Eval

• • High standardisation: victims wore same clothing, collapse at the same time when train passed 1st station --> replicable, reliable
  • Qualitative data: allowed further insights into bystanders' justifications
  • Quantitative data: time taken to receive help -> objective, comparable - High generalisability: large sample size (4450, 45% black 55% white)
  • High ecological validity: field experiment, real subway route, Ps unaware of being observed -> avoid demand characteristics"
• • Informed consent: naturalistic setting -> no permission to be apart of the study -> unaware, could cause psychological harm
  • Debrief: no debrief -> Ps may have had potential guilt for not offering help"
• • Lack of controls: naturalistic setting. some may have not noticed --> affect validity
  • Lack population validity: Ps predominantly consist of commuters on that specific New York subway route. America is an individualistic country, New York is busy -> findings may differ in a less busy, more collectivist culture"