If you were a policymaker

What you care about

A policymaker writes rules for millions of people: limits on screen content, school discipline rules, child-protection laws, parenting programmes. You care about whether a problem is real and serious across the population, whether a policy will reduce it, what the policy costs, and whether it restricts other rights, such as freedom of expression.

What worries you about the original study

• It cannot tell you about real violence. It shows that small children copy an adult hitting a toy, minutes later. Policy is about serious harm to people over years.
• It cannot tell you the size of the effect in the population. Laboratory effects with 72 children do not tell you how much violence in society would fall if a policy were introduced.
• It cannot tell you which policy works. It tests exposure to a model, not interventions.
• It has been stretched before. For decades, the Bobo studies were cited in arguments about television and video games, sometimes as if they showed more than they did.

Your redesign

Strictly, a policymaker would not redesign this single experiment. They would ask for a programme of evidence:

1. Long-term studies of real outcomes. Follow large, representative groups of children over many years, measuring exposure to violence at home, at school and in media, and later aggression and violence. Collect reliable records, not only self-reports.
2. Trials of interventions. Randomly assign schools or districts to policies, for example teacher training in non-violent discipline or a parenting programme, and measure children's aggression and wellbeing.
3. Studies in the population that the policy covers. For Indian policy, this means Indian children across states, classes and languages, not only Western samples.
4. Cost and side-effect analysis. What does each policy cost, and what does it restrict?

Why

Courts and policy bodies have repeatedly asked whether laboratory findings justify restrictions. In 2011 the US Supreme Court struck down a California law that banned selling violent video games to children. The Court's opinion called the state's research evidence "not compelling", saying the studies "do not prove that violent video games cause minors to act aggressively". Much of that research came from the social learning tradition that the Bobo studies began. Whatever your view of that decision, it shows the gap between laboratory effects and the evidence a policy needs.

India has made policy on related ground. The Right of Children to Free and Compulsory Education Act 2009, Section 17, bans physical punishment and mental harassment in schools. The National Commission for Protection of Child Rights has explained such bans partly with social learning reasoning, that adults who hit teach children that hitting is acceptable. Yet its supporting evidence came mainly from Western studies. A policymaker would want Indian evidence too, both to justify the policy and to design training that works in Indian schools.

What your redesign gives up

• Speed. Long-term studies and trials take years. Policymakers often must act before such evidence exists.
• Clean causes. Large field studies measure many influences at once, so they rarely isolate modelling as precisely as the Bobo experiment did.
• Simplicity. Real-world evidence usually shows small, mixed effects that make clear rules harder to write.