Attentional blink (AB) is a phenomenon that reflects the temporal costs in the allocating selective attention. The AB is typically measured by using rapid serial visual presentation (RSVP) tasks, where participants often fail to detect a second salient target occurring in succession if it is presented between 180-450 ms after the first one. Also, the AB has been observed using two backward-masked targets and auditory stimuli. The term attentional blink was first used in 1992, although the phenomenon was probably known before.
Contents
Research
The precise adaptive significance behind the attentional blink is unknown, but it is thought to be a product of a two-stage visual processing system attempting to allocate episodic context to targets. In this two-stage system, all stimuli are processed to some extent by an initial parallel stage, and only salient ones are selected for in-depth processing, in order to make optimum use of limited resources at a late serial stage.
One curious aspect of attentional blink is that it usually includes "lag 1 sparing", meaning that targets presented very close together in time (at "lag 1" or consecutively in the RSVP stream) are not affected by the attentional blink, even though items presented at slightly greater lags are significantly impaired. In attentional blink, participants in experiments have trouble reporting multiple targets that are in succession to one another, and will only report one accurately when these targets are presented to them 200ms to 500ms apart according to a study by Visser et. al (2015) . These targets are denoted as T1, T2, etc. The phenomena lag-1 sparing refers to the performance of T2 as opposed to what precedes it, T1. T2 performance was originally hypothesized to be reported far less often and less accurately by participants than T1, because the attention of the participant would still be on T1 while T2 was presented immediately afterwards. Visser also thought that the participants would be too focussed on finding the first target that they would miss the second target completely. However, participants actually did better identifying T2 than they were at identifying T1 when the targets were separated by one or two distractors, denoted also as lags.
A possible explanation for lag-1 sparing is that this phenomena is heavily interconnected with attentional blink, but does not operate on the same cognitive mechanisms and requires different stimuli to occur. Specifically, for lag-1 sparing to occur, it needs visual input as practice targets. These targets can be numbers or letters presented in rapid succession. When the first target, T1, is presented, it creates an attentional window because of its novelty, meaning that it attracts and holds more attention by the participant. The novelty that wears off between T1 and T2 creates a “boost” in attention and opens a metaphorical window for faster cognition. Participants now know what and how to look for targets, so they find targets more quickly. This attentional widow remains open long enough for T2 to be presented and processed at a much higher rate because of shared characteristics to T1. Targets are normally presented in less than .5 of a second from each other. Lag-1 sparing also occurred regardless of how information was visually presented. Of two RSVP streams— where T1 location was known in the first stream and unknown in the second stream, lag-1 sparing occurred whether T2 was in the same stream as T1, or in a different stream than T1.
There is as yet no conclusive explanation for the phenomenon of lag 1 sparing, although it is thought to be related to the first parallel stage of the two-stage system of stimulus selection and processing.
According to the LC-NE hypothesis, when a salient, or meaningful stimulus is presented, neurons in the locus coeruleus release norepinephrine, a neurotransmitter that benefits the detection of the stimulus. The effect of this release lasts for 100 ms after the salient stimulus is presented and benefits the second target when presented immediately after the first one, accounting for lag 1 sparing. Eventually the neurons in the locus coeruleus enter a refractory period, due to the auto-inhibitory effect of norepinephrine. According to the hypothesis, targets presented during this refractory period cannot trigger a release of norepinephrine, resulting in the attentional blink. The episodic distinctiveness hypothesis of the ST2 model suggests that the attentional blink reflects a limitation of the visual system attempting to allocate unique episodic contexts to the ephemeral target stimuli presented in RSVP.
The attentional blink can be moderated by changes in visual similarity between targets and distractor stimuli, but it can also be affected by conceptual similarities, suggesting that stimuli are processed to quite a deep level preconsciously, with much of the resulting information discarded before it reaches consciousness.
The attentional blink is related to, but distinct from the phenomenon of repetition blindness.
