Piliavin et al, 1969 Flashcards
Bystander
anyone who is present at an incident but not directly involved
Bystander effect
behavior of bystanders who don’t assist those who need help in an emergency
Diffusion of responsibility
When groups of people witness an emergency together, but they only assume a fraction of responsibility for helping
Background of the study
- famous murder in New York, 1964
- woman was stalked by a man
- stabbed 3 separate times
- 38 ppl saw, no one helped
What kind of study is this?
Field experiment
What is this study investigating?
to investigate if helping behavior was affected by:
- type of victim
- race of victim
- modelling effect
- size of group
Sample
- 4450 males and females
- commuters travelling in a tube
- racial mix: 45% black, 55% white
- opportunity sample
- 8 ppl in critical area
- 43 ppl in each carriage
- journey: 7 1/2 mins (unsolicited participants)
- time: weekdays 11am-3pm
Field situation
- non-stop ride from 59th 125th street
- no one was allowed to leave so that there was no change when an emergency takes place
Procedure
- 4 Columbia General Studies students
- 2 males, 2 fems - female confederates took seats outside critical are
- male model + victim standing - data collected 103 trials
- victim stood next to pole in critical area
- victim staggered forward, collapsed 70 sec after train
- until receiving help, victim remained on floor
- if victim got no help by the time train stopped, model helped him on feet
- at stop, team left and waited separately until other riders left the station
- proceeds to another platform to board train going in opposite direction for next trial
- 6-8 trials in given day
- all trials in same “victim condition”
- 59th - 125th station
Victims
- 4 victims = 3 white, 1 black
- males
- age: 26-35
- all identically dressed in Ein Sower jackets, old slacks, no tie
-38 trial victims - smelled of liquor and carried liquor bottle wrapped in BPB
- 65 trials, sober & carried black cane
- standardized procedure, victims dressed + behaved identically on 2 conditions
Models
- 4 white males
- ages: 24-29
- no identical clothing
- wore formal attire
4 model conditions
1. critical area (early)
- model stood in critical area
- waited until 4th passing station to assist victim
- 70 secs after collapse
- Critical area (late)
- model stood in critical area
- waited until passing 6th station to assist victim
- 150 secs after collapse
- Adjacent area (early)
- model stood in middle of car in area adjacent to critical area
- waited until passing 4th station
- Adjacent area (late)
- waited until passing 6th station
- 150 secs
Controls in study
- dressing the same
- behaving the same
- smelling of alcohol
- having black cane in 65 trials
Observers
Observer 1:
- noted race, sex, location of every rider seated/standing in critical area
- counted total no. of individuals in car
- counted total no. of individuals who came to victim’s assistance
- recorded race, sex, location of every helper
Observer 2:
- coded race, sex, location of all people in adjacent area
- recorded latency of first helper’s arrival after victim had fallen and on appropriate trials
- recorded latency of first helper’s arrival after programmed model arrived
Both observers:
- recorded comments spontaneously by nearby passengers
- attempt to elicit comments from rider sitting next to them
(qualitative data)
4 IVs
- race of victim
- model conditions
- drunk/cane condition
- size of group
2 DVs
- latency in observers (speed of helping)
- no. of people who helped
First helpers
- 60% of 81 trials, victim received help (received from 2 ppl)
- no significant differences between black/white victims or between cane/drunk victim
- 60% males in critical area
- 81% (first helpers), 90% were males
- men more likely to help than women because victim himself is male and he was drunk
- 81% first helpers, 64% were white
- this % doesn’t differ significantly from expected % of 55% based on racial distribution in cars
- “same-race effect” helping only in case of drunk victim may reflect those empathy/sympathy towards victims and one’s own racial group
Quantitative data
- Black = white:
–> 16 trials spontaneous help offered to black victim, half of first helpers were white (50% - 50%) - Male helpers > female helpers = 90% > 10%
- Slight same race effect for drunk condition:
–> same as victim = 16
–> different from victim = 3
Evaluation of qualitative data
STRENGTHS:
- descriptive
- holistic
- in-depth
WEAKNESSES:
- low generalizability
- subjective
- can’t make comparisons between 2 people
Evaluation of quantitative data
STRENGTHS:
- reliability
- comparative
- numerical
WEAKNESSES:
- reductionist
- not evaluating ethics
Other observations
- some passengers moved away from critical area on 21 of 103 trials, total of 34 ppl left in critical area
- ppl left critical area on higher proportion of trials that were drunk than cane
