Sunday 26 November 2017
Piliavin's subway samaritan study (1969)
Context/Background
Bystander behaviour= people who witness events and choose whether to intervene or not, e.g. the Kitty Genovese murder, screamed for help whilst being stabbed in an alleyway no one watching called the police
Key issue deciding whether we help or not is whether we see it as our personal responsibility or not
Bystander apathy= where people fail to help someone in need when others are present
Diffusion of responsibility= the more people there are, the less likely we are to help because we assume someone else will do it
The condition of the victim influences helping behaviour, also if the victim is the same race as us = more likely to step in
Altruism- doing something selfless for somebody
Extrinsic reward- external physical reward, e.g. money
Intrinsic reward- internal good feelings about ourself or less guilty feelings
Latene and Darley = Smoke in the waiting room, when there was stooges not reacting, the p was less likely to act on it
Piliavin wanted to improve ecological validity on bystander behaviour studies
Aim
To understand bystander behaviour outside of a lab, to test if people are altruistic and to see if helping behaviour is affected by
1. victim's responsibility for their situation, ill or drunk
2. race, b or w
3. effect of model on helping behaviour
4. size of potential helping group
Hypotheses
-the drunk person will get less help than an ill person
-people will help others the same race as them
-intervention of model will increase helping behaviour
-larger the group, less likely people are to help
Method
Field experiment on trains of 8th ave, NYC subway, 7.5 minute journey, took place over 3 months on weekdays
Iv's-
1. type of victim (drunk or with a cane)
2. race of victim (b or w)
3. effect of a model or no model (early or late help)
4. size of witness group on train (natural iv)
Dv's-
frequency of help, speed of help, race of helper, movement out of the critical area and comments from other bystanders
Procedure
Sample
-4.500 men and women on the subway on weekdays, from 11am-3pm over 3 months
-45% black, 55% white and 4 teams of researchers, 2 F and 2 M
Victims were male, around 26-35 and either smelled of alcohol carrying a bottle or were sober with a black cane, their behaviour was all standardised
The models were also male aged 25-29
Victims stood near a pole in the critical areas on the train carriages, after 70 seconds they stepped forward and collapsed, and remained on floor until helped. If he received no help the entire journey, the model helped him.
There was 6-8 trials a day, all using the same condition
Findings
The cane victim got helped 75% of the time but the drunk one got helped 50%
Help was offered faster for the cane condition, (on average 5 seconds) but (109 seconds) for drunk condition
60% of help was from 2 or more helpers
90% of helpers were male
People tended to help victims of their own race, especially in the drunk condition
No evident diffusion of responsibility
More comments made about the drunk victims
Conclusions
ill/lame victim = more likely to get help than if drunk
Men more likely to help other men
No relation between number of people and speed of help
When escape isn't possible = more likely to help
Helping = praise from others but also brings fear or disgust
No such thing as altruism- an opinion
Evaluation
Research method- field, so ecological validity was high but it was hard to control variables, and standardised behaviour from the participants but the passenger's behaviour was unpredictable and many possible extraneous variables, impact of time not measured.
Data type- both types, but more quant. qual was gathered from the comments of other passengers
Ethics- harm, distress, guilt and anxiety, no consent, deception, no chance to withdraw and no debriefing. basically horrible for ethics. only had confidentiality.
Validity- high ecological because the setting was natural, actor being male and young may have affected the results.
Reliability- low internal reliability and not all p's had the same experience
Sample- large sample so representative of the American public at the time, in the 60's, and no control over participant variables or effects, ethnocentric as they're likely all new yorkers, same p's may have been on the train for more than 1 trial
©
Psychology OCR notes | All rights reserved.