Repetition blindness has a perceptual locus: Evidence from online processing of targets in RSVP streams.

Abstract

Four experiments tested whether repetition blindness (RB; reduced accuracy reporting repetitions of briefly displayed items) is a perceptual or a memory-recall phenomenon. RB was measured in rapid serial visual presentation (RSVP) streams, with the task altered to reduce memory demands. In Experiment 1 only the number of targets (1 vs. 2) was reported, eliminating the need to remember target identities. Experiment 2 segregated repeated and nonrepeated targets into separate blocks to reduce bias against repeated targets. Experiments 3 and 4 required immediate "online" buttonpress responses to targets as they occurred. All 4 experiments showed very strong RB. Furthermore, the online response data showed clearly that the 2nd of the repeated targets is the one missed. The present results show that in the RSVP paradigm, RB occurs online during initial stimulus encoding and decision making. The authors argue that RB is indeed a perceptual phenomenon.

Repetition blindness and repetition priming: Effects of featural differences between targets and distractors on RSVP dual-target search

“…In addition, it has been demonstrated that repetition blindness occurs when memory demands are reduced (Chun, 1997;Dux & Marois, 2007;Luo & Caramazza, 1995 and even when no recall is required (Hochhaus & Johnston, 1996;Johnston et al, 2002). Thus, although retrieval failure contributes to some effects of repetition seen in RSVP streams, it is not responsible for all of them.…”

“…Research in which this paradigm has been used has identified two fundamental limitations of conscious visual perception: the attentional blink (Broadbent & Broadbent, 1987; Ray k -mond, Shapiro, & Arnell, 1992;Weichselgartner & Sperling, 1987) and repetition blindness (Chun, 1997;Johnston, Hochhaus, & Ruthruff, 2002;Kanwisher, 1987). The attentional blink refers to subjects' impaired ability to report the second of two different targets in an RSVP stream if it appears within 200-600 msec of Target 1 (T1).…”

When Words Collide: Facilitation and Interference in the Report of Repeated Words From Rapidly Presented Lists.

“…In fact, as the final draft of this article was being prepared, a published report appeared verifying the prediction that repetition blindness can be obtained with nonwords (Harris & Morris, 2004). Johnston et al (2002) carefully considered the question of whether repetition blindness is more appropriately characterized as a perceptual or as a memory phenomenon. Their evidence for online influences of repetition in a target detection task supported a strong argument against the idea that repetition blindness can be attributed entirely to postlist memory recall processes.…”

“…Thus, tokenization of the second presentation of a repeated word fails, and it cannot be reported. In the type-token account of repetition blindness, then, the impaired report of a repeated item is attributed to a failure of encoding or perception that occurs at the moment a repeated word is presented in an RSVP list (see also Hochhaus & Johnston, 1996;Johnston, Hochhaus, & Ruthruff, 2002). …”

“…Search tasks involving designation of a target set (e.g., vowels or two specific letters) in advance of an RSVP list led to a form of repetition blindness in which detection of two occurrences of the same target within a list was less likely than detection of two different targets within a list (Johnston et al, 2002;Kanwisher, Kim, & Wickens, 1996;Park & Kanwisher, 1994). This task has no postlist recall demands and provides no opportunity for reporting biases such as a bias against reporting an identified item more than once.…”