- far more likely to leave on trial which help wasn’t offered by 70 secs, compared to trials on which help was received before that time
- more comments made in drunk trials than cane trials
- more comments made when no passengers spontaneously helped
No. of by-standers
- no. of evidence of diffusion of responsibility
- mild effect in opposite direction
Conclusion
- ill person is more likely to receive help than drunk person
- men more likely to help another man than a woman
- people help their own ethnic group slightly more when drunk
- no strong relationship between size or group and likelihood of helping
- no support for diffusion of responsibility
- longer the incident goes on, less likely people help
- ppl may leave area and more likely to discuss with others
Cost reward model
- Emergency
–> empathy
–> proximity
–> duration of emergency - Heightened arousal
- Arousal reduction
–> help directly
–> get help
–> leave the scene
–> decide the victim is underserving
Ethical issues
- no debriefing
- no right to withdraw (pps couldn’t leave the situation)
- ethical informed consent
Improvements to be made
- inter-rater reliability
(make both observers record same situation)
Cost for helping and not helping
Helping - COST:
- effort
- physical harm
- embarrassment
- distasteful experience
Not helping - COST:
- disappointment
- blame
- guilt
Reward for helping and not helping
Helping - REWARD:
- praise from self and others
Not helping - REWARD:
- continuation of other activities
Application (AO2) -
> drunk
> women
> same race
> diffusion of responsibility
> longer emergency situation
- The drunk is helped less because the costs…
- smells bad
- distasteful experience
- potentially harmful
- have to use physical strength to pick person up
- can get thrown up on - Women help less because the costs…
- may not have physical strength
- dealing with a male
- more likely to get hurt/harmed - Same race helping for drunk condition takes place because…
- helpers can feel more empathy with another person of the same race
- can put themselves in the victim’s shoes - Diffusion of responsibility is not found because…
- people in larger groups were more likely to help faster in shorter amount of time
- people in smaller groups were less likely to help (took longer time) - Less helping takes place when longer the emergency situation takes place because…
- other people see others helping more
- people who see others stepping in more have more courage to help victim
Evaluation - methodological
STRENGTHS:
- high reliability (standardized procedure)
- opportunity sampling
- high eco validity
- less time consuming
- high generalizability
- high mundane realism
- high validity
WEAKNESSES:
- way too many variables!
(divides trials in to too much of small data/limited data, makes it difficult to find which variable applies to which set of data)
- low generalizability (this exp only takes place in New York, certain times, not inclusive of other commutes, only weekdays, not weekends)
- field experiment (not going to be able to control all extraneous variables)
Evaluation - ethics
STRENGTHS:
- confidentiality
- privacy?? (debatable)
WEAKNESSES:
- no debriefing
- no informed consent
- deception
- no protection from psychological harm
- no right to withdraw
Individual VS situational debate
- same race effect of drunk trials (situational)
- 16/19 times (same race as victim)
- 3/19 times (different from victim)
- timing of helping (situational)
- help varied from ppl depending on time of situation
- women who didn’t end up helping man/victim (situational)
–> “It’s for men to help him”
–> “I wish I could help him - I’m not strong enough”
–> “I never saw this kind of thing before - I don’t know where to look”
Interpersonal Reactivity Index (IRI)
- 28 items
- 5 points
- likert scale
- 4 subscales (each made up of 7 different items)
(PT) = perspective taking
(FS) = fantasy scale
(EC) = empathetic concern
(PD) = personal distress
What is perspective taking?
It’s being able to understand someone else from their own perspective
What is fantasy scale?
It’s a person’s tendency to fantasize and put their feelings and actions into fictitious imaginations like characters in books/movies/plays
What is empathetic concern?
It’s having feelings of empathy and sympathy for others in ways of concern and feelings of unfortunate situations for others
What is personal distress?
It measures “self-oriented” feelings of personal anxiety/unease in tense settings/situations
Psychometric tests
- standard and scientific method
- measures a person’s capabilities (mental, behavioral, personality, intelligence)
- E.g. IQ test, EQ test, AQ test
Strengths of psychometric tests
- simple
- subjective
- accessible
- standardizes (ensures high reliability)
- easy comparison (quantitative data)
Weaknesses of psychometric tests
- reductionist (limited)
- not valid??
- cultural bias
- social desirability bias
