Fourteen Passers-by in the Square
CCTV from a town square shows a 78-year-old man falling heavily and lying on the pavement at 3:10 p.m. on a Saturday. Over the next nine minutes, 14 people walk past. Two stop and look, one takes a photograph, and nobody calls for an ambulance until a shopkeeper comes out at 3:19. In interviews afterwards most give similar reasons: there were lots of people, so someone else must have called; I was not sure whether he was drunk or just resting; I did not want to touch him in case I made it worse; and I assumed a doctor would be along. Nobody mentions being afraid of the man. The local council wants to run a one-off Be a Good Samaritan poster campaign showing a phone and the words Just call 112.
1. Explain the passers-by behaviour using at least three social-psychological concepts, showing why each concept fits this footage.
Read the worked answer
Diffusion of responsibility fits directly. With 14 people present, each person's share of the responsibility for acting is diluted. The witnesses say it themselves: someone else must have called. The more bystanders there are, the less each one feels personally accountable.
Pluralistic ignorance also fits. People read an ambiguous situation from other people's behaviour. Nobody was rushing forward, so each passer-by interpreted the scene as not an emergency. The comment about the man possibly being drunk or resting is exactly this: the inaction of others confirmed a non-emergency reading.
Evaluation apprehension explains the reluctance to intervene in public. Not wanting to touch him in case it made things worse is a plausible reason, but it also protects the person from looking foolish or being blamed. People avoid acting when they might be judged.
The assumption that a doctor would come adds assumed expertise: someone more qualified will deal with it.
Note that nobody mentions fear of the man, so the danger of intervening does not explain the inaction here. The barriers are psychological, not physical, and the number of bystanders is likely to have lowered the probability of help rather than raised it.
2. Evaluate the proposed poster campaign. What would research suggest is more likely to work?
Read the worked answer
The poster is a single-factor intervention: it assumes the problem is that people do not know what to do. The interviews suggest people did know about emergency numbers; the problem was the social psychology of the crowd. A poster saying just call 112 does not address diffusion of responsibility, pluralistic ignorance or evaluation apprehension, and it may even reduce helping if people assume the campaign will handle it. It also has no measurement and no follow-up.
More promising components would include: making responsibility personal and specific, such as you, the person nearest, call now; training that rehearses the behaviour rather than merely informing people; removing ambiguity with a clear rule, for example if someone is on the ground and not moving, call and stay with them; addressing evaluation apprehension by normalising helping and showing that most people would help; and using multiple channels over time rather than a one-off poster.
The council should also plan to evaluate the campaign rather than assume it works. A before-and-after measure of helping in comparable locations would be inexpensive and would tell them whether the money was well spent.
3. Design a field experiment to test whether the campaign works, and explain the main ethical safeguards you would build in.
Read the worked answer
Select 12 comparable locations or 12 comparable time slots in the same square and randomly allocate half to the poster condition and half to a control poster on an unrelated topic. Stage a standardised, low-risk incident, for example a confederate dropping a bag of shopping and stooping to gather it, or dropping a clearly labelled envelope and walking away. Record from CCTV whether anyone helps, the delay before the first helper, and how many people pass by. Keep the confederate, clothing, time of day and weather as constant as possible, and run a fixed number of trials. The outcome measures are the proportion of passers-by who help and the time to first help.
Ethical safeguards: the victim must be an actor who is never at risk; the incident must be brief and harmless; the council must give permission; the public should be debriefed afterwards through signage or a leaflet for those who stopped; and no identifiable footage should be kept longer than the coding requires, then destroyed.
A staged fall by a real elderly person would be unethical, so a lower-risk incident must be used even though it is less similar to the real event. That trade-off between realism and harm should be stated explicitly in the report.