The Christchurch Civic Creche Case

News Reports Index


Television New Zealand
July 1995

Interview with Professor Stephen Ceci, Cornell University
Interview with Russ Francis


(Highlighted text is dialogue used in TVNZ Assignment programme
"Ellis through the Looking Glass" broadcast on July 27 1995)


Russ Francis:        ………Christchurch Creche children were conducted?

Prof. Ceci:            In my opinion having read the material that was sent, those interviews were what I would call typical of interviews done five, six, seven years ago. they're not .. interviews that scientists would have designed to maximise accuracy and to minimise inaccuracy, nor are they the worst interviews I've ever seen but rather interviews that were traditional to the late eighties early nineties.

Russ Francis:        But I gather from that you do have some concerns about the way they were conducted?

Prof. Ceci:            Yes they were not conducted in accordance with the currently understood scientific principles.

Russ Francis:        But what in particular did they not do correctly?

Prof. Ceci:            Well there were many things that the interviewers didn't do, perhaps I can step back though and say that as a researcher reading these trial transcripts and interview protocol one is struck with the possibility that problems may have existed in these interviews that pre-dated or preceded the interviews themselves.

There apparently was quite a lot of conversation among the parents of the children in the creche, the parents with their own children over long periods of time before the very first electronically preserved interview took place and as a researcher we understand just how that kind of conversation and communication with the children concede reliability problems that later show up in the official interviews.

Ah I'd say that because no matter how good an interviewer is, if you have that kind of previous activity it may present insurmountable problems. But to get to your question there were a number of strategies and techniques that were not used by these interviewers and I'll just give you a few examples.

One of the most important tactics that an interviewer ought to use when you're interviewing pre-school-age children, and in particular three and four year old children which were the subject of this particular trial, is what we call proof by disproof...

Proof by disproof is just a fancy way of saying that while you're going about interviewing children and trying to get them to confirm your hunch about what happened you must also in the course of the very same interview try to falsify your hunch.

This can be done by a sensitive interviewer without conveying scepticism to the child's story. If the interviewer thinks that Peter touched the child's bottom, fine you can try to elicit confirmatory statements from the child in that regard but at the very same time one ought to also ask the child well what about the police officer - Did he touch your bottom? What about your mum, What about your dad? and so on....

It's necessary to see what kind of reliability problems the child presents in the course of the interview and that wasn't done at all in any of the interviews unfortunately, but that was rather characteristic for its time.

Russ Francis:        But it does call into question the conclusions which ??? from the interviews?

Prof. Ceci:            Ah I'm afraid so, I mean as researchers we can design studies that mimic many of the very conditions that were used by these interviewers and we can show that you... it is possible to bring children to give incorrect statements when you use some of these techniques.

Russ Francis:        Some of the critics of this case have charged that the people doing the interviewing approached the children from the viewpoint that the children had been abused. From your reading of the transcripts is there evidence to support that?

Prof. Ceci:            Well there certainly wasn't any effort to falsify the hunches throughout the interviews. Often there was repetition of questions. It was almost as though the interviewers were surprised that the child said that's all that Mr Ellis did to them and therefore they would repeat the same question over and over again.

There wasn't an effort to try to talk the child out of it by saying things such as ah... "Children have said some very silly things to us, so it's very very important that you tell us only what exactly happened".

There wasn't this effort to rein the children back into reality when they roamed into these fabulous claims ... which sometimes they did.  They were conveniently ignored by the interviewers. 

Whether or not the interviewers' minds were made up prior to the interviews I can't say but what I can say as an outsider reading them that there was no serious attempt to test an alternative hypothesis to the State's claim that Mr Peter Ellis molested these children.

Russ Francis:        Professor Ceci you just referred to repeating questions, could you give an example of that and what sort of effect that can have?

Prof. Ceci:            Certainly. We know from scientific studies that the one age group more than any other that you have to be extremely careful about repeating questions within and across interviews is three and four year old children.

If you say to them "tell me how Russ touched your bum?" and the child says "he didn't" and then five minutes later you say "now I'm going to ask you again, tell me how Russ touched your bum?" …... Children of that age are more likely than older children and more likely than adults to interpret the repetition to mean that the first answer they gave was incorrect, so they're more likely to change their answer.

