Best of the Week
The fascinating science of child witnesses
Eyewitness accounts are notoriously fallible, and that's without the extra challenges posed by interviewing children. A network of scientists is trying to ensure that abused kids can bring abusers to justice - and let falsely-accused adults go free, reports Eloise Gibson.
The girl is six. She kicks her dangling legs against the low table and fidgets, avoiding eye contact with the blonde woman sitting across from her.
“Okay, Frieda,” says the woman, “tell me about…”.
That got her attention. She interrupts. “Hey! That’s not my name!” she says.
‘Frieda’ just followed a rule she was taught by her interviewer, Deirdre Brown: If Brown gets something wrong, the child is supposed to correct her. Brown’s second rule is, if she doesn’t know an answer, she shouldn’t guess or ‘have a go’.
“What’s in my pocket?” Brown asks her, testing. “Don’t know,” the young girl mumbles.
‘Frieda’ is really Brown’s daughter, Freya, and the topic of conversation is a recent family holiday to Disneyland. This interview is being staged so Newsroom can film it, but the rules that Brown just taught Freya are part of a real study.
Brown studies childhood memory, including how to help victims of child abuse give the most accurate testimony. Equally important is helping children reveal when there’s been no abuse, she says. “We want children to be protected, but we also don’t want lives ruined because of wrongful convictions.”
Brown’s field of study grew out of Peter Ellis’ conviction in the 1990s for abusing children at Christchurch Civic Creche. In that case and a couple of similar cases in the United States, kids were asked closed, leading questions that threw doubt on their testimony. Researchers wanted police to do better, and the field of child interviewing was born.
Brown helps train the specialist child interviewers working for the police and Oranga Tamariki. Especially with sexual abuse, children are often the only witnesses apart from the alleged perpetrator. Yet children are suggestible, still learning language, and not always tellers of logical stories. They can give highly reliable accounts, says Brown, but they need to be questioned correctly.
Her job is to find the tactics that elicit the truest responses. For example, she’s about to embark on a study of how best to explain ground-rules to children at the start of an interview. Using a grant from the Marsden Fund, she’ll see how children of different ages respond to being told the ground-rules that she just explained to Freya. Brown wants kids to know it’s okay saying, ‘I don’t know’ or ‘That’s not right’, even though it often goes against what they’ve been taught at school. If a child guesses an answer, or lets an interviewer’s mistake slide, it can mess up their evidence.
Here in Brown’s toy-lined interview room at Victoria University in Wellington, Freya has squirmed down in her chair until she’s almost horizontal. She’s getting confused recounting her story about a visit to Disneyland. Pop psychology might say she’s lying, but really she’s just feeling awkward being questioned by her mum in front of a video camera and a journalist.
In fact, there are no physical cues that reliably tell us if someone’s lying, especially when it comes to children, says Brown. Complicating matters is that children – and adults – can come to believe untrue versions of events.
“Memory is fragile,” she says. “And usually that fragility serves us well because it allows us to develop our understanding and be flexible. It’s really only a problem if you need to give an exact account of an event, which isn’t very often.”
In child abuse cases, it’s crucial.
Telling truth from fiction
Brown has a vivid childhood memory that her grandfather died when a tree fell on him. It’s not true, but that doesn’t make it feel any less real in her mind.
Strangely, our brains don’t seem to distinguish between real memories and false ones, which we’ve unknowingly built using outside information, she says.
Hollywood would have us believe that traumatic memories stay clearly preserved forever, like a video. But Brown says they break down, just as boring or pleasant ones do.
Every time we revisit a memory, it is liable to be altered by conversations we are having at the time or other gap-fillers. “We know that everybody’s memory decays over time, and we tend to plug in those gaps with assumptions from our general knowledge, conversations with other people, stories, TV, or things we saw in the newspaper,” says Brown. “If there are conversations [about an event before a forensic interview], each of those conversations is a chance to re-remember the event and for it to become slightly altered.”
People of any age can give false accounts without knowingly lying, Brown says. “Children may have come to develop a story or a memory that feels real and sounds convincing and that’s full of lots of little detail. They may have come to believe quite strongly in the account, just as adults can do.”
The risk of memory contamination is why, ideally, parents and caregivers are told not to discuss suspected abuse with a child again until after a formal interview. But parents get worried. They want to prepare the child, even if they don’t mean to coach them.
This is one of the reasons why Brown says trained child interviewers should get involved as early as possible.
“The best-case scenario would be that something happened to a child, they immediately told someone, that person said ‘we need to go and tell a police officer’, the police immediately took their statement and a specialist child interviewer is arranged within the next couple of days, and during the next couple of days the parents understand that they shouldn’t revisit the issue with the child,” says Brown.
“I suspect that’s unlikely.”
Last year the Sunday Star Times reported that some suspected child abuse cases were waiting up to a year to be investigated.
Courts now allow children to pre-record their evidence, so they can have therapy before the case comes to trial, without contaminating the memory. But the child will still need to be cross-examined much later, when the case comes to court.
Parents might think they can speak to a child themselves and get the full account.
Research suggests that’s wishful thinking, at least for many of us.
For a start, the way that parents tend to question their children – especially when anxious – has little in common with the techniques Brown advocates.
