Simons and Chabris Inattentional Blindness PDF

Background: Neisser created a task where participants had to monitor basketball passes of either the white- or black- shirt team, which were videoed separately and superimposed over each other. They found that only 48% noticed the unexpected event of a woman holding an umbrella, or a gorilla being p...

Background: Neisser created a task where participants had to monitor basketball passes of either the white- or black- shirt team, which were videoed separately and superimposed over each other. They found that only 48% noticed the unexpected event of a woman holding an umbrella, or a gorilla being present. When they did not have to monitor passes, all noticed the unexpected event. Aim: To investigate inattentional blindness for unexpected objects in live, dynamic (moving) scenes. Sampling method: Self-selected in response to advert for a study at Harvard University o ering compensation. Not much else to talk abt in the sample Sample: 228 majority Harvard undergraduate students. 12 di erent participants for the controlled observation. 36 participant's data was discarded as they either didn't pay enough attention or had already heard of the study (since it had already been done - background) Method: Independent measures design lab experiment. 16 conditions of 4 IVs: Unexpected event conditions (umbrella woman, gorilla woman) Cognitive load conditions (Hard: count the bounce and aerial passes separately; Easy: count total passes) Shirt colour conditions (monitor passes of black-shirt team, or white-shirt team) Transparency of video conditions (Live/opaque video: live lming of the white and black teams and unexpected event occurring at the same time; Transparent video: same as superimposed in that black, white team, and unexpected event are lmed separately, made transparent, and digitally superimposed over each other). Controlled observation procedure: participants watched an opaque video where the unexpected event lasted longer- 9secs- and was more salient- the gorilla stopped in the center of the scene and thumped its chest. Opaque- everything lmed at same time, whereas Transparent is recorded seperately and layererd. Materials: 2 transparent and 2 opaque, 75sec videotapes were produced of a white-shirted and black-shirted team of 3 players, bounce- and aerial-passing basketballs. In the middle of all videos an unexpected event would occur, lasting 5secs (tall woman with umbrella or shorter woman in gorilla costume). Gorilla was more unexpected, but umbrella women was taller. Procedure: Participants were tested individually and gave informed consent in advance. 1. Participants told they would be watching the basketball video and to count the number of passes (Easy condition), or number of bounce and aerial passes separately (Hard condition), of the black or white team. 2. After viewing the video, participants were immediately asked to write down their counts. 3. They were then asked the same 4 questions on the unexpected event, which became gradually more speci c, starting with whether they noticed anything unusual to whether they saw the gorilla/ umbrella woman. To verify their answer, they were asked to describe the unexpected event. 4. The researchers debriefed them, and asked if they’d heard of the study beforehand. They excluded their data if they had. Controls: The researchers analysed the standard deviations of the total counts to verify that the Hard condition was more di cult than the Easy condition. Participants found it harder to keep count, evidenced by higher standard deviations in the Hard condition than the Easy condition, showing the Hard condition required more concentration, increasing cognitive load. The researchers analysed whether there was a correlation between noticing the unexpected event and accuracy of count, and found no signi cant relationship, suggesting how well participants  were able to keep count did not confound results. Findings/ Conclusions: The controlled observation found that only 50% noticed the event when attending to the white team and carrying out the Easy monitoring task. This suggested inattentional blindness occurred even when the unexpected event lasted for a greater duration and was more salient. Signi cantly more participants noticed the unexpected event in the opaque condition than the transparent. This shows that inattentional blindness found in previous research was not a result of the superimposed event being more di cult to see. Signi cantly more participants noticed the unexpected event in the easy condition (64%) than the hard condition (45%). This shows that we have limited visual attention, and when our attention is being loaded by a di cult monitoring task, we’re less likely to notice other, unexpected events. The gorilla was noticed more by participants paying attention to the black team than the white. There was no di erence in how often the umbrella woman was noticed by those monitoring the black or white team (shows HOLISM: interaction between two IVs). Individuals more likely to notice unexpected events if events are visually similar to the events that they are paying attention to. Same experiment as background, but making the video better and measuring more things