Twenty of 24 radiologists failed to report a gorilla that researchers had inserted into a chest CT scan in a 2013 experiment. Twelve of those twenty had looked directly at its location while it was visible. The unusual image was on the screen, but looking toward it did not guarantee that it entered awareness.
Trafton Drew, Melissa L.-H. Võ and Jeremy M. Wolfe reported the finding in The Invisible Gorilla Strikes Again: Sustained Inattentional Blindness in Expert Observers, published in Psychological Science. The participants were carrying out a familiar task: searching lung images for nodules.
The gorilla was outside the search goal
The experiment used a series of CT cases, with the animal inserted into the last one. Its presence was unexpected. The question was whether people with extensive experience examining these images would notice something conspicuous that fell outside the object they were trying to find.
The Association for Psychological Science’s account emphasizes the expertise involved in that search. Radiologists inspect complex medical images for small features. Their experience helps them distinguish relevant patterns that may be difficult for an untrained observer to interpret.
That focus was the setting for the surprise. The task did not ask them to find an animal or inventory every recognizable object on the display. They were looking for a particular kind of medical finding.
The gorilla was visible, but unexpected.
Once a reader knows the headline, the situation changes. Looking at the image specifically to find a gorilla is not a recreation of the experiment. The warning supplies exactly the expectation that the original participants lacked.
Eye position is not a transcript of awareness
The eye-tracking result makes this more than a story about someone never directing their gaze to the right part of an image. Some participants looked at the gorilla’s location without reporting it. That is the distinction between where the eyes point and what the observer consciously notices.
It does not mean an eye tracker can read a person’s thoughts. The study combined gaze records with reports of what participants noticed. Each measure contributes something different: the first concerns visual orientation, while the second concerns the object entering reported awareness.
For interpretation, the denominator matters too. Twelve was the count among the twenty radiologists who missed the gorilla. It was not twelve out of all the people who saw it, or twelve separate failures by the same observer.
The reported numbers describe this experiment. They do not establish a fixed personal property, such as a permanent inability to notice unexpected events, for each participant.
A gorilla miss is not a cancer miss rate
The researchers’ original paper makes clear that the experts were substantially better than untrained participants at finding lung nodules. Expertise was useful for the assigned task even though it did not make the radiologists immune to the unexpected image.
This is why converting the result into a claim that radiologists miss 83 percent of disease would be a fundamental error. The gorilla was an artificial, nonmedical insertion. The measured outcome was whether it was reported during a particular search, not the accuracy of a clinical screening service.
The study also involved a small group in a controlled setting. A clinical workflow can include patient history, prior scans, different reading goals and subsequent review. Those factors need investigation before a laboratory effect can be translated into an estimate of real-world harm.
None of that removes the finding. It identifies what makes the demonstration informative: trained observers, working within their expertise, could still miss an unexpected object.
Unexpected events are difficult to test repeatedly
A later scholarly review, Inattentional Blindness in Medicine, discusses a central methodological problem. Researchers need enough experimental control to know what happened, while also making the task representative of medical work. They must preserve the participant’s lack of expectation that an unusual object will appear.
After someone notices the first surprise, later surprises are no longer equivalent. The observer may start deliberately searching for oddities. Combining all those detections into a single score can therefore blur the distinction between genuinely unexpected events and anticipated ones.
The review also separates inattentional blindness from other reasons for missing something, including distraction and difficulty detecting rare but expected targets. A missed finding by itself does not identify which process caused the miss.
Those distinctions prevent one memorable image from becoming an explanation for every error. Attention research needs to specify the task, the observer’s expectations and the way noticing was measured.
The lesson is about the search, not just the searcher
The same review discusses additional review by people with different task goals or expectations as a possible avenue for addressing unexpected findings. It presents this as a direction to consider, not a proven universal remedy arising from the gorilla experiment.
That shifts the useful question from whether a professional is generally observant to how a particular task directs observation. A person can be highly skilled at finding the requested feature without taking in every other meaningful detail at the same time.
Looking and noticing are related, but they are not interchangeable.
The 2013 experiment made that gap unusually easy to describe. A gorilla was placed inside a medical image, and most expert readers did not report it. The striking detail is not that the image defeated their eyesight. It is that something their eyes encountered could remain outside the experience they later described.