If you’re like me, the last time you really wanted to know something, you probably used a stern voice and asked closed questions like ones starting with “did you…”, “did he…”
What works better, according to research, is acting encouraging, asking open questions (‘tell me anything you can about…’), following the child’s train of thought and leaving the when/what/who questions until last. Research advocates not (repeat, not) jumping in after every answer with the next burning question on your mind.
At least most parents get to keep their dubious questioning tactics private. Pity the poor Scandinavian mum whose interviewing of her child ended up being studied by researchers.
This mother had the foresight to pull out her cellphone when her child made a troubling statement, so she could record the rest of the conversation. She took the tape to the police, where she presented it as evidence of child abuse. Later, when researchers got permission to listen to it, they found the conversation hadn’t happened as she remembered.
Statements the mother remembered the child saying had come from the mother’s mouth, and things she thought the child had disclosed spontaneously turned out to have been prompted by her highly leading questions.
Finnish researchers tested more parents, to see if this was common. It was. Parents were good at recalling the gist of a conversation, but they didn’t remember if the adult or child had made most of the allegations. Undeterred, the parents in the study were still likely to think that a fictional uncle had probably abused the child.
The main concern in the Ellis case was that investigators and parents had asked closed, leading questions, some of them containing information that the children hadn’t disclosed themselves.
The techniques wouldn’t fly under today’s evidence-based guidelines. “In part, it was because we didn’t know as much back then as we know now,” Brown says. “Lots of other techniques were embedded to encourage the children to share, all very much based on the premise that something had happened and they needed to get the children to describe it rather than coming from a neutral perspective.”
Even today, she says, problems can arise when an interviewer decides part-way through an interview that something bad has probably happened, then begins to focus on that. In their efforts to elicit more detail, they might miss other things the child is saying that could provide an alternative explanation.
At its worst, Brown says, botched questioning can create false memories of being abused. “Thinking back to the Peter Ellis case I often think that for so many of those children, who will now be adults, they’ve lived with the memory of having been victims,” says Brown. “Whether or not the abuse happened isn’t for me to say, but they’ve lived that.”
To try to ensure that doesn’t happen, a network of New Zealand researchers work with researchers from the United States and elsewhere to constantly improve the guidelines.
They run experiments on children of different ages to see how well they remember events and to test what kinds of questions glean the most accurate answers.
Researchers learn from footage of real police interviews. And they test kids’ recall by staging events in schools, where they film exactly what happened, so it’s easy to compare kids’ answers.
Brown recently set up a pretend ‘nurse’s’ visit, where children were touched for a health check. Brown was curious to know whether children would ‘remember’ something that didn’t happen, but that some of them might have expected -- being touched with a stethoscope. (A few did).
Another time, Brown tested children using human figure drawings, similar to the ones police interviewers use. When children were asked to mark on the figure where they’d been touched during a game of dress-up, some got it right, some left marks out, and others added false marks, including marks on their genitals. Conclusion: drawings probably shouldn’t be used except as a final resort to confirm a child’s verbal account.
Other studies are more encouraging. When Brown tested children with mild intellectual disabilities, she found them just as capable of giving accurate evidence as other kids, so long as their interviewers and the courts adjusted for their mental age.
The more researchers understand, the better they can help kids be believed when they should be.
Most parents, hopefully, will never have to send a child to a forensic interview. But some of the research in Brown’s field applies to everyday parenting.
Elaine Reese at Otago University has spent 25 years researching childhood memories and storytelling, and she says the way we speak to our kids from toddlerhood can help them develop more accurate memories.
Children start referencing the past “from around 18-19 months,” she says. “At first, they are talking about a biscuit they just ate or a toy they can’t find, but very quickly these references start going further back in time. Parents start having conversations with them about whole experiences, so talking about going to the museum and seeing the dinosaurs and that kind of thing. You may not be aware how often those conversations about the past are happening, but it’s a couple of times an hour,” she says.
Rees says not all parents carry out these talks in the same way. “Some parents are very elaborative in these conversations, asking more open-ended questions, like, ‘what part did you like’, and then the kid says something and they seize on it and say ‘Oh, you like the lion, what did you like about the lion?’.”
On the other hand, when a parent asks a closed question like ‘did you like that ice-cream?’, the child can answer with a yes or a no, says Rees. “But when you say, ‘what did you like about the ice-cream?’, you’re asking them to say more and there’s something about describing it in their own words that really helps their narrative skills and memory skills.”
“By the time they are 3 or 4 they are telling longer or more detailed stories about things that have happened to them,” she says. “What this is doing is helping the child build up the child’s representation of the past and put what happened into words.”
Rees says developing the memories and story-telling of toddlers can have spin-off benefits for children later, after they start primary school.
By the age of 5, it can help with their reading comprehension, managing their emotions, and telling stories about their past experiences during ‘mat time’.
Should it ever be important that a child can recall something accurately, the time spent talking to them as a toddler may just help them out.
At the very least, it might give them the confidence to speak up when an adult gets their name wrong. Just ask Frieda.
Can you help our journalists uncover the facts?
Newsroom is committed to giving our journalists the time they need to uncover, investigate, and fact-check tough stories. Reader donations are critical to buying our team the time they need to produce high-quality independent journalism.
If you can help us, please donate today.