Skip to main content

Sheeple in an Information Cascade


This video and accompanying article posted in August of last year describes an experiment involving herd behavior and social conformity. In the waiting room of a doctor’s office hidden cameras were set up, and about ten actors were seated when an unknowing person checks in and sits down in the waiting room with them. About every minute or so a beep goes off in the office, and all of the actors briefly and up and then sit back down. The person initially looks confused but after the first two beeps she begins standing up with them on the beeps without asking any questions. Each actor gets “checked in” and they leave the waiting room one by one leaving only the initial unaware subject of the experiment, who continues to stand for the beeps even when she is the only one left. Eventually more real patients come into the office, the first person asks the girl why she stands for the beeps and she shrugs and says everyone else was doing it before she came in. Successive patients fill the waiting room and without asking questions join in the standing behavior even without knowing why, and continue it once the initial girl leaves.

I had seen this video a while ago and thought it was amusing then, but now that I have some background information from the class I think it is even more interesting.  I think this experiment can be loosely defined in terms of Information-Based Imitation, or an Information Cascade. There are four components to an information cascade and going through them individually clearly shows the experiment meets the qualifications. The first is that there is some decision to be made, in the experiment the decision is to stay seated or to stand when the beep goes off. The second is that people make the decision sequentially over time, which is analogous to the new patients coming in the office and observing the standing behavior and deciding if they want to copy it or not. The third component is that each person has some private information, this one is a bit of a stretch but I think that in the context of the experiment each new patient witnesses a different group of people following or not following the behavior when they enter the room, but they don’t know what was happening before. The new patients could think that perhaps the receptionist explained why patients must stand for beeps before they got there, or maybe the patient who watches people stand has just never been to the office before and decides the people there know something that they don’t, so they should go with the crowd. The final qualification is that you can see what earlier people did but not what they know. Following from the previous example, new people entering witness people standing and sitting but do not know why. Based on these conclusions I think that this social experiment is a unique example of an information cascade taking place without the participants even knowing that they are involved!


Leave a Reply

Blogging Calendar

October 2017