Lesson 1 of 10 · 16 min
Latané and Darley's bystander studies
Describe the seizure, smoke-filled room and lady in distress studies and interpret their findings on how the presence of others changes whether and how quickly people help.
What you will be able to do
- Describe the procedure of the seizure, smoke-filled room and lady in distress studies.
- Interpret their data on the proportion who helped and how quickly, by number of bystanders.
- State the bystander effect and what Latané and Darley concluded about why it happens.
Imagine you hear someone collapse in the next room. Would you go and help? Almost everyone says yes. Now imagine four other people heard it too. Most people say that makes no difference. Bibb Latané and John Darley showed that it does.
In the 1960s they ran a series of experiments in which participants faced a staged emergency, either alone or with other people present. In 1969 they brought the results together in an article called Bystander "apathy". The quotation marks around "apathy" are the point: they argued that bystanders who fail to help are not uncaring. Something about the presence of others stops them.
History
The case that started it
Their work was prompted by the 1964 murder of Kitty Genovese in New York, which a newspaper reported had been watched by 38 witnesses who did nothing. Later investigations found that report was inaccurate: far fewer people saw the attack, and at least one called the police (Manning, Levine & Collins, 2007). The story was still a powerful question, and Latané and Darley tested it under controlled conditions.
The seizure study (Darley & Latané, 1968). Seventy-two New York University students took part in what they were told was a discussion about the pressures of student life, held over an intercom so that everyone stayed anonymous. Each participant sat alone in a booth. In fact the other voices were recordings.
Early on, one "student" mentioned that he sometimes had seizures. Later in the discussion he began to choke, call for help and say he was having a seizure. The independent variable was how many people the participant believed could hear him: just themselves (a group of two, the participant and the victim), themselves and one other (a group of three), or themselves and four others (a group of six). The dependent variable was whether, and how quickly, the participant left the booth to find the experimenter.
| People the participant believed were present | Participants | Helped before the end of the seizure | Mean time to help |
|---|---|---|---|
| Participant and victim only (alone) | 13 | 85% | 52 s |
| Participant, victim and 1 other | 26 | 62% | 93 s |
| Participant, victim and 4 others | 13 | 31% | 166 s |
The more people the participant believed could hear, the less likely they were to help in time.
Which statement best describes the results of the seizure study?
Change the number of bystanders and switch between the seizure, smoke and lady in distress studies.
The smoke-filled room (Latané & Darley, 1968). Male students at Columbia University sat in a waiting room filling in a questionnaire. Through a wall vent, a stream of white smoke began to fill the room. The researchers watched through a one-way mirror to see whether the participant left to report it within six minutes of first noticing it. Participants were alone, with two confederates (actors working for the researcher) who had been told to notice the smoke, shrug and keep working, or in a group of three real participants.
| Condition | Reported the smoke within 6 minutes |
|---|---|
| Alone (24 participants) | 75% of participants (18 of 24) |
| With two passive confederates (10 participants) | 10% of participants (1 of 10) |
| Three real participants (8 groups) | 38% of groups (3 of 8) had anyone report it |
Look at the last row carefully. A group of three has three chances to report. If each person were as likely to report as someone alone (75%), then the chance that nobody in a group reports is , so about 98% of groups should have had at least one reporter. Only 38% did. Being in a group did more than fail to help: it actively held people back.
In the lady in distress study (below), 70% of participants alone helped. If two strangers each behaved like a person alone, independently, what percentage of pairs would be expected to have at least one helper? Give a whole number.
The lady in distress (Latané & Rodin, 1969). Male Columbia students filled in a questionnaire for a market research company. The woman who had greeted them went into her office next door, behind a curtain. A few minutes later they heard (from a tape recording) her climb on a chair, a loud crash and her crying out that her ankle was hurt and she could not move it. The measure was whether anyone went to help or called out an offer of help.
| Condition | Helped |
|---|---|
| Alone | 70% of participants |
| With a passive confederate | 7% of participants |
| Two strangers | 40% of pairs (at least one helped) |
| Two friends | 70% of pairs (at least one helped) |
Key idea
The bystander effect
Across all three studies, a person was less likely to help, and slower to help, when other people were present than when alone. The effect was strongest when the others were passive, and weaker between friends, who could read each other's reactions and felt less embarrassed.
Worked example
Interpret the seizure study's findings
In Darley and Latané's (1968) seizure study, 85% of participants who believed they were alone with the victim helped before the seizure ended, compared with 62% of those who believed one other person could hear and 31% of those who believed four others could hear. The mean times to help were 52 s, 93 s and 166 s.
Interpret these findings.
Step 1: Describe the trend in the proportion who helped
As the number of other people the participant believed were present rose from none to one to four, the percentage who helped before the seizure ended fell from 85% to 62% to 31%. With four others, fewer than half as many helped as when alone.
Step 2: Describe the trend in speed
The mean time to help rose from 52 s to 93 s to 166 s, more than three times as long with four others as alone.
Step 3: State what the results show
Both measures show the bystander effect: the presence of others made helping less likely and slower.
Step 4: Explain the result
Participants could not see or hear the other "bystanders" react, so they could not have been influenced by their calm. The best explanation is diffusion of responsibility: each believed someone else could help.
Interpreting means saying what the data mean, not just reading them back. The design rules out social influence, which points to diffusion of responsibility.
Try it yourself first, then check the working one step at a time.
Answer
Helping fell (85% → 62% → 31%) and slowed (52 s → 93 s → 166 s) as the believed number of bystanders increased: the bystander effect. Because participants could not see the others, the effect is best explained by diffusion of responsibility.
Method
Describe each trend with the numbers from the data.
State the overall finding in one sentence.
Explain it with the process the design points to.
Watch out for
Writing "people are selfish" or "apathy". Darley and Latané found non-helpers were anxious and conflicted.
Naming social influence: participants in separate booths could not see how others reacted.
Common mistake
Non-helpers were not uncaring
Darley and Latané noted that participants who did not leave the booth were not calm or indifferent: many were nervous, sweating and asked the experimenter afterwards whether the victim was all right. They had not decided not to help; they were caught in conflict. That is why the explanations in the next lessons are about how others change our interpretation and our sense of responsibility.
In the lady in distress study, pairs of friends were much more likely to have a helper (70%) than pairs of strangers (40%). Which explanation best fits this finding?
Summary
Latané and Darley (1969) brought together staged emergency experiments: the seizure, smoke-filled room and lady in distress studies.
Seizure study: 85% alone, 62% with one other and 31% with four others helped before the seizure ended; mean times 52 s, 93 s and 166 s.
Smoke study: 75% alone reported the smoke, 10% with passive confederates, and only 38% of three-person groups had anyone report it (about 98% would be expected if people acted independently).
Lady in distress: 70% alone and 7% with a passive confederate helped; 40% of pairs of strangers and 70% of pairs of friends had a helper.
The bystander effect: the presence of others makes helping less likely and slower. Non-helpers were conflicted, not apathetic.
Finished the lesson?
Try the 3 quick checks above and the lesson completes itself, or mark it complete now.