Pilliavin

• Background

  Bystander behaviour and diffusion of responsibility Bystanders in an emergency situation often do not offer help. This is described as bystander behaviour. in a large group each person feels less responsibility and this may explain bystander behaviour,
• Aims

  To investigate the effects of four independent variables on helping behaviour in a real life setting.
• Method (Design) 


  Field experiment, independent measures design, using observational techniques. It is also a snapshot study. IVs: Type of victim (drunk or cane), ethnicity* (black or white), gender of helpers (men or women), presence of model, group size. DVs: Time it took for help to be offered and number of people who offered help.
• Method (Sample)

  Opportunity sample of 4,450 passengers on the New York subway, weekdays 11 am to 3 pm from 15 April to 26 June 1968. 60% of passengers were men.
• Method (Materials/apparatus)

  Each trial involved the 'victim' carrying either a black cane or a liquor bottle wrapped in brown paper, standing next to a pole in the middle of the carriage.
• Procedure

  1. 103 trials of one ride on the New York subway lasting 7½ minutes

  2. Four students boarded using different doors

  3. Victim stood next to the pole in the centre of the critical area

  4. After 70 seconds, the victim collapsed and remained on the floor until he received help or until the train stopped and the model helped him up

  5. On 38 trials the victim smelled of liquor and carried a liquor bottle (drunk condition)

  6. On 65 trials, the victim appeared sober and carried a black cane (cane condition)

  7. The model was in the critical or the adjacent area, and either helped early (70 seconds after initial collapse) or late (150 seconds later)

  8. The observers sat outside the critical area

  9. Observer 1 noted (a) ethnicity, gender and location (EGL) of people in the critical area, (b) number of people in the car, (c) number of people who helped, (d) EGL of people who helped

  10. Observer 2 noted (a) EGL of people in the adjacent area, (b) how long it took for the first person to help after the victim collapsed and/or after the model appeared

  11. Both observers recorded comments spontaneously made by nearby passengers
• Results

  Type of victim (drunk or cane) 1. A person using a cane is more likely to receive help than one who appears drunk. 2. Help is more quickly forthcoming for a person with a cane than a 'drunk' victim. 3. The median latency for cane trials was 5 seconds.
• Results

  Ethnicity and gender of helpers 4. The black victim received help less quickly than white victims. 5. A slight 'same-ethnicity effect' was found in the drunk condition. 6. 90% of first helpers were men, whereas only 60% of passengers were men.
• Results

  Presence of model 7. The model intervening early (after 70 seconds) had slightly more effect than the late model (150 seconds).
• Results

  Group size 8. 'Diffusion of responsibility' was not found in this study. Helping was greater in seven-person groups than three person groups.
• Results

  Comments from passengers 9. For example, 'It's for men to help him' or You feel so bad when you don't know what to do'.
• Conclusions

  A cost-reward model can predict when help will be forthcoming in an emergency situation where escape is not possible: • The emergency situation creates heightened arousal. The decision to help is motivated by a selfish desire to rid oneself of this unpleasant emotional state. • Action will depend on whether the rewards of helping are greater than the costs of not helping.