Review: Social Approach: Piliavin (Subway

Samaritans)
Psychology Diffusion of responsibility
Being The reason people do not help people in need when in a group is the responsibility is equally shared
Investigated among the group

AIM 1. study bystander behavior outside the lab, in a realistic setting where participants would have a
clear view of the victim
2. see whether helping behavior was affecting by 4 variables: victims responsibility, race, effect of
modeling and group size
Hypothesis 1. An individual would be more inclined to help someone of their own race
2. help would be offered more and in a quicker fashion for the ill victim (cane) over the drunk victim
3. It is expected that modeling would increase help
4. Group size will not matter when you can see the victim.
Research Field experiment: On a New York city subway
Method/
Design
Independent Groups- study s repeated on different days and with different participants for each
condition
Variables: IV
IV & DV 1. Victims personal responsibility: Cane (ill- low responsibility) or smelling of alcohol and carrying a
bottle in a paper bag (drunk- high responsibility)
2. victims race- white or black
3. presence of a model: whether the male model either close or distant helped after 70 0r 150
seconds
4. # of bystanders: how many people were present in the vicinity
DV

Helping: measured as
Time taken for the first passenger to help
Total number of passengers who helped
Sampling Opportunity sample: participants were passengers on the train, not deliberately selected
Method

Sample 4450 people; 45% black 55% white; avg 43 people per carriage; 8.5 in critical area

Procedure Where:
“lab on wheels” NY Subway 1968 between Harlem and the Bronx weekdays between 11am and
3pm. 7.5 minute gap between stations

A team boarded the train using different doors. Female confederates sit and record data covertly.
Male model and male victim stood. Victim always stood next to the pole in the critical area. Shortly
after the train had passed the first station (70 seconds) the victim staggered and collapsed. Until any
help was given he lay on the floor looking at the ceiling.

If the victim did not receive help by the time the train got to the next station the model helped him
back to his feet, the team gets off the train, waited and boarded a train in the opposite direction
and repeated. 6-8 trials per day. On any given day all victims were the same condition
Procedure Teams:
Continued Groups of 4
2 observers both female- sat in critical and adjacent areas
Victim is male- stood at the pole in the center of the critical area
Model is male remains standing
Female observers: Critical Area: race, sex, and Adjacent Area: recorded the race,
What did they location of every person in the sex location of everyone in the
observe (QL data) critical area and the total # of adjacent area. Time taken for the
people in the subway car. first person to help. Amount of
Recorded the total number of time it took for someone to help
people who gave help (race, sex after model. Comments of people
and location). Comments of people in the car and attempted to get
in the car and attempted to get comments from the person sitting
comments from the person sitting next to them
next to them
Male confederates: Victim: male 26-35. 3 white 1 Model: male 24-29, dressed
What did they do? black; dressed identically in identical in casual clothes;
Where? What did Eisenhower jackets; old trousers
they look like? and no tie.
38 trials they smell like alcohol
65 trials sober with cane
Conditions: Victim- Drunk: smelling like alcohol and Ill: cane
How did they act? carrying s bottle in a paper bag
Props?

Conditions: Model Critical Early: model in critical area Adjacent Early: model in adjacent
Where were they and and helped after 70 seconds area and helped in 70 seconds
how long did they
wait until they helped Critical Late: model in critical area Adjacent Late: model in the
No model also and helped after 150 seconds adjacent area and helped after 150
seconds
Data Collection Female observers: Critical Area: race, sex, and Adjacent Area: recorded the race,
What did they location of every person in the sex location of everyone in the
observe (QL data) critical area and the total # of adjacent area. Time taken for the
people in the subway car. first person to help. Amount of
Recorded the total number of time it took for someone to help
people who gave help (race, sex after model. Comments of people
and location). Comments of people in the car and attempted to get
in the car and attempted to get comments from the person sitting
comments from the person sitting next to them
next to them
Results Spontaneous:
Overall, 93% spontaneous help (subway riders helped before the model)
60% had more than one helper
No difference between black/white, drunk/ill
st
1 helper is the crucial thing, if one person helps the others join in
 Drunk: Ill:
• 75% help w/model (12/16) Cane victim-
• 50% spontaneous help of total • 100% help w/model (3/3)
trials (19/38) • 95% spontaneous help of total
• Over 70 seconds in 83% of trials trials (62/65)
• avg of 109 seconds to help (21x • 100% of the cane trials had help
longer!) • Over 70 seconds in 17% of trials
• avg of 5 seconds to help

