I have no main concerns regarding this contribution. The information provided is adequate, the arguments are valid. The following suggestions may improve readability:
The definition of the Attentional Blink at the top is a little dense for a broad readership. What about: “… is the phenomenon that the second of two targets cannot be detected or identified when it appears close in time to the first.”?
Part of a sentence is missing in the “What is not an AB” section. The sentence is in the HTML code, it just does not show up in my browser… (IE6).
The section “A few AB C’s” suggests that a T1 mask is typically required for an AB to occur. Nieuwenstein, Potter, and Theeuwes (in press; JEP:HPP) recently found an AB without a T1 mask.
Section “Not to be confused with the AB”: One can add that task switching costs and spatial switching costs are also different from the AB. The AB occurs even when T1 and T2 involve the same task, and are presented in the same location (e.g. Chun & Potter, 1995). Some have added a task switch between T1 and T2, which may add to the T2 deficit. By the way, this entire section might be merged with “What is not an attentional blink”.
Section “AB approaches”: The part on the neuropsychological approach may also mention studies showing that parietal and frontal lesion hemineglect patients suffer from a prolonged AB (Husain et al., 1997).
Section “AB theory” I’m not sure that Sergent et al (2005) see the AB as a failure to retrieve information from a short-term memory store. Their work reads more as an argument for a central bottleneck, but this is open to discussion I guess.
We have addressed virtually all of the points raised by Chris Olivers with the following exceptions:
We do not think it is necessary to add another reference to the point that the AB can occur without masking as there is already one report cited to this effect. There were many reports that could have been cited for each point but we had to draw the line somewhere in the interest of space limitations.
The request to insert something about 'task switching' and 'spatial switching' has been done in the section "What is not an AB" as the section where this edit was suggested to go ("Not to be confused with the AB") is reserved in our minds for 'paradigms' that can be confused with the AB.
The part of the sentence missing in the "What is not an AB" section is not really missing; the link to "Edit" the figure is obscuring a bit of the text. We are searching for a way to remedy this.
We thank Chris Olivers for his excellent suggestions.
This is a fairly comprehensive and current review of the AB and its related effects in the world of visual attention. It reads well. It is also pleasantly short, which in this case (given the really vast literature that now exists on the AB) bestows on it a huge plus. The few categories that are used in the article are really the most important ones! I will recommend the piece to all my students.
What follows are a few recommendations for clarification and some idiosyncratic (on my part) suggestions for focusing some of the theoretical debates that are ongoing about this incredibly generative effect (“generative” in the sense of sparking new research).
In the section “What is not an attentional blink?” The sentence: “No AB occurred in this condition suggesting attention to T1 and not perceptual masking…” I recommend using the term “visual masking” as opposed to perceptual masking, and linking the term to Bruno Breitmeyer’s Scholarpedia article on visual masking. Perceptual masking is not well defined in this context, whereas visual masking is, as the authors point out when they say “…thus equating the two conditions perceptually…” Again, they should use the term “visual” here rather than perceptual, as perception is determined by much more than the nature of the visual input, which is controlled in this comparison. Clearly, perception is not controlled, as evidenced by the very existence of the AB itself.
In the section “A few AB C’s” The sentence: “Finally, the issue of T1 difficulty and its effect on the AB has been heavily investigated with some investigators finding a relationship between T1 difficulty and the magnitude of the AB (e.g., Seiffert & Di Lollo, 1997; Ward Duncan, & Shapiro, 1997) and others not (McLaughlin, Shore, & Klein, 2001)….” I recommend referring here to two other papers of Visser in 2007 (than the one that is referenced): 2007: Visser Troy A W T1 difficulty and the attentional blink: expectancy versus backward masking. Quarterly journal of experimental psychology, 60(7):936-51. 2007: Visser Troy A W; Ohan Jeneva L Data-limited manipulations of T1 difficulty modulate the attentional blink. Canadian journal of experimental psychology, 61(2):102-8. In my mind these papers are the most thorough investigation so far of the relations between T1 difficulty and the AB. They pinpoint processing time, not difficulty per se, as the critical ingredient. They deserve mention, even if someday they will be proven wrong or incomplete, because they focus on what is an important issue in behavioral cognitive psychology more generally: the distinction between processing delays vs processing effort.
