conducted in a train station on the
subway weekdays 11am-3pm
IVs: The race and type of victim
(drunk or ill), presence of
models and no. of bystanders.
DVs: Speed and
frequency of help.
Aim: To investigate the effect on
helping behaviour of the type of
victim (drunk or respectable) and
the race (black or white).
Evaluation
Strengths
high ecological validity in a
real-life environment
no demand characteristics
large sample
4450 people- more generalisable
Weaknesses
No informed consent,
deception and lack of
debriefing.
chance of causing distress and
social and emotional harm.
field experiment
with low controls
Results
90% of helpers were
male.
Slight
'same-race'
effect.
Black victims received
less help less quickly.
cane victim received
spontaneous help 95% of
the time.
cane victim was helped on
average within 5 seconds.
drunk victim was helped on average after
109 seconds.
drunk victim was spontaneously
helped 50% of the time.
Procedure
4 teams aged 24-29
4 students per team
2 female observers,
1 male confederate,
1 male victim
each trial lasted
7 1/2 minutes
70 seconds after the train
leaves, the victim collapses
if no help is given,
confederate steps in after 70
or 150 seconds in both critical
and adjacent
Triangulation- both qual and
quan data gathered
length of time, amount of helpers, race,
gender, comments and location of
surrounding people
e.g. "It's for men
to help him"
Background
March 1964- Nurse
Kitty Genovese was
stabbed in New York
City at 3am.
There were 38
witnesses who didn't
attempt to help her.
Diffusion of responsibility- the idea that people are less likely to help someone
if there are others present, as they perceive responsibility as shared between
all present, and see themselves as being less personally responsible.
Conclusions
An emergency situation
arouses the attention of a
bystander
people are more likely to help if they
feel physically close to the victim, for
example, the same race.
no evidence for diffusion of responsibility- no more or
less likely to help when surrounded by others.