Automatic change detection: Mismatch negativity and the now-classic Rensink, O’Reagan, and Clark (1997) stimuli

Abstract

Change blindness experiments had demonstrated that detection of significant changes in natural images is extremely difficult when brief blank fields are placed between alternating displays of an original and a modified scene. On the other hand, research on the visual mismatch negativity (vMMN) component of the event-related potentials (ERPs) identified sensitivity to events (deviants) different from the regularity of stimulus sequences (standards), even if the deviant and standard events are non-attended. The present study sought to investigate the apparent controversy between the experience under the change blindness paradigm and the ERP results. To this end, the stimulus of Rensink, O’Reagen, and Clark (1997) was adapted to a passive oddball ERP paradigm to investigate the underlying processing differences between the standard (original) and deviant (altered) stimuli measured in 22 subjects. Posterior negativity within the 280–330 ms latency range emerged as the difference between ERPs elicited by standard and deviant stimuli, identified as visual mismatch negativity (vMMN). These results raise the possibility that change blindness is not based on the lack of detailed visual representations or the deficiency of comparing two representations. However, effective discrimination of the two scene versions requires considerable frequency differences between them.

Attention and schema violations of real world scenes differentially modulate time perception

“… 32 , 40 , 41 ) stimulus, in a stream of identical ones, can affect subjective duration judgments and lead to duration overestimations of the odd stimulation. Given that, to our knowledge, no studies on timing had tested whether more complex stimuli (i.e., natural scenes) could induce similar effects in an oddball setting (see also 42 – 44 for the use of more complex stimuli in variations of the oddball paradigm yet not in terms of time perception), we reasoned that our findings could further expand the literature around the oddball effect, as well as provide evidence on the role of attention on intervals’ duration estimates.…”

Orientation and contrast deviance examined: Contrast effects mimic deviant-related negativity yet neither produce the canonical neural correlate of prediction error

“…If not, it would be unlikely that our foraging ancestor would have been able to detect those changes in visual properties, such as when a predator was approaching. A recent study by File et al [69] found that change-blindness does not imply the absence of visual representations or deficits in distinguishing regular and irregular visual representations [69]. Their findings indicate that even for the most subtle perceptual changes, our brain's encode such disparities.…”

“…The question is, what neural signatures reveal detection? File et al [69] reported posterior negativities �250 ms for undetected changes in scenery stimuli. This does not appear here.…”