Uploaded by iuytfrdesdfrtyuiolk

Piliavin et al

advertisement
PILIAVIN et al.
Asim Masood
asimmm196@gmail.com
+923002626209
 A young woman by the name of Kitty
Genovese was attacked late at night by a
black man near her apartment compound
 38 people witnessed this incident but no one
took any immediate action
 When interviewed later by police, most of
the witnesses reported saying they assumed
someone else might have already taken
action and called the police
 This lead to the theory of diffusion of
responsibility being developed which
suggests that in an emergency situation,
when there is a large number of people
present, each person will diffuse or pass
their responsibility of helping on to one
another, and hence, there is a less likely
chance for the victim to receive immediate
help
 Conversely, the smaller the group size, the
greater the chance of help being received
 Two researchers by the name of Darley and
Latane decided to test the diffusion of
responsibility hypothesis in a lab
environment
 They conducted two studies in controlled
environments and proved the theory of the
diffusion of responsibility
 However, they did not test the theory in an
ecologically valid or natural setting
 Diffusion of responsibility/bystander
behavior/bystander helping/bystander
apathy: in an emergency situation, when
there is a large number of people present,
each person will diffuse or pass their
responsibility of helping on to one another,
and hence, there is a less likely chance for
the victim to receive immediate help
 Conversely, the smaller the group size, the
greater the chance of help being received
 Piliavin wanted to test the diffusion of
responsibility hypothesis in the natural
setting of the New York Subway
 To test whether helping behaviour is affected
by the condition of the victim; the race of
the victim; and the modeling conditions
 Research Method:
 Field Experiment: Conducted at the New York
Subway from 59th Street (Harlem Station) to 125th
Street (Bronx Station)
 Observation: Through two covert female observers
 IV:
 The condition of the victim –
Drunk Condition (operationalized by having the victim
hold a brown paper bag with an alcohol bottle, and
smelling of alcohol himself)
 Ill/Cane Condition (operationalized by having the victim
pretend to be ill and hold a black colour cane)

 IV (continued):
 The race of the victim –
Black
 White

 Modeling Conditions (operationalized by having the
model stand in either the critical or adjacent area
and help after either 70 or 150 seconds) –
Critical Early
 Critical Late
 Adjacent Early
 Adjacent Late

