A man is trying to unload shopping from a car. He is having difficulty because one of his hands is bandaged. He asks a young girl, aged about nine or 10, who is walking along the pavement, to help him carry his purchases into the house. What does she do?

A child is playing in the kitchen, and overhears his mother telling a friend that she has lost her job. She is not going to tell his father yet, and when she realises the child has overheard asks him to keep it a secret. Later the child is sitting with his father who is looking through holiday brochures and planning a holiday which relies on the mother’s income. What does the child do?

If three studies undertaken by a team of psychologists from three Glasgow universities are right, most children aged between six and l0 would comply in both these situations. The young girl would go with the man into the house, and the boy would keep the secret, however bad he felt about it.

Extrapolating these findings to situations in which children are in danger of harm including sexual abuse, the team concluded that, while children were very quick to learn the rules of “stranger danger”, they often got it all wrong in practice, when presented with real situations.

Moreover, Gillian Mayes, one of the psychologists, says: “Many programmes on personal safety work on teaching children that if they feel bad about a secret they should tell. Our study suggests that even when a child feels bad about a secret they will still be more likely to keep it than to tell.”

Of the three studies, the two more recent ones focussed on “stranger danger” (being lured by someone into harm) and “secrets”, an important feature of sexually abusive situations. They followed an evaluation in 1990 of the Kidscape Safety Training Programme which aimed to provide ways to help children deal with potentially dangerous situations, including sexual entrapment. This found that children performed badly on the secrets and “intimacy” (inappropriate touch from adults) themes. None of the schools involved in the later studies had recently implemented safety training programmes.

Scenarios were presented on video or in storyboard form to children in several Glasgow schools and their reactions to them assessed by questionnaires and interviews.

Children were often shown in situations in which they are called upon to be helpful to or to obey adults. Asked to identify with these children, the Glasgow pupils seemed to lose touch with their safety instinct. In one situation, where a child is caught by a man in his garden picking flowers and he orders her to bring them into the house, most of the children in the study indicated that they would go with him.

Apart from confusion and the desire to do as they were told, an important factor, suggested John Gillies, one of the psychologists, was that many children appeared to think a stranger would look odd or “bad” or sinister, whereas the stranger in this scene looked normal. “Ingratiation is a major art of paedophile strategy, they work at looking and behaving in a normal way”, he points out.

The problem of how to simulate situations in order to test children’s reactions occupied the researchers a great deal. They rejected as unethical the idea of setting up “real-life” situations, as has been done in some American studies.

None of the video or cartoon stories had any explicit sexual content but the researchers argued that the events depicted were analogous and the children’s reactions to them would reflect how they would react in more explicitly sexual situations. The scenarios included “safe” ones: a mother asks her daughter not to tell their father about his surprise party, for example.

The team concluded that just teaching children rules for dealing with someone intent on harming them - saying “no”, running away if possible, telling an adult, for example - won’t necessarily help. A lot of situations require a more sophisticated assessment of risk. “Children should be provided with multiple examples of all kinds of situations and secrets and encouraged to discuss and practice possible ways of dealing with them”, says Gillian Mayes.

“You can’t underestimate the conflicts that these children feel. We must discuss conflict - and try and get across that you don’t always have to be polite, nice, and helpful. They have to be able to express their concerns”.

Michelle Elliott of Kidscape, a charity concerned with child safety and bullying, agreed that helping children to understand what were dangerous situations and secrets was complicated. “In the UK we raise our children to be polite, and abusers often use fear and respect for authority to subdue children”. The only way to make sure that the safety message “gets into their bones” was to ensure teachers integrated it into the whole curriculum.

Children had to practice many different “what if...?” situations to learn what to do. A useful strategy, which didn’t involve talking at all, was to pretend they hadn’t heard and walk away, she said.

Kidscape has stopped defining secrets as good or bad, and now calls them “safe” or “unsafe”. “We’re not trying to teach them to mistrust, but to know if the situation is safe or unsafe”, she said.

An Evaluation of a Children’s Safety Training Programme (1990). Ellen Moran, John Gillies, Gillian Mayes and Lindsey MacLeod from Departments of Psychology at Glasgow University and the Universities of Strathclyde and Glasgow Caledonian. Funded by the Scottish Office Education Department, Home and Health Department and Health Educational Group and the Economic and Social Research Council. Stranger Danger: What do Children Know? (1993) same authors. Funded by the ESRC. Children’s Conception of Secrets (1995). Same authors without Lindsey MacLeod. Funded by the ESRC.