You get that regardless of the ground truth of the manner so its a worry and interviewers really have to be quite careful when they do that. 

One finds an awful lot of question repetition in these interviews...again that... that's quite characteristic for its time, its not a hallmark of an awful interview ah.... a vigilante interview. I don't think these interviewers were vigilantes. Certainly their statements in the press struck me as the statements of very reasonable professionals. It's just that what we know in the most scientifically adequate way dictates against using these techniques.

Russ Francis:        Would you say.... call what... the way the questions were phrased a way of coercing the children into making statements?

Prof. Ceci:            Well I wouldn't say coercing in a sen... you know I didn't see videotapes of the interviews and perhaps I would feel differently had I seen that but ah... my analysis ??? wholly on a written transcription so therefore the demeanour, the affect, the gesturing, all that is omitted from the material I was sent, but just looking at the sanitised written transcription they didn't seem terribly coercive. They also didn't seem ah... how shall I put it... very interested in alternative scenarios?

Russ Francis:        Based on your reading of the transcripts what do you think the chances are that Ellis did abuse the children for something like five years without any... anyone even noticing, without a complaint?

Prof. Ceci:            You know these claims and I have... I've seen hundreds of claims that are very similar to this one, ah... often have an amalgam or a blending of credible and highly incredible claims and sorting the wheat from the chaff becomes extremely difficult. Some of the things the children said I would be exceedingly sceptical that they ever occurred and I think, in fact, the prosecution was exceedingly sceptical because they chose in many cases not to pursue some of the children's more bizarre claims.

It, in my experience, is exceedingly unlikely that you can coerce a group of children this age into silence for prolonged periods of time when the following were allegedly involved... anal insertion; forcing children to walk over precarious ladders perched high above buildings; defecating and urinating on children in bathtubs, in beds and the like. These are events which cause almost instant revulsion in children, night tremors, unwillingness to go to school, fear of the perpetrator and ah.... despite claims to the contrary, scientifically its now been established quite convincingly in my view that it is ah... very very unlikely that you could persuade children to be silent about that for long periods and also to exert external manifestations of affection for the perpetrator which many of these children did.

So on that level I'm extremely sceptical, I don't think the bizarre stuff happened. Does that mean nothing happened? Well I simply don't know. No-one else knows either except ah... God and Mr Peter Ellis.

Russ Francis:        The parents at the Christchurch creche were very close to each other and there was a strong network. I've given that ???? network do you think that the close questioning of the parents ?? their children, do you think its possible that there was some contamination there of the     ?

Prof. Ceci:            Yes, in reading the materials one is struck by the fact that the parents are understandably were extremely moved by the possibility that their children may have been molested. They talked with each other and they conveyed to their children often information that was passed on to them from other children.

Not only as a scientist but as a parent I can understand that reaction. Were my daughter involved in a scenario such as this creches I too would have strong impulse to pursue my daughter persistently over long periods of time saying to her well are you sure don't remember? So and So's mom told me that she said you were there when it happened and so on.

While that's a perfectly understandable human impulse scientifically we know that this.... is really the seeds for getting children to later assent to things that may not have happened and I think, if you'd like I can describe how that can be done?

Russ Francis:        Can you do that?

Prof. Ceci:            Yeah.... yeah.  If you bring mums into the laboratory with their children and you say to your research assistant take the child into the adjoining room, and you say to mum look we're very interested in how much your child can remember about events and we're showing your child a video, and we turn the VCR on and show the mum and say your child's watching this video next door and when she's done we're going to bring her in here and we want you to interview her and see how much she can remember from this video.

Now we do this scientifically so on one condition the mum's watching the same video the child's watching whereas in another condition unbeknownst to the mum the child's watching a totally different video. But you see here you have the ingredients where the parent thinks she knows what the child's experience was but she was wrong.

Now what happens in that situation? Well what happens typically is the parent starts interviewing the child and the child provides a limited verbal recall and what they do provide sometimes doesn't coincide with what the parent believes to have occurred so the parent starts becoming more and more suggestive, providing more and more structured ??? and close ended questions and you start to see three and four and five year olds assenting, they start nodding, they're giving one word answers.

