If you were a clinician

What you care about

A clinician here means a child psychologist, psychiatrist or counsellor who works with individual children and families. You see children who hit, bite or bully, and children who have grown up around violence at home. You care about two things: which children are most at risk, and what helps them change.

What worries you about the original study

• It averages away the children you treat. The study reports group means. Your clients are the children at the extremes, the ones who copied the most, or who were already very aggressive.
• It ignores children's histories. A child who has seen a parent being beaten may react to an aggressive adult very differently from a child who has not.
• It shows how harm is learned, not how to undo it. You need the second half of the story.

Your redesign

1. Turn the study around. Use modelling to teach a helpful behaviour. For example, show children who struggle with anger a short video of a child their age who gets frustrated, says what they feel, and asks an adult for help.
2. Work with the children who need it. Include children referred for aggression, with careful consent and safeguards, alongside a comparison group.
3. Measure each child's background. Record exposure to violence at home, temperament, language skills and any developmental difficulties, so you can see for whom modelling works best.
4. Measure change over months, at home and at school, rather than in one session.
5. Combine modelling with practice and praise, as therapy normally does, and compare this with modelling alone.

Why

Bandura's own later work did exactly this. In one study, young children who were afraid of dogs watched another child approach and play with a dog, and afterwards they avoided dogs less (Bandura, Grusec & Menlove, 1967). Modelling is now part of many therapies, including participant modelling for phobias, in which the therapist demonstrates and then guides the client through each step, and video modelling in programmes for children with autism.

Measuring children's backgrounds matters because exposure to real violence is common. In India's National Family Health Survey 2019-21, women whose fathers had beaten their mothers were more than twice as likely to have experienced violence from their own husbands, 58% compared with 25% (IIPS & ICF, 2021, p. 645). A clinician wants to know how such histories change what children learn from the models around them.

What your redesign gives up

• It answers a different question. It tests whether modelling can help, not whether aggressive models harm. That is useful clinically but no longer a test of the original hypothesis.
• Combined treatments blur causes. If modelling, practice and praise are combined, it is hard to tell which part did the work. That is why the design compares modelling alone with the combination.
• Small and varied clinical groups. Children referred for aggression differ greatly from one another, so results are harder to generalise, and recruitment is slow.