Section Four

Changes in Best Practice


(4) Changes in Best Practice

(4.1) The Prue Vincent case
(4.2) Non-verbal leakage and lie detection
(4.3) Truth testing
(4.4) The official word in changes to interview protocols
(4.5) Continuing inaction from the NZ Psychological Society
(4.6) Elizabeth Loftus’s recommendations
(4.7) Conclusions

Next Section
Previous section
Document Contents | Intro | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | References | Appendices |

(4) Changes in "best practice."

"Eye-poppingly bad"

Dr Maryanne Garry describing the work of psychologist Prue Vincent (December 2001)

(4.1)     The Prue Vincent case

The startled face of psychologist Prue Vincent staring out of page one of the Dominion of 4 December 2001 is indicative of the changes in "best practice" since the Ellis case. Vincent had "botched a sex abuse investigation that left a man wrongfully accused of molesting his young children." (Leah Haines, The Dominion). As was quickly pointed out by other commentators, Vincent in many respects had operated in a manner similar to that of Zelas, Sidey and others in the Ellis case. She pleaded guilty to charges of conduct unbecoming of a psychologist. Vincent was once head of Social Welfare’s psychologists team. Here are some of the charges she pleaded guilty to, alongside similarities to the conduct of Zelas and Sidey in the Ellis case

o        Vincent used books dealing with sexual abuse during her assessment. I assume that they were books read to the children about bad touching, as happened in the Civic childcare centre case where the parents read such books to their children. In Christchurch, some children brought along books they had made themselves, with their parents’ help, to the evidential interviews. These contained pictures and statements accusing Ellis of abuse. They were not told to stop doing this by Sidey or Zelas. Some parents had also introduced their children to books describing sexual abuse in children’s terms, before any allegations arose (Hood 2001, Sas 44 and see section (6.1) this report).

o        Vincent used leading questions when interviewing children. On page 301, Hood discusses this issue. Leading can include questions that are closed to the extent that a yes or no answer is elicited. Most parents, and certainly at times Sidey, asked leading questions in Christchurch.

o        Vincent did not consider other explanations for the children’s behaviour. Hood’s book is full of countless examples of this in regard to Sidey’s and Zelas’s work (see this report). The Eichelbaum report does not even mention most of the alternative interpretations discussed in Hood’s book.

o        Vincent did not observe the children in their wider environment. In the Ellis case, there was ample opportunity in the early parts of the police inquiry to observe the children at the Civic childcare centre with Ellis. Only the E.R.O office did so, and found no evidence that anything was amiss.

o        Vincent did not interview the father as a reference source. Until he was arrested late in the enquiry, no one thought to talk directly to Ellis. This would have been the obvious thing to do, especially in looking for alternative explanations. Ellis did in fact have plenty of alternative explanations for much of what the children said, but these were not heard until the trial, and then only through his lawyer. There may be ground here for an official complaint to the medical council about Zelas’s approach.

o        Vincent accepted without question the mother’s testimony whilst asking the father to put his rebuttals in writing. Ellis was not even asked to express his rebuttals at all during the interview phase. Had Zelas and Sidey followed this procedure, the case may never have come to court. In this respect, they are even more culpable than Vincent, and would certainly have to plead guilty to this charge as well.

o        Vincent allowed the mother to be present during the child interviews. I have no evidence that this happened during the evidential interviews in the Ellis case, but during the Ellis trial Justice Williamson, on the advice of Zelas, allowed a mother to sit with her child in the video room. (Hood p485). In addition Detective Eade interactively influenced the course and focus of some interviews by being in communication with the interviewers during interviews.

Vincent was fined $5,000, and given a letter of censure from the Psychologist’s Board. CYPS have stated that they will not be employing her again. Victoria University senior lecturer, Maryanne Garry has described Vincent’s work as

"eye-poppingly bad."

Hood appears to describe the work of Zelas and Sidey in much the same way. The Dean of Law at Otago University, Mark Henaghan, has called for a judicial review of the Family Court as a result of the Vincent case.