When the mum comes back and re-interviews the child a second time the assents become sentences and the third or fourth time they get further embellished and you begin to get paragraph level narratives about events that the child never experienced but the mum thought that they did.

That's why it's exceedingly important for the interviewer to also try to falsify their beliefs.

Russ Francis:        The judge and jury did believe some of the things that the children said and actually found Ellis guilty of those but they dismissed some of the more bizarre charges. Is it... is it preferable... is it acceptable to make distinctions between claims like that or if some are rejected should all be rejected?

Prof. Ceci:            No I think ah the realities of child sexual abuse and the realities ???????? are such that you often will find the child where you can have a documented case of sexual abuse, and there are unfortunately thousands of these in my country and I suppose there are many hundreds in your own, where you have documented cases where there's probative medical evidence, where the defendant himself has plead guilty, where there may have even been an eye witness who had corroborated the child's claims. Even in those cases you will often find a blending of fabulous as well as plausible claims. So the fact that there are fabulous claims doesn't on its face require one to say that nothing the child said can be accurate. Having said that it is a very serious matter to ignore the incredible stuff and chalk that up to a vulnerable child striving for self-empowerment, for regaining control over their alleged victimisation.... yes he... he said that he karate chopped the defendant and tied him in chains and ran away and so on and we of course know that didn't happen but we think these other things did. You see while, in documented cases you can get fabulous and plausible intermingled, in false cases you also can.

Russ Francis:        Is it possible for the ???? for an expert to tell if something a child says is fact or fantasy?

Prof. Ceci:            Read my it is not. One of the things that bothers me more than anything else is the cadre of instant experts who make their way into North American courts and I suppose also into New Zealand, Australian, other Commonwealth country courts that profess with a great deal of certainty to judge and jury that they have a Pinnochio test - that they can tell when a child's statements are accurate or inaccurate.

There is no Pinnochio test, were there one I would bottle it and try to make a lot of money. I've spent my life doing ??? analysis of very young children's statements and there is no diagnostic set of criteria, there's no set of affective emotional reactions, there's no linguistic analysis, there are no memory demands, there's nothing that has ever been established in a scientifically adequate way that can discriminate between accurate and inaccurate statements when the child has been repeatedly interviewed over long periods of time.

Russ Francis:        Well given that is it safe to convict someone on the uncorroborated evidence of children?

Prof. Ceci:            Well you have a complicated situation in that sexual abuse is almost always a private act; there are only two people involved; there's almost never probative medical evidence; there's almost never a third party eyewitness to be asked, so it's always "He said, She said." Fact finders are always faced with this very private act believing one person, disbelieving another. If you say that you're going to require corroboration for sexual abuse you're essentially saying you'll never have convictions.

Russ Francis:        From what you've looked at ???? New Zealand, how does New Zealand practice compare with the best international methods of interviewing young children?

Prof. Ceci:            Ah this creche case in Christchurch could have been in any North American cities that I'm familiar with, the very same constellation of factors that one would find in New York, Los Angeles, London, Sydney, were also evident in this case.

Russ Francis:        So it's not unusual?

Prof. Ceci:            Not at all.

Russ Francis:        How common are false convictions do you believe?

Prof. Ceci:            Well no-one knows. We... we know that in my country there are approximately a half million cases of alleged sexual abuse of minors each year... major state maltreatment hotline.... and of these approximately 40% are called indicated or substantiated and that means simply that an initial interview provides information that's consistent with the claim that sexual abuse, and only a small percentage of those actually get filed with the police departments and wind their way to full trial.

Of those if you said that 90% or 95% resulted in true carriages of justice that would still leave 5 or 10% that will not. To be quoting these numbers you're dealing with it could result in some real sad miscarriages of justice but the actual numbers are anyone's guess. There's been studies that have said seven percent, some studies have said fifteen percent of sexual abuse claims are invalid... and I take no position on the actual percentage, we need far better data I think.

Russ Francis:        Professor Ceci you've studied hundreds of similar cases to the Christchurch creche case. From what you've read of this case would you send your child to a creche where Peter Ellis worked?

