If you were a participant

What you care about

Imagine being four years old and one of the 72 children. You care about simple things: being safe, knowing what is going on, being allowed to say no, not getting into trouble, and having a nice time. You cannot read a consent form. You trust the grown-ups.

What worries you about the original study

• You were tricked. A kind lady gave you wonderful toys and then said they were for "the other children" because she "did not let just anyone play with them". That was not true. It was a planned step to make you a little upset.
• You did not know you were being watched. People behind the mirror were writing down what you did every five seconds.
• You saw a grown-up do something frightening or exciting. A big adult shouted and hit a clown doll that was taller than you. Some children may have found this confusing or scary. Nobody asked.
• Nobody told you what happened afterwards. The paper reports no explanation for the children, no chance to play with the special toys again, and no check on how you felt.
• You could not really say no. The paper explains that the experimenter stayed in the test room because otherwise a number of the children "would either refuse to remain alone or would leave before the termination of the session" (p. 577). Her presence kept you in a room you might have chosen to leave.

Your redesign

1. Ask the child as well as the parent. Explain in simple words and pictures: "You will watch a grown-up play, then you can play with some toys. You can stop any time." Get the child's assent and respect a "no" at any point, including during the session.
2. Give the child a way to stop. A stop card or a bell that brings a familiar teacher in.
3. Remove the trick with the toys. Drop the deliberate frustration, or replace it with something honest and mild.
4. Make the model less intense. Shorten the aggressive sequence and remove shouting.
5. Let a familiar adult be nearby. A nursery teacher the child knows could sit in the room instead of a stranger.
6. End kindly. After the test, let every child play with the special toys. Have the model come back and say, "That was pretend. We don't hit people or toys to hurt them." Check that the child is happy before returning to class.
7. Ask the child what they thought. "What was the grown-up doing? Why did you play the way you did?" This respects the child as someone with a view, and adds data the original never collected.

Why

These changes follow from treating children as people with rights rather than as sources of data. Today's research ethics codes, such as the British Psychological Society's Code of Human Research Ethics, require valid consent, minimal deception, the right to withdraw, debriefing, and avoiding harm. With children, they also require the child's own agreement in a form the child understands.

The redesign also improves the science in one way. Asking children what they thought helps separate playful copying from aggression, which was the study's weakest point.

What your redesign gives up

• Less arousal. Without frustration, children may be less aggressive overall. This makes effects smaller and harder to compare with 1961, although Bandura's 1965 study found copying without a frustration step.
• A familiar adult changes behaviour. A teacher in the room may make children more careful, so they copy less in front of her. This could hide learning that has taken place.
• The ending may contaminate later measures. If the model says "we don't hit", children's later behaviour, if measured, reflects both the aggressive model and the correction. Researchers would need to measure first, then correct.