Back to section contents

(4.2)     Non-verbal leakage and lie detection

A new development has occurred in the field of lie detection. Videos of interviews are analysed for evidence of non-verbal leakage in the form of fleeting facial expressions, difficult to see under normal conditions. This is based on the work of Dr Paul Ekman. There are other forms of non-verbal leakage such as maintaining eye contact for longer than usual (an attempt to appear ‘sincere’ whilst lying). A problem with non-verbal leakage analysis is that once a subject knows what you are looking for, they may try to suppress those behaviours. Because children can believe their own untruths, such an approach may be less reliable than when used interviewing adults.

With regard to the children, there is probably little value in a re-evaluation of the videos using Eckman’s techniques. The child who retracted, described feelings of self-revulsion, and feeling sick whilst in the courtroom due to her act of lying. Unfortunately, she was the eldest witness, so the others, if mistaken, probably were sincere, as younger children are more susceptible to memory construction. Nevertheless, there may have been some leakage occurring in some cases. If the research into non-verbal leakage is as good as claimed by Ekman, then this type of analysis needs to be part of modern best practice.

Back to section contents

(4.3)     Truth testing

In section (3), I discussed effective truth tests. Kim Hill interviewed Professor Stephen Lindsay on her National Radio programme in 2000. He claimed that his research had shown that his new technique applied at the start of an evidential interview reduced the incidence of false assertions made by children on one question from 67% to 30%. This has implications for best practice. Obviously the children in the Ellis case were frequently telling untruths about cages suspended from the ceiling, and teachers dancing naked in circles, that Ellis had dug up Jesus, that Ellis had made children explode etc, etc, the truth tests applied were rather ineffective, and possibly capable of considerable improvement. In some cases where they were mistaken, however, the children may still have thought that they were telling the truth. In those cases, the truth test would not have been relevant. The important point is; if Lindsay is right, then the truth testing procedure used in the Ellis case interviews was clearly less than ideal.

Back to section contents

(4.4)     The official word in changes to interview protocols

In 1996 police legal adviser Mark Copeland and child abuse interview training co-ordinator Wendy Burgering publicly stated that a maximum of one formal interview is now the standard. (McLoughlin 1996). This is very significant, because few allegations arose during the first interview with each child in the Ellis case.

In 2002, Child, Youth and Family’s Karen Wilson provided on request the following statement about current practice. Her role at the time was manager and part-time specialist interviewer at the CYF’s, Manuwai Specialist Services in Hamilton.

"The New Zealand model includes explaining ground-rules to children eg. permission to correct the interviewer, permission to use don’t know, can’t remember, and don’t understand. Interviewers are taught to use a neutral tone and expression throughout, to ask source monitoring questions and to avoid introducing any prior knowledge into the interview.

Anatomical or genitally detailed dolls are not recommended as tools in New Zealand interviews. Clothed cloth dolls can be used to clarify body positioning if the child’s verbals
[sic] are unclear. Body outline diagrams (not genitally detailed) may be used late in the interview following a verbal allegation and if clarification is required eg. if the child has used the word "bum" s/he may be asked to put a mark on the diagram where they meant by that part."

This description differs radically from the way interviews were conducted in the Ellis investigation. Genitally detailed dolls were used, the interviews are generally lacking in source monitoring, and interviewers were obviously often briefed in advance as to what allegations the children had made to parents.

CYFS are now understood to be planning on training interviewers to carry out up to six "diagnostic" interviews before the evidential one. This seems to be an attempt to carry out more than the one or two recommended interviews. Davies, in the Eichelbaum report, states that most guidelines recommend a maximum of one or two interviews. The extra six initial interviews can be seen as an attempt to create a loophole, as they "don’t count" and are not recorded. Yet there is ample scope within those interviews to get impressionable young children to make allegations (or, "disclose" as some interviewers say) through invalid means. There is also no way a jury later can judge the conditions under which the allegations may have initially come about. Psychologist Dr Parsonson also makes the point that in his opinion, some of the people having an input into current interviewing protocols "don’t know how to analyse research." i.e., in his opinion they are not well equipped (perhaps through inadequate research training) to discriminate good science from bad.

