Piliavin et al.

PSYCHOLOGY BEING INVESTIGATED

BYSTANDER APATHY

Bystander: person present at the scene of a crime but does not intervene. 38 people ignored Kitty Genovese’s murder. Usually, bystanders don’t help because they perceive the victims negatively and judge their behaviour. 

DIFFUSION OF RESPONSIBILITY

The greater the number of people present at the scene of a crime, the lesser responsibility each of them bear to help. Duty to help is shared amongst everyone, so their guilt for not helping reduces. 

BACKGROUND 

People are less likely to help a victim if others are present. most studies on bystander apathy are done in labs. Piliavin a did this experiment to investigate factors affecting bystander apathy and good samaritanism.

AIM

Piliavin et al. wanted to see how the following factors affected help offered to a passenger who collapsed in the carriage: the type of vic: drunk or ill the race of the vic: B or W modelled help provided by another passenger number of people in the carriage (group size). 

HYPOTHESES 

• A helper would be more likely to help someone of their own race
• Ill vics would receive help faster and more frequently than drunk ones, because: drunk people receive less sympathy as they are responsible for their situation, people’s desire to help drunk people is overcome by the fear of being embarrassed, harmed, and disgusted by them. This also makes observers leave the scene of the emergency.

RESEARCH METHODOLOGY 

Field experiment; took place on a 7.5-min express train bw 2 NYC stations on the 8^th Avenue line. 

SAMPLE 

Approx. 4,450 people, weekdays between 11 a.m. and 3 p.m., from April 15 to June 26, 1968. Racial composition: 45% B, 55% W. Average number of people per car: 43. Average number of people in the critical area: 8.5. 

PROCEDURE 

A team of 4 Columbia students boarded a carriage on the New York subway train via separate doors. Two female observers took separate seats, while the two male students playing the roles of vic and model remained standing. The vic always stood in the “critical area”. Seventy seconds in, the vic staggers forward & collapses on the floor and is laid in a supine position until they receive help. If no help was given by the time the train reached the next station, the model helped the vic to his feet. The team would then re-board a train along the same route in the opposite direction, completing around 6-8 trials per day [t=103 trials]. 

VIC 

All vics were males aged 26-35, three W and one B. They were identically dressed in Eisenhower jackets, old slacks, and no tie. In 38 trials, the vics smelled of liquor and carried a bottle wrapped in a brown bag. In 65 trials, the vics were sober and carried a B cane. the vics behaved identically and were part of both conditions. 

MODEL 

4 W males aged 24-29, wearing informal unidentical clothes. There were 4 conditions for models used in the vic conditions: 

• Critical area- early: The model was in the critical area and waited 70 seconds [till they passed the 4^th station] to help.
• Critical area- late: The model was in the critical area and waited 150 seconds [till they passed the 6^th station] to help.
• Adjacent area- early: The model stood in the middle of the adjacent area and waited 70 seconds [till they passed the 4^th station] to help.
• Adjacent area-late model stood in the middle of the adjacent area and waited 150 seconds [till they passed the 6^th station] to help.

Equal model conditions, including the no model condition, were assigned to each team. It was later discovered that the trial conditions were distributed unevenly across B and W vics. Teams 1 & and 2 [W vics] started with the cane condition and Teams 3&4 [W and B vics] started with the drunk condition. The teams were told to alternate conditions but team 2 went against that on the first day; the vic didn’t like acting drunk so they ran the cane trial instead. Then, the Columbia student strike caused the teams to disband and the study was over. Now, teams 1 and 3 run trials on 3 days each while teams 2&4 run trials on 4 days each. In each trial, one confederate/observer recorded the race, sex, and location of all other riders in the critical area as well as the helpers. She calculated the number of people in the car and the number of people who helped. The second observer did the same for the adjacent area, as well as recording the latency of the first helper’s arrival in conditions where the model helped and didn’t help. Both observers recorded comments passed by other riders and attempted to elicit comments from riders next to them. 

RESULTS 

ILL/DRUNK: “Ill” person more likely to receive help (62/65 of trials) than a “drunk” one (19/38 of trials). The ill vic was helped 100% of the time when there was no model. 

RACE OF VIC: No tendency for same-race helping unless the vic was drunk. Drunk W vic helped 100% of the time, B vic helped 73 % of the time & more frequently by B helpers. On 9% of trials with a B vic, people left the critical area vs 5 % of trials with a W vic. 

Modelled help: Early models were more likely to elicit additional help than late models (4 vs 2). 

Group size: Weak positive correlation between group size & helping behaviour. Groups of ≥ 7 were faster to respond than groups of 3. Males were more likely to help the man than females. 90% of first helpers were male and 68% of helpers were W. In 20% of trials, people left the critical area and a total of 34 people left. The researchers reported a few interesting findings from bystander comments. Female comments included: ‘It’s for men to help him’, ’I wish I could help him, I’m not strong enough’, ‘I never saw this kind of thing before, I don’t know where to look’, ‘You feel so bad that you don’t know what to do.’ 

CONCLUSIONS 

• A drunk person is less likely to be helped than an ill person.
• If the vic is a male, males are more likely to help than women.
• There is a slight tendency for same-race helping to be frequent, if the vic is drunk.
• Diffusion of responsibility did not occur; no relationship between the number of bystanders and the speed of helping.
• The longer the vic doesn’t receive help: the less of an impact the model’s action will have on observers; people will leave the critical area; the more people discuss and comment on the event taking place.

ETHICAL ISSUES

INFORMED CONSENT 

Participants were not aware that they were part of a study so consent could not be taken. Being unaware of their responsibility to help a person in need in a subway could have caused them psychological harm. Some people’s comments were published in journals which could have embarrassed them, 

DEBRIEFING 

Due to the large sample size couldn’t be debriefed about the aim of the study and couldn’t answer any questions regarding their psychological state. Participants may have left the subway feeling guilty or distressed if they had not helped. 

METHODOLOGICAL ISSUES 

RELIABILITY 

Standardisation: The experimenters controlled many variables like what clothes confederates were wearing and when the collapse took place [after the first station].

VALIDITY 

Confounding variables 

Only one B vic being used makes it hard to identify whether he was helped because of his race or his condition. Some students felt embarrassed to act drunk, which contributed to how they were behaving and how people interpreted that behaviour. 

Lack of controls 

Some people may have been distracted and not noticed the collapse, showing that them not helping was not a conscious choice. 

Qualitative data 

For example, the comments passed by the bystanders allowed for deeper insight into the behavior of the bystanders so researchers could draw more accurate conclusions.

 OBJECTIVITY AND SUBJECTIVITY

QUALITATIVE DATA 

Allowed researchers to compare the results of the help received by drunk/cane and W/B vics; eg: the time taken for a drunk vic to receive help can be compared against the time taken for ill vics to be helped. 

GENERALISATION AND ECOLOGICAL VALIDITY 

Since the sample size was very large [4450 men and women, 45% B and 55% W], it can be argued that they represent the public of New York. The participants were only commuters on the New York subway, but if a different group like urban citizens, they would most likely not help due to their individualistic nature. Therefore the study lacks population validity and may not be applied to people living in more rural areas. The participants being unaware of the study prevented the display of demand characteristics and therefore the study has high ecological validity as it shows researchers how people would react if this was a real problem and not an experiment. 

LINKS TO APPROACHES AND ASSUMPTIONS 

Behavior can be associated with the presence of a group, assuming others will help causes individuals to feel less responsible. On all cane trials, someone helped; supports situational explanations for behaviour.