Watch People Try To Act Cool In A Room Filling With Smoke

This video is a recreation of an experiment first done in 1968. Subjects were observed reacting to the room they were sitting in slowly filling with smoke. Surprisingly few of them left. Take a guess as to why they stayed.


Bibb Latane and John Darley called up male Columbia students, and asked them to come to the psychology department and fill out a survey. When the students arrived, the scientists showed them to a room, gave them their questionnaire, and left. Within a minute, smoke started pouring into the room from a vent on one wall. The psychologists watched to see what the subjects would do, and found that what they did varied greatly depending on who their companions in the room happened to be.

When the students were alone, three-quarters of them got out of the room, found a person in hallways outside, and reported the smoke. They did so within about two minutes.

Sometimes the students shared the room with two paid confederates of Latane and Darley. The subject's two companions were told to act as if nothing at all were happening as smoke slowly filled the room. In these cases, the subjects waved their hands in front of their faces, and rubbed their eyes, and totally failed to try to save either themselves or anyone else in the building. Only one in 10 left the room and told the researchers outside about the smoke. Even when the poor students had to open a window to breathe, they didn't think to talk to anyone.

It's natural for people to look at what the people around them are doing in an unsettling situation. Some might call it conformity, but really it's just gathering information. Though today most subjects walk into a psychology experiment expecting some dirty tricks, it's wasn't the same in 1968. What's more bizarre, everyone in a room being a paid member of a conspiracy, or that the room sometimes gets a bit smoky?

Still, there was an element of the herd in the subjects' behavior. The scientists calculated that, as 75% of subjects on their own reported the smoke, a group of three "naive" subjects all put in the same room would report the smoke 98% of the time. They over-estimated to the tune of 60%. Only about 38% of the time did any single member of the group of three report the smoke.

The video is a recreation of the experiment. As we can see, the person on their own gets out, while the person with calm companions suffers in silence.


[Via Group Inhibition of Bystander Intervention In Emergencies]



Though today most subjects walk into a psychology experiment expecting some dirty tricks, it's wasn't the same in 1968.

So true. On the extreme end of that, I stayed suspicious even just walking into certain buildings on campus known for psych shenanigans. I remember going to my stats class one night, and some students were recording people's reactions to keys they'd left in the middle of the hallway. They wanted to see who'd pick them up to turn them in and who ignored them. But I was wise to their game and shared a laugh with the experimenters as I passed by.