In an experiment in which a student simulated a seizure, helping was inhibited by

The Kitty Genovese murder in 1964 where 38 neighbors watched and listened but did not act to help or call police shocked the nation. However, the neighbors reactions were not much different than in other emergency situations where people "watch the drama in helpless fascination". Why do people who are so willing to help in non-emergency situations don't in emergency situations?

Characteristics of Emergencies
First, there are few positive rewards in an emergency. Life is threatened for the victims and the helpers. Second, it's an unusual event so reactions are untrained and unrehearsed. Emergencies come without warning with no practiced responses to fall back on. Yet it requires instant action. It puts the potential helper in alot of stress.

A Model of the Intervention Process
An intervener must make a series of decisions. First, he must notice the event and then interpret it as an emergency. Then he must decide if he has responsibility to act, and if so what form of assistance he should use. Should he help directly or call the police? Then he must decide how to act and implement his choice.

Of course, in a real emergency a person isn't so rational as this. Also, his decisions affect himself as much as the victim. The bystander can only gain with pride and a hero's status -- but he risks being a failure, getting sued, or even attacked or wounded himself.

Social Determinants to Bystander Intervention
When a person happens upon an ambigious "situation", the person may look to other's behavior to see if they observe it as an emergency. An individual, seeing the inaction of others, will judge the situation as less serious that he would if alone".

Males are expected to react to stress by being calm and collected. If they react to the emergency by intial calm inaction, this may be interpreted by others as an assessment of non-emergency. A 'pluralistic ignorance" my develop.

Thus, people may react less to an emergency if they are in a group situation than if they are alone.

Experiment 1. Where There's Smoke, There's (Sometimes) Fire
They had subjects began to fill out questionnaires in a room to which they began to add smoke. In one condition the subject was alone. In another three naive subjects were in the room. In the final condition one naive subject and two confederates who purposely noticed and then ignored the smoke (even when the room became hazy from all the smoke).

75% of alone subjects calmly noticed the smoke and left the room to report it. But only 10% of the subjects with confederates reported it. Surprisingly, in the three naive bystander condition only 38% reported the smoke.

Most subjects had similar initial reactions. Those that didn't report it all concluded that the smoke wasn't dangerous or was part of the experiment. No one attributed their inactivity to the presence of others in the room.

Other studies have shown that togetherness reduces fear even when the danger isn't reduced. It may have been that people in groups were less afraid and thus less likely to act. Or people were inhibited to show fear in a group situation. However, from post-interviews it was clear that people didn't act because they concluded the situation wasn't a threatening situation.

Experiment 2: Lady in Distress
In this experiment subjects either waited alone, with a friend, with a passive confderate, or with a stranger in a room. The room was separated from another room by a curtain (which they passed on their way to their waiting room). The experimenter who led them there returned to other room and left, turning on a tape recorded that simulated a fall and subquent moaning about a hurt leg (total time 130 seconds).

They measured the % who took action and how long it took them to act.

Results
Overall, 61% pulled back the curtain to check on the experimenter. 14% entered via another door, and 24% simply called out. Nobody went to report the accident.

70% of alone subjects reacted, but only 7% of those with passive confederates reacted. The subjects with confederates became confused and frequently looked over at the confederate. Only 40% of stranger pairs offered to help.

70% of friend pairs helped (same as alone group), which shows some inhibition because given the 70% alone rate we would expect a 91% rate with no inhibition.

The interveners claimed they acted because the fall seemed serious and it was "the right thing to do". The non-interveners said they were unsure what happened but decided it wasn't serious, and some felt they didn't want to embarass the researcher. Again, people felt they weren't highly influence by others in the room.

The results confirm results in the Smoke study. It seems that the risk of inappropriate behavior is less with friends, and friends are less likely to develop "pluralistic ignorance".

Experiment 3: The Case of the Stolen Beer
This experiment tested whether group influences would increase intervention if a villian was involved.

The staged a shoplifiting theft of a case of beer at a liqour store. They had two variables - one or two customers in the store, and one or two "robbers". Overall 20% of subjects reported the theft spontaneously, and 51% reported upon prompting by the store owner (who had gone in the back during the robbery).

One or two robbers made no difference. Sex made no difference. 65% of single customers reported the theft. But only 56% of two-customer setups made a report (less than expected).

Social Determinants of Bystander Intervention, II
All of the above experiments concerned whether a bystander noticed and concluded there was an emergency. He must also decide what responsiblity he has and what form of assistance it would take.

If there are multiple people at an emergency, the overall responsibility for one individual is reduced. Or they may assume that others have already responded to the emergency, so no one acts first.

Experiment 4: A Fit to Be Tied
This experiment tested what people would do if they witnessed an emergency with the knowledge others are present but can't see or hear them.

They put a naive subject in a room and told him that they were to talk with others about normal stress problems with other student who were similarily in isolated rooms to ostensibly preserve anonymity. Actually, all the other students were on tape. One of the other students became a victim that suffers a seizure and calls for help.

They varied the perceived number of people in the discussion group for two people (subject and victim), three person, and six person. They also varied the three person group by changing the other bystander (female, male, and a male pre-med student with emergency training).

Finally, they set up two more conditions. One with the subject and a real friend as bystanders, and one where six real subjects had prior contact and a brief "encounter" with the percieved victim.

95% of all subjects responded within the first 3 minutes. 85% of perceived alone subjects left their cubicle before the victim finished speaking to report it. Only 31% who thought there were four other bystanders did so. 100% in the two real person condition, but only 62% in the six person condition reported the emergency. Sex of bystander and medical competence had no effect on the results.

Being in the perceived presence of a friend significantly increase the speed of response. It seems that responsibility does not diffuse across friends. Also people who had briefly met the victim were significantly more likely to respond quicker to his pleas. It seemed the ability to visualize the victim help spur action.

Even those who didn't report the emergency showed signs of genuine concern. They were often nervous and trembling. They seemed to be in a state of indecision about responding.

Again, subjects were aware of others, but did not think they influenced their thinking.

Social Determinants of Bystander Intervention III
Basically, these experiments show there are strong situational factors that can inhibit people from acting in emergencies.