Who Helped:
Men were significantly more likely to help
• 90% of first helpers even though critical area was 60% male and 40% female
Neither race was more helpful overall
• 64% of first helpers where white (not significant) critical area was 55% white 45% black
Was there same race helping?
• When the victim was white 68% of first helpers were white (32% black) significantly above
55%
• When the victim was black only 50% of first helpers where white (50% black)
• Drunk condition Tendency toward same race helping
• In the ill (cane) condition there was no difference between black and white helper
In the drunk condition mainly members of the same race came to help, black were less likely to be
helped
Effective of Modeling:
The model intervening after 70 seconds was more likely to lead to help from other passengers (9
cases) than the one intervening after 150 seconds (3 cases)
Researchers noted that because passengers helped spontaneously in the vast majority of trials,
there were to few cases of helping after modeling to analyze in detail
People Move Away:
• In 20% (21/103) of the trials, subway riders moved away from the critical area (34 people)
with more leaving in the drunk than ill trial
• Mostly females, noting their size/strength
Qualitative:
• In 20% (21/103) of the trials, subway riders moved away from the critical area (34 people)
with more leaving in the drunk than ill trial
• Mostly females, noting their size/strength
Conclusion Was diffusion of responsibility found?
No ‘diffusion of responsibility’ was found
1. against hypothesis derived from laboratory studies that the more people present, the less help
would be offered In fact, there was a trend for more help being offered with the larger group size
2. Authors suggest this may be because the Ps were face to face with the victim unlike in the lab
experiments where they could not see them and maybe because they had no where to go? Also, the
more people, the more potential helpers present?
Main Conclusions:
1. An individual who appears to be ill is more likely to get help then one that appears drunk
2. Given a group of both men and women, when a male is the victim, men are more likely to
help then women
3. Given a mixed racial group; Same-race helping is more likely, particularly when the victim is
drunk compared to ill
4. There is no strong relationship between the number of bystanders and the speed of
helping: Diffusion of responsibility was not found
5. The longer the emergency continues without help being offered:
1. The less impact a model has
The more likely it is that individuals will leave the immediate area

Evaluation: Difficult to generalize past the sample of urban city dwellers. People in urban areas are more used
Generalizabilit to deindividuating (loosing their sense of identity) and feeling anonymous whereas people who live
y in rural areas may act differently. Also people in different countries may help at different rates

Evaluation:
Reliability

Evaluation: Results can be used to educate people about intervention, try to breakdown stereotypes, children
Application to could be educated to help others no matter what
everyday life

Evaluation: The study told us the type of victim can affect how long a person takes to help or whether people
Useful help at all.
Can use this to educate people that we should help people no matter what because the longer it
takes to help, the more likely the person will suffer
New Theory:
Cost benefit model: involves a decision-making process in which a person weighs up both the
advantages and disadvantages of helping. If it seems beneficial to help, them the person is likely to
do so, if the risks are too great, they may refrain.

Observation of an emergency creates an emotional arousal state.

Arousal is greater: if we can empathize with the victim; are close to the emergency; and the
longer the emergency continues
Arousal can be reduced if: helping directly; getting help; leaving the scene; rejecting the victim as
undeserving of help

Response chosen is a function of the cost-reward matrix

Cost- effort, embarrassment, disgusting or distasteful experiences, physical harm, self blame,
censure from others
Rewards- praise from self, victim or others; reward associated with not helping (continuing other
activities)
Evaluation: S: realistic setting: Ecological Validity: subway is not artificial, people travel daily and the event is
Validity realistic, people falling

Evaluation: Deception: participants did not know they were part of a study and did not know they were
Ethics watching a fake event

Informed consent: participants did not know they were part of a study so they could not give
permission

Protection: witnessing someone collapse can be distressing (and some people did leave the scene);
participants were not protected from experiencing psychological stress

Debates: Situational: the types of victim formed the situation and this caused different behaviors of helping
Individual v. and not helping
Situational
Individual: the sample was large and varied, it may have been a certain personality type that was
making people help or not
Debates: N/A
Nature v.
Nurture
Debates: Use N/A
of Children

Debates: Use N/A

of Animals