Piliavin et al. (1969) Good Samaritanism: An
Underground Phenomenon?
Procedure: Aims:
The ‘type’ of victim was portrayed as either drunk or ill To investigate the effect that the type of victim would have on the speed and frequency
The ‘race’ of the victim was either black or white with which a person would respond to aid them in an emergency (as well as the effect of
Sample: 4,450 men and women passengers were present over the duration of the the race of the victim)
scenes being staged Investigated the effect of role model behaviours in emergency situations and the
● Approximately 55% were white and 45% black relationship between the size of the group and helping a victim
● The mean number of bystanders in the train carriage was 43 people
● The mean number of bystanders deemed to be in the ‘critical area’
(closer proximity to victim) was 8.5 people Results: Conclusions: Strengths:
The staged scenes took place on a 7.5 minute continuous train journey between Help before 70 seconds for the ill, white victim was Spontaneous helping occurred within 70 Conducted in a naturalistic
59th Street and 125th Street stations in the New York subway between 11:00am recorded in 52/57 trials seconds, and where a model assists a environment on passengers who
and 3:00pm during the period 15th April-26th June 1968 Help before 70 seconds for the ill, black victim was victim early it elicits further help from regularly use the subway to
There were four groups of confederates consisting of four students from Columbia recorded in 7/8 trials bystanders commute, giving high ecological
University, two male and two female Help before 70 seconds for the drunk, white victim was Helping behaviour is not affected by validity
The victims in all conditions were dressed identically recorded in 5/24 trials proximity to the victim (in this study) Participants were unaware of being
● The ‘drunk’ victim carried a bottle carried a bottle of alcohol, and Help before 70 seconds for the drunk, black victim was Larger groups of people are more likely to observed and therefore acted
smelled of alcohol recorded in 4/14 trials offer assistance (no bystander effect) naturally, reducing the risk of
● The ‘ill’ victim carried a cane In 62 out of 65 (95%) of trials where the victim had a cane, The longer the emergency continues, the demand characteristics
After 70 seconds into the train journey, the victim staggered forward, collapsed and passengers helped the victim before the model planned to more likely people are to leave the area
remained on the floor until they received help and the less impact intervention from a
intervene Weaknesses:
If no help was received by the time the train stopped at the station, the model In 19 out of 38 trials (50%) of trials where the victim was model will have on others helping
There was no informed consent as
helped him to his feet and they left the train drunk, passengers helped the victim before the model Bystanders are more likely to offer
passengers were unaware they were
The two female confederates also left the train planned to intervene immediate help for an ill victim than a
observed, raising ethical
The group of four would then make their way to the other platform, get on the next Therefore in 81 out of 103 trials the victim was helped drunk victim
considerations
train heading in the opposite direction and repeat the trial before the model was scheduled to help ● The ‘costs’ (embarrassment,
The situation could have caused
Approximately five to eight trials were completed each day Of the 81 spontaneous trials (within 70 seconds) first aggression) are higher for
distress to participants as they would
● On each day, the victim’s condition (drunk, cane, black, white) helpers, 90% were male helping a drunk victim
have been witness to an upsetting
remained the same In 21 out of 103 trials, 34 people left the ‘critical area’ There is a tendency for same-race helpers
situation and so could have felt
A total of 103 trials took place ● This occurred more when the victim was to be more frequent, and also more likely
pressure to offer help or guilt for not
The two female confederates were seated away from the ‘critical area’ in an drunk than when they were ill to assist the drunk victim
helping the victim
adjacent area, and they covertly recorded the date from the trial (participants A content analysis of comments made by train passengers ● There was no race effect
were unaware they are being observed) highlighted that discomfort was felt if no one had eloped when helping the ill victim
● One made a record of the race, sex, number of, and location of the within 70 seconds with the cane
passengers in the ‘critical area’ and the race, sex, number and location ● Other comments attempted to seek are more likely to help a male victim than
of bystanders who came to help the victim confirmation that not helping was the right women, as the costs of not helping are
● The second confederate recorded the race, sex, number of and decision lower for women
location of the passengers in the ‘adjacent area’ and the time between There did not appear to be a strong correlation between
the model helping and the first bystander helping, as well as comments the number of bystanders and helping behaviour
made by passengers around them
Underground Phenomenon?