In the sentence: “The improvement of the Raymond et al. study over that by Broadbent and Broadbent was the finding of a similar outcome but using stimuli, i.e., letters, that did not require access to ‘meaning’,…” please clarify for the general reader in what sense “meaning” in stimuli reduces the validity of an experimental effect.
The section “When does an AB not occur?” could benefit from a somewhat more comprehensive review of the fate of “blinked” second targets. I’m thinking of work by Giesbrecht, B., Sy, J. L., & Lewis, M. K. (2008). Personal names do not always survive the attentional blink: Behavioral evidence for a flexible locus of selection. Vision Research. Giesbrecht, B., Sy, J. L., & Elliott, J. C. (2007). Electrophysiological evidence for both perceptual and post-perceptual selection during the attentional blink. Journal of Cognitive Neuroscience, 19, 2005-2018. Giesbrecht, B., Sy, J. L. (2008) Target–target similarity and the attentional blink: Task-relevance matters! Visual Cognition 17(3), 307-317.
Overall, this is an informative review that should be very useful to readers who want a quick survey on the attentional blink.
The AB is defined as having occurred when T1 is reported correctly but report of T2 is inaccurate at short T1/T2 intervals, typically between ~100 to 500 ms, but recovers to the baseline level of accuracy at longer intervals (see Figure 3) <-- change 'but recovers' to 'recovering'?
A few AB C’s (I think this is an important section and could read more smoothly, benefiting from this scholarpedia format. This could run as a concise list of points. It is subjective how to do so, but I were editing this, I would recommend something like the following (I've added some citations as well; although one has to draw limits, these are amongst the well-cited attentional blink papers in the literature):
The question is often asked, “What are the requirements to produce an AB?”
The full RSVP stream is not required. Ward, Duncan, and Shapiro (1997) presented masked targets, but without any other stimuli, in the centre of a visual display and obtained an AB outcome, though not revealing the typical AB function (see lag-1 sparing below).
T1 and T2 masks are important but in different ways (Seiffert & Di Lollo, 1997). A seminal finding for understanding the attentional blink is that removing the T1 mask completely attenuated the AB (Raymond et al., 1992). However, Visser (2007) discovered an AB can be obtained without the use of a T1 mask, though the T1 task was novel and thus likely more difficult. Moreover, Giesbrecht and Di Lollo (1998) showed that T2 must be appropriately masked; simply degrading T2 does not produce an AB.
Related to the masking issue, T1 should be attentionally demanding to process. The issue of T1 difficulty and its effect on the AB has been heavily investigated with some investigators finding a relationship between T1 difficulty and the magnitude of the AB (e.g., Seiffert & Di Lollo, 1997; Ward Duncan, & Shapiro, 1997) and others not (McLaughlin, Shore, & Klein, 2001).
Is T1-sparing a defining feature of the attentional blink? Although it might be expected that the AB should be maximal immediately following T1, a phenomenon referred to as lag-1 sparing often arises, where T2 accuracy is preserved when T2 occurs immediately after T1. Lag-1 sparing is most likely to occur when the T1 and T2 are selected from the same category and spatial location (Chun & Potter, 1995; Visser et al., 1999).
Is the AB strictly visual? AB exists within other modalities (audition, touch) and between modalities (e.g., visual T1, auditory T2). Evidence suggests that both the above experimental situations can lead to an AB, although whether or not an AB occurs depends to a considerable extent on the method employed (Potter et al., 1998; Arnell & Jolicoeur, 1999).
Because conciseness is key here, I felt that the following sections did not add much and could be deleted if the authors are interested in tightening this further.
What came before -Broadbent and Broadbent was already mentioned earlier. -Weichselgartner and Sperling is worth mentioning, but this could be done in one line earlier as well. I just felt too much space was devoted to this issue that is not on people's minds.
AB parametrics -this section was not very informative and given the lack of citations, I just felt it could be dropped.