Prof. Ceci:            Well it's difficult to get a sense of the man from the written materials only and not having any audio-visual input. One gets the sense that he engaged in some behaviours that were at the very least unfortunate, for example rough-housing, locking children in the bathrooms until they pounded on the door to be released, um these aren't the kind of behaviours that perhaps one would prefer to have their child exposed to. On the other hand you get the sense that he was a very vibrant vivacious person that was able to get on the children's own level and quite a few children seemed to have a great deal of affection for him. But I'm really hesitant to say much more without really knowing something more about the person.

Russ Francis:        And what about... do you believe Peter Ellis should be in jail?

Prof. Ceci:            Oh dear I have no thoughts on that at all.

Russ Francis:        Do you have any other comments about the Christchurch Creche case?

Prof. Ceci:            Ah   I mean I could just elaborate. I suppose on some things but I don't have any burning new points....

Russ Francis:        I think we've covered just about everything. You talked a little bit before about ..


Russ Francis:        Can you tell us about some of the consequences of allegedly false allegations of sexual abuse in the United States?

Prof. Ceci:            The two most awful things that can happen are a) you are abused as a child and no-one believes you, and the other thing is to be accused of abusing a child and to be innocent. Because when the latter happens it's also exceedingly difficult to prove your innocence, people have a tendency to believe... you just simply weren't found guilty, there wasn't enough evidence for them to get you this time.

Lives are shattered, homes are mortgaged to pay for the legal bills, neighbours ostracise you, there's always the sinking suspicion that you did it but the state simply couldn't prove it.

And there have been in my country a whole string of cases that have been overturned on appeal, some as high as the Supreme Court, that most of us... most of us who are experts in this area feel were miscarriages of justice.

Quite recently the supreme court in New Jersey overturned a conviction of a woman named Margaret Kelly Michaels and forty five of us were scientists that worked on this problem signed an amicus brief, a friend of the court brief, where we analysed the children's testimony and the interviews in that case, and forty five of us put our names to this document saying that we felt very strongly that the conditions to foster unreliabilities were present in these interviews. Only three out of forty eight chose not to sign that. So you'll never get a hundred percent where passions run high but where you're able to get ninety plus percent of experts I think that's fairly telling.

Russ Francis:        I want to ask you about the pattern.  You were saying before that, I just want to ask you about the pattern.  Does the Christchurch creche case fit a pattern, is there a particular way that these allegations evolve?

Prof. Ceci:            Ah this case has the same constellation of ingredients that ninety plus percent of mass allegation creche cases have.

What do I mean by that? Well there's always a trigger that starts it and the trigger is a single allegation. It may come in spontaneous discussion, the child might be bathing or having a conversation before naptime with a mum and make a statement saying "so and so touches my bottom" or "so and so puts a stick in my bottom or puts a pin in my penis or has a black penis". But a single statement gets made.

That causes the parent to have some worry, mull it over, maybe they talk to a paediatrician, maybe they call social services, maybe they re-interview their child, they talk to parents of other children, before you know it the other parents start talking to their children. They say "well so and so already told his mum that this person did this naughty thing and he said he did it to you as well, don't you remember?"

And there's this air of accusation that starts after the initial trigger where parents now are starting to paint the person as a bad person, he defecates on kids, he urinates on kids, he karate chops and hurts kids etc etc etc.

Before you know it there are confirmatory biases, by that I mean there's no attempt to test an alternative hunch but simply let's get the goods on this person. At that point parents begin talking to each other. That information that gets swapped gets translated to the children. They start saying "well so and so told her mum already and don't you remember that, don't you want to help them keep so and so in jail so he doesn't hurt any other children?"

All these cases including the Christchurch Creche case have these kinds of features to them. The allegations grow over time. They start with one or two criminal charges and they often end up with dozens, even hundreds of complaints by the time all the interviews are done.

They all include an amalgam of highly plausible events and highly implausible events. And I could go on but there is a real script that these cases follow.

Russ Francis:        So the Christchurch case is certainly not unusual?

Prof. Ceci:            No it's very typical..