Exploring the bystander effect
I just finished reading an astounding book, The Unresponsive Bystander: Why Doesn’t He Help? by Bibb Latané and John M Darley (Prentice-Hall, Inc, 1970). Latané and Darley set out to understand what makes the modern bystander so apparently apathetic and callous, watching but not helping as others are hurt, maimed and even die. They ask whether urbanization has created such extreme apathy and alienation that people have lost their humanity and no longer care about each other?
For motivation, they cite the killing of teenager Andrew Mormille, witnessed by many but helped by none, on a NYC subway; and the 1964 stabbing, rape and murder of Kitty Genovese.
Of the Genovese case, the authors repeat Abe Rosenthal’s version in which 38 neighbors watched and did nothing to help, a version that has been debunked by Kevin Cook more recently and decades after The Unresponsive Bystander was written. Despite the fact that the Thirty-eight Witnesses (AM Rosenthal, McGraw Hill, 1964) version of the Genovese story now appears to be apocryphal, there are plenty of anecdotes, then and now, of groups doing little or nothing to help others in distress.
The apparent diminution of social affiliation evidenced by a seeming lack of concern for others is the starting point for Latané and Darley.
Latané and Darley are psychologists and they painstakingly and methodically investigate the effect that the presence of others has an individual’s actions. They start with the premise that a change in the “basic motives underlying altruism” is unlikely. Evolution of a seismic shift in psychological motives over the last few hundred years due to the industrial revolution and the growth of urban living appears to them, and me, exceedingly unlikely.
In order to understand the psychology of helping, Latané and Darley start out with several innocuous forms of “help” that incur a very low cost. New Yorkers acquiesced to strangers’ requests for the time (85%) or directions (84%). Yet, only a third responded when asked, “Excuse me, I wonder if you could tell me what your name is?” or when asked for a dime. So far, so good; for low cost, non-dangerous and non-intrusive requests, New Yorkers are helpful at a high rate. The finding that New Yorkers balk at intrusive requests appears reasonable if one considers that yielding one’s privacy is, in and of itself, a cost.
The crux of the bystander effect was revealed by a series of experiments in which subjects (college students) went into a waiting room where they started to fill out a questionnaire.
In all of the experiments, there was a cover story, a ruse if you will, for the experiment. For example, in the smoke experiment, subjects were told that the experimenters were interested in the students’ reactions to New York. They then are asked to fill out preparatory forms prior to testing. The actual experiment takes place during this time, when the subjects believe that they are simply waiting for the experimental interview to occur.
After a few minutes in the room, white smoke started to flow into the room and continued to do so for four minutes. Within this time period, 75% of the subjects who were in the room alone (18/24) left the room to report the problem but only three of 24 (13%) people in the room as part of eight triplets did so. The difference between 75% and 13% is both large and significant. Furthermore, consider that if the three people in the triplets acted independently at the frequency that the individually tested people did, then >98% of people in triplets would be expected to act.
If you’re wondering about this calculation, think of it this way. Any member of the triplet could report the smoke. The only way that the smoke would not be reported is if all three people in a triplet failed to report the smoke. If there is a 25% chance that one person will fail to report smoke, then the chance that all three will fail to report smoke is 0.25 x 0.25 x 0.25 or (0.25)3 which amounts to a 1.6% chance.
So, the impact of two other strangers is to decrease the likelihood of helping from an expected 98% to 13%. Latané and Darley find the same effect – far more helping by single people than by individuals within groups – in a series of experiments employing a variety of issues:
- money being stolen from a receptionist’s desk
- beer being stolen from a liquor store
- a bookshelf falling on a woman in an adjacent room **
- one child bullying another child **
- a man having a seizure **
[In the situations involving people being hurt (marked with ** above), subjects heard previously recorded sounds but were not in visual contact with the victims. Of course, no one was in fact harmed.]
When the in-house experiments (all but the liquor store experiment) were over, the experimenters talked to the participants, asking them whether they “had experienced any difficulty while filling out the questionnaire.” The responses in the smoke experiment are exemplary. Participants that reported the smoke said that they thought something was wrong, that there may be a fire and that in any case, checking into the situation was warranted. In contrast, participants that did not report did not believe that the smoke was due to a fire. They had other explanations for the smoke. Some of these explanations were truly bizarre; not one but two subjects shared their thought that the smoke was a truth gas intended to make sure that they accurately filled out the questionnaire. In the other experiments, there were similar justifications for the strange goings-on, but only from those that did not act. Thus it appears that the just-so stories that reporters and non-reporters tell themselves are different. Could this be because, reporters and non-reporters sense the incidents differently?
