Pivilian et al

• Psych Investigated

  bystander apathy: showing no concern

  murder of Kitty Genovese: 38 individuals around, no help

  Darly and Latane(1968): perceived obligation help reduced when others are present (diffusion of responsibility hypothesis)

  -more likely to assist w witness modelling helping behavior

  -help people similar to ourselves/not responsible for circumstances
• Background

  investigate whether people demonstrate bystander apathy toward a person in need of help in a realistic setting
• Aims

  study bystander apathy and diffusion of responsibility in a natural setting

  investigate 4 variables on helping behavior/being a "good Samaritan"

  -type of victim

  -face of victim

  -behavior of a model

  -size/group of bystanders
• Research Method

  field experiment, NYC subway

  design: independent groups

  -repeated on different days, different participant involved each condition
• Independent Variables

  4IV: outlined aims, operationalized

  -type of victim: levels of 'drunk' or 'ill victim

  -face of victim: black/white

  -behavior of model: levels close/distant from victim helped either early/late

  -size of group: naturally occuring # passengers in subway carriage
• Dependent Variable

  level of bystander helping (quantity operationalized)

  -time taken, number of passengers who helped

  qualitative data operationalized: verbal remarks made by passengers
• Procedure

  4 teams researchers, standardized

  2m, 2f boarded dif doors

  -fem: sat adjacent from 'critical area'+ observed passengers, recorded data during each trial

  -males: victim (stood pole center critical area) model (rem standing throughout)

  7.5min gap between 2 stations, 70 sec in: victim staggered toward, collapsed

  -rem lying on floor upwards

  -no help= victim helped at next stop
• Victim

  played by different males made to look similar

  26-35 years 3 white/ 1 black

  -identical casual clothing, jacket, trousers, no tie

  -38/103 victim smelled like alcohol, carried bottle wrapped in brown bag

  -65/103 sober carried black cane; behaved identically
• Model

  all white males 24-29yrs; dressed informally

  -model raised victim to sitting position, stayed with until next stop
• Split in conditions

  critical/early: model stood in critical area, waited 70 seconds to help victim

  critical/late: waited 150 seconds to help

  adjacent/early: model stood adjacent area, waited 70 seconds to help victim

  adjacent/late: waited 150 seconds
• Results

  78% of victims recorded spontaneous help (before model interviened/no model cond)

  -60% cases more than 1 person helped
• Type of Victim Difference in levels helping*

  more likely to help victim w cane than drunk, occurring earlier

  cane: received help 62/65 trials

  drunk: 19/38 trials

  
  

  white victim: no model cane/drunk: 100%; model: cane 100% drunk 77%

  black victim: no model: cane 100% drunk 73%; model drunk 67%
• Race

  b/w equally likely receive help (minor ev of same-race helping in drunk cond, noticed help of own race)

  -drunk cond: black victims received less help overall (73/67)

  suggests people are more likely to help those similar to themselves
• Variable of sex

  majority of helpers (90%) male

  -women comments: 'its for men to help him' 'im not strong enough'

  -suggest difference between men and women in terms bystander helping behavior (or result victim being played by male)
• cost-benefit model

  effect of modelling: difficult to analyze, helping was spontaneous

  -early model interventuon 70sec slightly more likely result in helping beh

  
  

  NO evidence support diffusion of responsibility hypothesis: 7 person groups faster to respond than predicted (than 3 person groups)

  
  

  leads to cost-benefit model: witnessing an emergency raises individuals levels of arousal (feelings disgust, sympathy/courage)

  -prompted to respond, reduce any difficult feelings

  -less help occurred due to disgust, arousal participants prompted comment/move away
• Comments

  higher w/o help, especially w drunk victim

  -20% trials passengers moved away
• Conclusion

  natural setting offer spontaneous help even in group situation

  -no evidence diffusion of responsibility;

  factors influencing decision to help:

  -type of victim: cane > drunk

  -gender of helper: men > women

  similar victim helping own race, esp if drunk

  -duration of emerg: longer emerg continues, less likely anyone help, more likely find other way coping w arousal
• Strengths

  field exp, indep measures: good ecological validity

  -unaware taking part, behave naturally

  -limit/less extraneous variables

  quantitative data: objective record (# help/how long took)

  -more reliable w 2 observers (+qualitative data w remarks/movements)

  -understand par thoughts/beh associated w help in more depth
• Weakness

  methodological issue: not sure par took part only once

  -same route, exposed more than one cond = risk of dem characteristic

  unrepresentative: helping may vary in other countries

  -4500 individuals w mixed ethnics/gender could increase generalizability

  less control on extraneous variables: (possible subway delays/weather) lower validity/reliability
• Ethics

  no consent w research, not debriefed after study

  -deceived: believed victim genuinely collapsed and needed help

  -may have suffered psych distress as a result/guilt for not helping/concern for well-being
• Sample

  NYC subway between Harlem and Bronx weekdays 11am-3pm

  -considered as opportunity sample

  -est 4450 people 45% black 55% white

  -mean number passengers per carriage: 43

  -mean number people in critical area: 8.5