 DV: Helping Behaviour (operationalized by
recording the time taken and number of people
to help)
 Experimental Design: Independent Measures
Design. Different participants were intended for
each trial however, it is possible that some
participants might have been repeated as they
might take the same route on a daily basis, but
this is not what the researcher intended as it
was not in their control
 Approximately 4453 passengers who boarded
the subway
 55% white and 45% black
 Both genders and of different ages and
backgrounds
 On average 43 passengers were on the
subway per trial of whom 8-10 were in the
critical area on average
 Sampling technique: Opportunity (as they
were readily available)
 16 experimenters (4 victims, 4 models, 8
observers) in total who were split into 4 teams of
4 each
 They were General Studies students from
Columbia University
 Each team had 1 male victim, 1 male model, and
two female observers
 During each trial, there would be only one team
of experimenters
 Teams 1 and 2 began on day 1 with the cane
condition
 Teams 3 and 4 began with the drunk condition
 They were meant to alternate each day but on
day 4, Team 2 did not do as instructed as the
victim did not like playing a drunk man
 Victim – There were four victims in total, but
only one per trial
 Victim was always male and the age range was
25-36
 3 were white, 1 was black
 They were all identically dressed in Eisenhower
jackets and trousers
 The victim would pretend to be either drunk or
ill whilst also being either black or white
 The victim would always collapse in the critical
area of the subway 70 seconds into the journey
and lay on the floor staring at the ceiling till he
received help
 Model – There were four models in total, but
only one per trial
 Model was always male and the age range was
24-29
 They were all identically dressed in casual
clothing
 The model’s job would be to wait either 70 or
150 seconds after the victim collapsed, to see if
any of the passengers would help, before moving
from either the critical or adjacent area to go
and help the victim
 Observers – There were eight observers in total,
but only two per trial
 Observers were both female
 Both observers sat in the adjacent area
 Observer 1 was closer to the critical area and
recorded the race, sex, and location of everyone in
the critical area as well as the race, sex and location
of those who helped in the critical area
 Observer 2 was further away and recorded the race,
sex and location of everyone in the adjacent area
along with the race, sex and location of those in the
adjacent area who helped the victim. She also
recorded the time it took for the first person to help
as well as the time taken for someone to help after
the model helped the victim
 Both observers recorded comments of passengers as
well, particularly those sitting next to them
 On average, 43 passengers would get on
board the New York Subway on 59th Street at
Harlem Station
 It was a 7.5 minute journey to 125th Street at
Bronx Station
 The carriage was divided into two
compartments labeled the critical area (as
this is where the emergency took place) and
the adjacent area (as this was the adjacent
compartment)
 There would be 6-8 trials a day conducted from
11 AM to 3 PM, run by four teams of
experimenters who took turns running the trials
 A total of 103 trials were conducted over a
period of over two months
 70 seconds into the journey, the victim would
stagger forward towards a pole in the critical
area and collapse and lay there with his head
facing the ceiling till he received help
 The victim would either be black or white, and
drunk or ill, depending on whichever trial was to
be run that particular day
 The model would be present either in the
adjacent or critical area depending on the
modeling condition for that particular trial
and would wait either 70 or 150 seconds to
see if the passengers would help, before
going to help himself.
 The following are the modeling conditions:
 Critical Early – Model would stand in the critical
area and wait 70 seconds for someone to help
the victim. If 70 seconds passed and no one
helped the victim, then the model would go and
help
 Modeling Conditions (continued):
 Critical Late – The model would stand in the
critical area and would wait 150 seconds to see if
any passenger would help the victim. If no one
helped within this time, then the model would go
and help
 Adjacent Early – The model would stand in the
adjacent area and would wait 70 seconds for
someone else to help the victim, before going to
help himself
 Adjacent Late – The model would stand in the
adjacent area and wait 150 seconds for someone
else to help before going to help
 The two observers would record quantitative
data and note down the time taken for
people to help as well as the number of
people who helped
 They were covert observers in order to avoid
demand characteristics
 They also attempted to get comments from
passengers next to them for qualitative data
 When the train reached its destination, the
trial would end, and the passengers would
get off the subway without being debriefed
 The team would also get off and board the next
train heading in the opposite direction (125th
Street to 59th Street) and repeated the
procedure
 All team members boarded the train from
different doors
 Only one victim condition would be conducted
the entire day from 11AM to 3 PM (either drunk
or ill)
 From the 103 trials in total, 65 trials involved
the victim pretending to be ill, and 38 involved
the victim pretending to be drunk
 The dressing of the victim was identical –
Eisenhower jackets with trousers and no tie.
This was important to ensure that it was not
the appearance of the victim that led to
people helping as perhaps people would be
more likely to help a person formally dressed
rather than casually dressed
 The age of the victim was 26-35 – this was
important to ensure people would not be
helping on the basis of age because if an older
man would have played the role of the victim,
perhaps people would have helped because of
his age rather than his race or condition
 The gender of the victim was always male –
this was important as perhaps people would
have been more likely to help a female more
than a male, but since gender was not an IV of
the study, it was important to keep the gender
standardized of the victim
 The appearance, age, and gender of the model
– just as for the victim, it was important to
control these features of the model as well
 The train route was always the same to ensure
that the journey was always the same 7.5
journey. This was an important control because
if different journeys had been done, with some
being longer than others, then perhaps
participants would have helped more because
they were around the victim for a longer
period of time. Therefore they would not be
helping based on the IV, but rather, based on
the length of the journey duration
 The victim would always look and act the same
way when he was drunk – this was done to make it
standardized and easy to replicate