A clue that reporters and non-reporters indeed process identical situations differently came from the smoke experiment. In this experiment, it was easy to tell the moment when subjects noticed the smoke: they turned their head and looked up toward the vent. Of subjects tested alone, 63% noticed the smoke within five seconds whereas only 26% of subjects tested in groups did so. Could the presence of others have put a damper on helping by actually changing the processing of sensory stimuli and thus perception?!
The idea that the perceptions of reporters and non-reporters could differ came into sharp relief in the experiment on children fighting. In this experiment, subjects (again college students) came in purportedly for market testing on games. When the subjects arrived they were ushered into a room marked Student’s Testing Center, which was separated by a heavy curtain from a room marked Children’s Testing Center. Shortly after entering the room, the subject (who was alone) heard a rapidly escalating fight over a toy. At one point one boy says “…Get off me. You cut me. It hurts so much. Help me, please. I’m bleeding.” The other boy’s voice says, “You asked for it. I’m going to really beat you up.” And so on.
Only one of 12 subjects reported the fight. But what is really, really interesting is that most (9/12) participants in this study explained afterwards that they did not believe that it was a fight! They thought that the boys were listening to a TV show or that it was a recording. These results suggested to the clever Latané and Darley that the possibility that subjects “were motivated to disbelieve the reality of the situation by their desires to avoid involvement.” And being clever psychologists, Latané and Darley directly tested this idea. They set up a condition that was just the same except that subjects were told of the children, “Somebody is in there testing them…they do tend to get a little noisy but please leave them alone.” Thereby absolved of responsibility, only one subject (1/8) failed to believe that the fight was real. The difference between 75% (responsibility) and 12% (no responsibility) of the subjects perceiving the situation as faked was significant. This is strong evidence that expectations of responsibility can “distort” perceptions.
There was only one situation – the one involving a person having a seizure – in which the non-reporters consistently believed that there was a genuine emergency and were distressed. In other words, the non-reporters perceived the situation in the same way as did reporters. The non-reporters said into microphones (that they believed were turned off), “Oh, what should I do?” and asked about the health of the seizing confederate immediately when the experimenters came in to terminate the session. Latané and Darley conclude that rather than having decided not to help, these non-responders simply had not figured out how to help.
In the end, Latané and Darley find four ways in which the presence and appearance of others suppress helping by an individual in an emergency situation:
- By acting, we risk looking foolish in front of others. This may not be a conscious influence. In the bullying experiment, the burden of responsibility led to internal pressure to interpret the situation as not requiring intervention.
- Others’ inactivity guides us to act similarly. Whereas others may only be inactive transiently, en route to figuring out what to do, the influence of this inactivity reinforces collective inaction.
- The effect of the above two influences is super-additive so that an individual’s perception of situations is heavily influenced by the apparent lack of reaction shown by others. In other words, perception itself is altered by the social circumstances, enabling a bystander to interpret a situation as less serious than it is.
- The presence of others diffuses an individual’s sense of responsibility, making it more acceptable to not act because it is perceived that others may act.
The authors also found that no personality variable predicted whether or not any given person would help. The implication of this finding is that there is no “they” or “them” to not helping. It is highly likely that we are all capable of watching others in distress without helping. We can then rephrase the problem of “why do some help and some not?” as “what are the factors that influence us, all of us, to help or not help in specific circumstances?” Although no personality is either particularly susceptible to, or immune from, the social influence to not help, people consistently believed that the presence of others had no influence on whether they themselves helped. The authors speculate that if we all had an understanding of the social influences on helping, we may be able to counteract our natural tendencies and act more pro-socially, even in crowds.
As I read this book, my thoughts kept wondering to non-human animals: would other mammals show a bystander effect? The authors quote one rat study which is largely non-informative and I think that the jury is out on this intriguing question. However a recent anecdote suggests that other mammals may not show the bystander effect. In a YouTube video, a baby elephant does not quite make a step-up and falls over on her/his back. Two adult elephants rush over to the baby’s aid, with a third following close behind.
I will end this post by returning to the authors’ Preface, where they write, “We believe in the existence of that legendary totem figure of all academic authors, the intelligent, well-informed layman who wants to find out more about the world in which he lives.” Amen to that. The authors succeed at writing in an engaging and accessible manner. My summary is just a poor attempt at summarizing a fascinating, elegant series of experiments told in crystal-clear, literary prose. For those of you who are interested in learning more, I highly recommend curling up on a couch some afternoon with The Unresponsive Bystander: Why Doesn’t He Help?. If you do, you’ll be glad that you did.