Procedure: Aims:
The ‘type’ of victim was portrayed as either drunk or ill To investigate the effect that the type of victim would have on the speed and frequency
The ‘race’ of the victim was either black or white with which a person would respond to aid them in an emergency (as well as the effect of
Sample: 4,450 men and women passengers were present over the duration of the the race of the victim)
scenes being staged Investigated the effect of role model behaviours in emergency situations and the
● Approximately 55% were white and 45% black relationship between the size of the group and helping a victim
● The mean number of bystanders in the train carriage was 43 people
● The mean number of bystanders deemed to be in the ‘critical area’
(closer proximity to victim) was 8.5 people Results: Conclusions: Strengths:
The staged scenes took place on a 7.5 minute continuous train journey between Help before 70 seconds for the ill, white victim was Spontaneous helping occurred within 70 Conducted in a naturalistic
59th Street and 125th Street stations in the New York subway between 11:00am recorded in 52/57 trials seconds, and where a model assists a environment on passengers who
and 3:00pm during the period 15th April-26th June 1968 Help before 70 seconds for the ill, black victim was victim early it elicits further help from regularly use the subway to
There were four groups of confederates consisting of four students from Columbia recorded in 7/8 trials bystanders commute, giving high ecological
University, two male and two female Help before 70 seconds for the drunk, white victim was Helping behaviour is not affected by validity
The victims in all conditions were dressed identically recorded in 5/24 trials proximity to the victim (in this study) Participants were unaware of being
● The ‘drunk’ victim carried a bottle carried a bottle of alcohol, and Help before 70 seconds for the drunk, black victim was Larger groups of people are more likely to observed and therefore acted
smelled of alcohol recorded in 4/14 trials offer assistance (no bystander effect) naturally, reducing the risk of
● The ‘ill’ victim carried a cane In 62 out of 65 (95%) of trials where the victim had a cane, The longer the emergency continues, the demand characteristics
After 70 seconds into the train journey, the victim staggered forward, collapsed and passengers helped the victim before the model planned to more likely people are to leave the area
remained on the floor until they received help and the less impact intervention from a
intervene Weaknesses:
If no help was received by the time the train stopped at the station, the model In 19 out of 38 trials (50%) of trials where the victim was model will have on others helping
There was no informed consent as
helped him to his feet and they left the train drunk, passengers helped the victim before the model Bystanders are more likely to offer
passengers were unaware they were
The two female confederates also left the train planned to intervene immediate help for an ill victim than a
observed, raising ethical
The group of four would then make their way to the other platform, get on the next Therefore in 81 out of 103 trials the victim was helped drunk victim
considerations
train heading in the opposite direction and repeat the trial before the model was scheduled to help ● The ‘costs’ (embarrassment,
The situation could have caused
Approximately five to eight trials were completed each day Of the 81 spontaneous trials (within 70 seconds) first aggression) are higher for
distress to participants as they would
● On each day, the victim’s condition (drunk, cane, black, white) helpers, 90% were male helping a drunk victim
have been witness to an upsetting
remained the same In 21 out of 103 trials, 34 people left the ‘critical area’ There is a tendency for same-race helpers
situation and so could have felt
A total of 103 trials took place ● This occurred more when the victim was to be more frequent, and also more likely
pressure to offer help or guilt for not
The two female confederates were seated away from the ‘critical area’ in an drunk than when they were ill to assist the drunk victim
helping the victim
adjacent area, and they covertly recorded the date from the trial (participants A content analysis of comments made by train passengers ● There was no race effect
were unaware they are being observed) highlighted that discomfort was felt if no one had eloped when helping the ill victim
● One made a record of the race, sex, number of, and location of the within 70 seconds with the cane
passengers in the ‘critical area’ and the race, sex, number and location ● Other comments attempted to seek are more likely to help a male victim than
of bystanders who came to help the victim confirmation that not helping was the right women, as the costs of not helping are
● The second confederate recorded the race, sex, number of and decision lower for women
location of the passengers in the ‘adjacent area’ and the time between There did not appear to be a strong correlation between
the model helping and the first bystander helping, as well as comments the number of bystanders and helping behaviour
made by passengers around them
