Key concepts in bystander behaviour

Bystander behaviour refers to how individuals respond when witnessing someone in need of help, particularly in emergency situations. This area of psychology explores why people sometimes fail to intervene, even when assistance is clearly required.

Bystander apathy

Bystander apathy occurs when the presence of other people discourages an individual from helping in an emergency or high-danger situation. As a result, people are more likely to offer help when they are alone than when they are part of a group.

Diffusion of responsibility

Diffusion of responsibility is a key explanation for bystander apathy, where the responsibility to help is shared among the bystanders present. The more bystanders there are, the less responsible each person feels, and the less guilty they experience if they do not intervene.

Background to the study

Interest in bystander behaviour grew significantly after a real-world incident that shocked the public and prompted psychological research.

The Kitty Genovese case

In 1964, Kitty Genovese was murdered in New York, and a newspaper reported that 38 witnesses did not call the police. This case drew attention to why people fail to help in emergencies, sparking laboratory experiments by psychologists.

Previous research

Studies by Darley and Latané, such as the 'lady in distress' experiment, the 'epileptic seizure' study, and the 'smoke-filled room' experiment, investigated bystander behaviour in controlled settings. Their findings showed that as group size increased, helping behaviour decreased, which they attributed to diffusion of responsibility.

Aims of the study

The study aimed to investigate how the following factors affect helping behaviour in a real-life setting:

Specific factors investigated:

• Type of victim - Whether the victim appeared ill or drunk.
• Race of the victim - Whether the victim was black or white.
• Modelled help - The effect of a confederate providing help at different times (70 or 150 seconds after the collapse) and from different locations (critical or adjacent area of the carriage).
• Group size - The naturally varying number of passengers in the subway carriage and its impact on helping.

Methodology and design

This research was conducted as a field experiment, which involves manipulating variables in a natural environment to observe real behaviours while maintaining some experimental control.

Research method

The study used an independent measures design, meaning different groups of participants experienced different conditions without repeating trials. Participants were assigned to one of four main conditions: ill/black victim, ill/white victim, drunk/black victim, or drunk/white victim.

Data collection techniques

Data were gathered through covert observation, where observers secretly recorded behaviours without participants knowing they were part of a study.

Roles of observers:

• Observer 1 - Recorded the sex, race, and location of passengers in the critical area; the total number of people in the carriage; and details of helpers (including their race, sex, and location).
• Observer 2 - Recorded the sex, race, and location of passengers in the adjacent area, and the time taken for the first passenger to help (known as latency).
• Both observers noted any comments made by passengers and tracked movements of passengers out of the critical area.

The subway carriage was divided into two areas for observation:

• Critical area - Where the victim collapsed, including the motorman's box and exit doors leading to the next carriage.
• Adjacent area - The rest of the carriage, with additional exit doors.

Observers were positioned in the adjacent area to record behaviours discreetly.

Sample

The sample consisted of 4,450 unaware passengers on the New York subway, including both men and women. Approximately 45% of the passengers were black and 55% were white. Participants were selected using an opportunity sampling technique, meaning they were whoever happened to be travelling on the 8th Avenue subway line between 11 a.m. and 3 p.m. on weekdays.

Procedure

The study was carried out on express subway trains in New York, using a standardised seven-and-a-half-minute journey for each trial to ensure consistency.

Steps in the procedure:

1. Two male confederates (one acting as the victim and one as the model) and two female observers boarded the train through separate doors.
2. The observers took seats in the adjacent area, while the victim stood in the critical area. The model's position varied between the critical or adjacent area depending on the condition.
3. Seventy seconds after the train left the station, the victim staggered forward and collapsed, lying still and looking upwards until helped.
4. If no help was offered, the model (confederate) stepped in to assist the victim to his feet.
5. Observers recorded data throughout, and the team repeated the process for multiple trials across different conditions.

Variables

Variables were carefully defined to test the study's aims.

Independent variables

• Type of victim: ill (65 trials, carried a cane) or drunk (38 trials, smelt of alcohol and carried a liquor bottle).
• Race of victim: black or white.
• Modelled help: timing (70 or 150 seconds after collapse) and location (critical or adjacent area).
• Group size: naturally occurring number of passengers in the carriage.

Dependent variables

• Frequency of helping (how often help was offered).
• Speed of helping (time taken to respond).
• Race and sex of the helper.
• Movement of passengers in and out of the critical and adjacent areas.
• Comments made by passengers (qualitative data providing insights into thoughts and justifications).

Controls and ethical issues

To ensure reliability, several controls were implemented, but the study also faced ethical challenges due to its covert nature.

Controls:

• All trials used the same seven-and-a-half-minute train journey.
• Victims wore identical clothing (old trousers, jacket, no tie) and collapsed in the same manner and location after exactly 70 seconds.
• Team members started each journey in fixed positions, such as Observer 1 near the exit door in the adjacent area and Observer 2 in the far corner.

Ethical issues:

• Lack of informed consent - Passengers were not told they were part of a study.
• No debriefing - Many passengers left the carriage at stops, making it impossible to explain the experiment afterwards.
• Potential psychological harm - Witnesses might have felt shocked or distressed by the simulated emergency.
• Deception - Participants were unaware of the confederates' roles, which could have influenced their emotional state.

Results

The study produced both quantitative and qualitative findings, revealing patterns in helping behaviour.

Key quantitative results:

• Type of victim - Help was offered in 62 out of 65 trials for the ill victim, compared to only 19 out of 38 trials for the drunk victim.
• Race of victim - There was a tendency for same-race helping in the drunk condition.
• Modelled help - Early modelling (at 70 seconds) prompted more additional help than late modelling (at 150 seconds).
• Group size - There was a weak positive correlation between larger group size and helping, but this was not statistically significant.
• Sex of helpers - 90% of first helpers were male.
• Movement - Passengers left the critical area in 20% of trials.

Qualitative results:

Female passengers made comments such as: "It's for men to help him" or "I wish I could help him – I'm not strong enough."

Conclusions

The findings led to several key conclusions about why and how bystanders respond in emergencies.

Arousal and cost-reward analysis

Observation of an emergency creates unpleasant arousal (emotional discomfort) in bystanders, which they seek to reduce. The strength of this arousal depends on factors like empathy for the victim, physical distance from the emergency, and the time elapsed without intervention. Bystanders reduce arousal by helping directly, seeking help from others, or leaving the scene.

Cost-reward matrix

Individuals weigh the costs of helping (such as effort, personal ability, or disgust) against the costs of not helping (such as self-blame or guilt) and potential rewards (like reduced arousal or social approval). This decision-making process influences whether help is offered.

Challenge to diffusion of responsibility

Unlike lab studies, no strong evidence supported diffusion of responsibility here – help was offered frequently and quickly.

Evaluation

The study has both strengths and weaknesses, which affect its overall quality and applicability.

Strengths and weaknesses

Issues and debates

This research touches on key psychological debates about what drives behaviour.

Individual versus situational explanations

Situational factors, such as an ill victim collapsing, strongly triggered helping – aid was offered without a model in every ill trial. However, individual characteristics also played a role, with 90% of first helpers being male. The cost-reward matrix suggests that personal appraisals of costs and rewards vary between individuals, blending situational and individual influences on helping behaviour.

Applications to everyday life

Insights from this study can inform practical strategies to improve responses in emergencies.

Education and training programmes

Programmes could teach young people about the cost-reward matrix, helping them recognise their own biases and emotions in emergency situations.

Changing responses to emergencies

By raising awareness of factors like victim type and modelled help, training could promote quicker interventions in public settings, potentially saving lives in real-world crises.