Attentional Blink can be used in other studies and experiments as a way to measure attention. A study conducted by Morrison et. al 2015 used AB to measure attentional differences in people with mental disorders such as depression and social anxiety disorder. Measuring AB in the subjects was effective in showing that subjects with those disorders have more difficulty recognizing T2 than others. Another study used the magnitude of AB in order to see whether positive moods affected temporal attention.
Emotion
When AB is tested, most often the stimuli presented are inanimate objects. But research shows that not only do inanimate stimuli effect AB but human faces do as well. Despite the hypothesis that configural and featural information are processed by separate channels, thus avoiding AB, human faces in all different configurations, are still processed and affected by AB.
An important factor which influences the AB is the role of emotions. Research has shown that when the second target (T2) in RSVP is an emotionally relevant stimulus it is more likely to be perceived during the attentional blink. The attentional blink is not only modulated by emotional relevance of (T2) but also by the emotional relevance of (T1). In short: when (T1) is emotionally relevant the AB is lengthened, when (T2) is emotionally relevant then the AB is reduced. This research suggests that emotion mediates attention.
The sex of the stimuli presented also influences the AB. In a study, it was shown that emotions that are more often than not prescribed to different sexes, were more quickly identified. Not only that, but when the attentional blink would occur, whether it be at lag 1 or lag 2, and how deep of a blink it was, could be manipulated by the sex of and the emotion expressed by T1.
There have also been studies using images as emotional stimuli. Emotionally negative pictures preceding the target by 2 items were found to induce greater deficits in processing the target stimuli than neutral pictures did. Thus, it seems that emotional information can elicit attentional biases which temporarily prevent awareness of actively sought out stimuli.
Meditation
A study conducted by Heleen Slagter, Richard Davidson and colleagues, suggests that meditation, particularly vipassana, may reduce the duration of attentional blink. In an experiment, 17 people received three months of intensive training in meditation. Those 17 along with 23 meditation novices performed an attention task in which they successively picked out two numbers embedded in a series of letters. The novices exhibited attentional blink and missed the second number. In contrast, all the trained meditators consistently picked out both numbers. This indicated meditation practice can improve focus.
The Inhibition Theory
Raymond et al. (1992) suggested that the attentional blink is produced by perceptual uncertainty amongst the target (T1) and following target (T2). They suggest that this confusion happens at some point in the target identification processes. When confusion is eliminated, attentional blink isn’t observed. The researchers also suggested that one way to eliminate confusion is to have items that cannot be named.
The Interference Theory
Shapiro et al. (1994) suggested that an interference model may better explain the attentional blink effects than the inhibition model. In this model, the attentional blink is thought to take place because of an out of place item which is selected out of the series because of the interference within the items in the series. Shapiro proposes that the amount of interference increases or decreases with the length of the series.
The Delay of Processing Theory
Giesbrecht and Di Lollo (1998) suggests that the attentional blink deficit arises from the wait of target 2 when the person's cognitive processes are busy registering target 1. It is suggested that anything that increases the difficulty of processing the first target will result in a greater attentional blink deficit.
The Attentional Capacity Theory
Duncan et al. (1996) suggests that the initial target takes over parts of our attentional capacity to the loss of sight of a target right after the initial target. This theory suggests that the time for which target 1 continues to occupy attentional capacity is related directly to the difficulty of processing target 2.
The Two-Stage Processing Theory
Chun & Potter (1995) proposed that quickly processing a series of items requires two back to back stages. The first stage is the initial rapid-detection. Here, the possible targets are noticed. The second stage is the capacity-limited in which items are taken in order to report later. Stage 2 occurs after the acknowledgement of targets in Stage 1. Here, stage 2 must finishing processing target 1, until then, target 2 will not be recognized in stage 2. If there is a situation where the second target comes in the first stage, the highway to stage two is delayed. Attentional blinking occurs when the second target is in stage 1 which causes a delay. The attentional blink mirror a restriction in the process of combining information from an unstable representation to a stable representation (Johnson & Proctor, 2004).