Back to section contents

(4.5)     Continuing inaction from the NZ Psychological Society

Some psychologists and commentators have suggested that the New Zealand Psychological Society could sort out the issues through a panel of level-headed scientists. Cheryl Woolley, a senior consultant clinical psychologist and then president of the society, was asked about the idea of a panel late in 2002, and gave only one curious and ambiguous statement, "convening a ‘panel’ would solely report on the issues of that individual grouping which already occurs". Unfortunately she did not elaborate on what grouping she was referring to. It appears that the society is unlikely to act under Woolley’s leadership.

The society did publish a book containing practice guidelines and a review of issues in 1997. This publication is currently under review and a second edition will be published in 2003. Some of the material is relevant to the Ellis case, especially a contribution on memory of traumatic childhood events co-authored by Professors Corballis, Dr Mel Pipe (then at Otago) and Ms Judith McDougall of Victoria University.

Corballis says, 

"The New Zealand Psychological Society does not really speak for all psychologists. Many (perhaps most) academic psychologists do not belong to the Society, or have resigned. There is really no alternative body representing psychological science, as there is in the US and other countries. I do think that the Society has done its best to deal with the Ellis case and related problems, and some influential members (such as Barry Parsonson) have spoken out. But the membership is probably too skewed away from mainstream psychological science to be able to deal easily with matters that involve basic psychological processes."

Back to section contents

(4.6)     Elizabeth Loftus’s recommendations

In her 1993 study, Loftus concludes with this comment:

"Techniques that are less potentially dangerous would involve clarification, compassion, and gentle confrontation along with a demonstration of empathy…"

I contend that the concept of gentle confrontation would not have been accepted at the time of Ellis’s trial.

Back to section contents

(4.7)     Conclusions

Eichelbaum did not have the benefit of knowledge of the Prue Vincent case when he prepared his report. However, the case illustrates how widely he missed the mark when he claimed (on p109) that "Even by present day standards, it [the interviewing] was of a good overall quality." As discussed earlier, Eichelbaum should have been made aware that a number of practices in the Ellis investigation are no longer in general use by CYPS or psychologists in New Zealand now.

o        The most obvious is the use of anatomically correct dolls.

o        The number and average length of the interviews was excessive, the truth test procedure may have been inadequate.

o        There was insufficient source monitoring in the form of challenging allegations.

o        Books were brought into the interviews

o        Interviewers mentioned other children’s allegations

o        Interviewers were primed with details of other children’s allegations

Although evidential interview standards appear to have improved, there appears to be cause for concern in that there may still be insufficient scientific input and monitoring. In addition, new moves seem to allow unrecorded multiple informal interviews that precede evidential ones. This may be a backward step.

Memory expert Maryanne Garry stated to the parliamentary select committee looking at the case on 10 December, 2003 that interview techniques have changed significantly since the Ellis case. Specifically:

"Anatomically correct dolls are no longer used as children are curious and their actions could be misinterpreted; instead, they would [now] be encouraged to do such things as draw pictures.

"When there's no evidence, we don't try and dig evidence out.

"We don't use techniques that we know now are dangerous, in the same way we wouldn't be putting asbestos in buildings any more. It was fine then. It's not fine now."

Michael Corballis (2003) states that our understanding of how memory works has also greatly improved in the last few years.

" It is now widely recognised…that episodic memory (memory for specific events) is highly fallible, and experiments showing how easy it is to create strongly believed false memories are becoming standard undergraduate laboratory exercises."

Karen Zelas, Louise Sas and Eichelbaum all recklessly disregarded these advances in knowledge when accepting the dated and now unacceptable interviewing practices carried out in the Civic childcare case. None of these practices (bulleted in the forgoing) would be likely to be acceptable today. Eichelbaum could be considered reckless in saying in his conclusion that they would be.

End Section Four

Next Section
Previous section
Back to section contents
Back to document contents