 The victim would always look and act the same
way when he was ill – this was done to make it
standardized and easy to replicate
 Both the above controls were important for
reliability
 The above controls were also important for validity
and the way they acted and looked made it
believable and hence would have avoided the
participants from showing demand characteristics
 62/65 trials received spontaneous helping for
the cane condition (people helped before the
model)
 19/38 trials received spontaneous helping for
the drunk condition (people helped before
the model)
 The median time taken to help the ill/cane
victim was 5 seconds
 The median time taken to help the drunk
victim was 109 seconds
 90% overall helpers were male
 55% overall helpers were white and 45% were
black
 When the model helped, more people joined in
 When the victim was white, 65% helpers were
white, but when the victim was black, it was a
50-50 split
 60% helpers in the critical area were male
 20% of trials, 34 people walked away from the
critical area
 Comments such as “I am not strong enough” and
“it is for men to help” were made by female
participants
 Hypothetical 3 people groups were compared
with hypothetical 7 people groups and it was
assumed that there would be more and faster
helping in 3 people groups than 7 people groups
(negative correlation between group size and
frequency of helping)
 However, actual 3 people groups had less
frequency of helping compared to actual 7
people groups as people were more likely to help
in the latter (positive correlation between group
size and frequency of helping)
 The results rejected the diffusion of
responsibility hypothesis
 Men are more likely to help than women in
emergency situations
 An ill person is more likely to receive help
than a drunk person
 Piliavin suggested an alternative explanation
for the results that were observed in order to
explain why the hypothesis was not proven
 She called this the Cost-Benefit Model (aka
Cost Reward Matrix)
 Cost-Benefit Model – According to this, the
participants experienced an unpleasant
arousal when the victim collapsed, and in an
attempt to reduce that arousal, they would
either help, get someone else to help, or
walk away from the situation. Each person
according to Piliavin, before making a
decision, went through the following four
possibilities in their minds:
 Cost of helping – e.g. getting harmed
 Cost of not helping – e.g. guilt
 Reward of helping – e.g. praise
 Reward of not helping – e.g. not getting harmed
The study by Piliavin aimed at testing the diffusion of
responsibility hypothesis in the natural setting of the New York
Subway. One strength of the study was that it was a field
experiment and was conducted in a natural environment of a
subway which makes it ecologically valid. Furthermore, the task
of helping someone is close to everyday life and it is a task
people can expect to perform in their daily lives and is
therefore, high in mundane realism. A second strength is that it
is high in generalisability as the sample size is large of 4453
passengers who were 55% white and 45% black. They were also
both genders and people of various backgrounds and hence, it is
quite diverse and generalizable to a larger population. One
weakness of the study is that the validity might have been
reduced due to the fact that the same route was taken for all
the trials. This means some participants may have been repeated
for more than one trial, even though the researchers intended
for it to be an independent measures design. These possible
repeat participants might have suspected they are part of a
study and could have shown demand characteristics which would
lower validity. A second weakness is that of ethics as participants
were deceived into believing the victim had actually collapsed.
They did not give consent to be part of the study, and were not
debriefed which meant they could have experienced long term
emotional harm as they might have felt uneasy by the drunk
victim or guilty for not helping the ill victim. To conclude, the
study’s results rejected the diffusion of responsibility hypothesis
 Other points that could have been considered for
the evaluation:
 Strengths –
 Quantitative data (objective and numerical. Allows
for comparison of results. Give examples of
quantitative results)
 Qualitative data (subjective and in the form of
comments and emotions. Allows to understand
reasons behind behaviours. Give examples from
study)
 Validity (participants were deceived and therefore
may not have shown demand characteristic)
 Reliable (high controls – was repeated 103 times to
test for consistency of results through the same
standardized procedure)
 Weaknesses –
 Generalisability (all participants were people
travelling the subway at New York, and therefore
cannot be generalised to people of other cultures)
 Reliability (difficult to control all variables such as
number of people on the train or possible train
delays. This would make it more difficult to
standardize or replicate)
 Inter-Rater Reliability (the two observers were not
recording the same thing)
 Sampling technique (opportunity sampling – difficult
to control the features of the participants. They all
might be similar in certain ways as they are readily
available and perhaps from the same area)
 Situational – 62/65 trials the participants provided
immediate help when the victim was ill, whereas
19/38 trials the participants provided immediate help
when the victim was drunk. This suggests that it was
the situation of the victim being ill that lead to more
instances of spontaneous helping as perhaps the
situation of being around a drunk victim might have
been more uncomfortable for the participants than
being around an ill victim which is why the time
taken to help was longer.
 Individual – Some participants did not help due to
individual differences, such as females making
comments such as “I’m not strong enough to help”,
suggesting it was an individual characteristic that
affected their decision to help the victim, rather
than the situation itself
 Nurture: The study supports the nurture side
over nature as peoples’ helping behavior may
differ due to environmental factors. One
individual may be more helpful than another due
to the way they have been brought up, or due to
their surroundings, which could have led to them
learning to become a helpful individual.
 Nature: However, perhaps the victim collapsing
in a state of being drunk or ill may have created
an arousal in the passengers which may have
made them feel uncomfortable that might have
caused them to walk away from the situation, or
perhaps even go help the victim. This arousal
would be natural and therefore, can support the
nature side of the debate.
 The findings are useful to everyday life and
individuals can be made aware of the
concept of bystander apathy/diffusion of
responsibility, and this may allow them to
take immediate action in emergency
situations, particularly in large groups, so
that the victim is helped immediately, and
others may also join in to help
Download