommel and Akyurek (2005) identify this as Lag 1 Sparing. Previous research (Raymond et al., 1992; Chun & Potter, 2005, Olson et al., 2001) suggests that this is due to the parallel processing of T1 and the item immediately following it. Therefore, if the adjacent item is T2, then they are processed together and so T2 is not suppressed. However, if this item is a distractor, it interferes with the processing of T1 and expends attentional resources, resulting in a suppression of T2 and thus an AB.

Figure 1: the attentional blink (Chun & Potter, 1995) as a U-shaped function of lag

Enhancing these findings, there is evidence to suggest that T2, although not processed at a level sufficient to report, can prime a semantically similar subsequent target, suggesting a rather high level of processing at least at some unconscious level (Shapiro et al., 1997; Maki, Frigen & Paulson, 1997). In their second experiment, Chun and Potter (1995) found that when T2 was missed, the probability of reporting T3 was greater and so they proposed that processing of a target at a reported level causes a transitory deficit for a successive target. This finding contradicts the inferences of the gating model of the AB (Raymond et al., 1992) that suggests that unless an item is adjacent to T1 it cannot be processed. However, these studies exhibit that T2 and even the distractors are processed at some semantic level. This proposal is however consistent with Chun & Potter’s (1995) Two- Stage Model for Multiple Target Detection. The first stage of this model initiates target detection at an unreported level, whilst the second stage is capacity-limited and involves target identification and consolidation into memory at a level sufficient to report. It is at this second stage where the AB of the second target occurs, as the demands of T1 prohibit consolidation of T2 into a more durable store; namely short term or working memory and thus remains processed at a level inadequate to report.

Various researchers have been able to disclose those factors that contribute to the size of an AB. The most prominent of these include lag size between the targets (Hommel & Akrurek, 2005; Shapiro et al., 1997; Chun & Potter, 1995), target-distractor visual similarity (Maki, Bussard, Lopez & Digby, 2003; Chun & Potter, 1995) and degree of difficulty of T1, for instance word length, which can increase processing time and attentional capacity requirements (Olson et al., 2001; Coltheart & Langdon, 1998). The purpose of this study is to extend on these findings