Article
A video viewer can count passes correctly and still miss an unexpected event that enters the scene. That result asks a narrow question about attention under one task, not whether sight is generally unreliable. Studies of expectation and perception bring their own tasks and definitions. To compare them carefully, we need to separate noticing, sensing, deciding, remembering, and later reporting.
A striking result with a specific denominator
Simons and Chabris asked observers to count passes in a dynamic video while an unexpected event appeared. Across the study's combined conditions, 46% of 192 observers missed the event. The number describes those participants and that task. It is not a general rate of how often people fail to notice important things, and it should not be repeated as an unqualified “gorilla statistic.”[1]
Inattentional blindness names a failure to report an unexpected stimulus under defined conditions. The finding demonstrates that attention can limit what people notice in a task; it does not show that the stimulus was unreal or that perception is wholly an illusion. A study's method, instructions, and measure matter to any conclusion drawn from its result.[1][4]
Expectation is studied in defined tasks
Expectation research asks how prior probabilities or cues shape decisions about sensory evidence. Summerfield and de Lange review neural and computational accounts in perceptual decision-making paradigms. Their work concerns specified tasks and response choices, not a rule that belief determines whatever a person sees. A measured influence needs its own conditions, effect, and limits.[2]
A participant's response can reflect several stages: sensory processing, allocation of attention, a decision criterion, memory, and the words used to report an experience. If an expectation changes an answer, researchers have to ask where in that sequence the change occurred. The same verbal report does not by itself identify a single mechanism.[2][3]
The meaning of top-down effects is debated
The phrase “top-down effect” can cover claims that cognition alters perception, but scholars disagree about what some experiments establish. Firestone and Scholl argue that apparent effects may sometimes be explained by attention, judgment, memory, or response rather than a change in perception itself. Their critique is part of an active methodological debate, not proof that expectations never affect perceptual processing.[3]
That debate makes careful language useful. A result can show that expectations influence performance in a particular experiment while leaving open whether the sensory experience itself changed. Different experimental designs can test different links in the chain. The conclusion should remain no broader than the task and measurement support.[2][3][4]
A real world, filtered attention
Attention helps select information for action and report, but it does not create the external world. Perception is shaped by sensory input and by processes researchers still work to distinguish. Showing that noticing is selective does not make everything seen false, just as a confident report does not prove that every detail was attended to.[1][2][3]
When an expectation changes what someone reports in one task, what additional evidence would help you decide whether attention, perception, memory, or judgment changed?