Ellis' lawyer Rob Harrison says the children's memories were "contaminated" by a number of factors – including having conversations with their parents about the allegations, overhearing conversations between other adults and being given access to therapy and books or dolls which depicted sexual abuse.
Professor Harlene Haynes also said the children's interviews fell far short of best practice, even for the time – and says experts for the Crown agree with her.
On Friday Haynes told the court Professor Gail Goodman for the Crown agrees there was moderate to high risk of contamination for a number of the children.
"[Professor] Goodman and I are in large agreement on these contamination opportunities. Goodman rated child 3 as having moderate risk and child 4, 5, 6 and 7 as extreme."
Haynes says in one of the pre-recorded interviews no fewer than 21 suggestive questions were asked of the child.
However, Dr Suzanne Blackwell for the Crown says Haynes' analysis of what counts as a leading question "suggested a risk of being over inclusive as abuse-related".
Blackwell says some of the things that were raised as being suggestive included "talk of dolls or drawings".
"Does this include talk about dolls which may refer to neutral aspects of the doll such as clothing? Another was anything to do with the creche – does this include things like when they attended, what they did?"
Blackwell says the age of the children is an important factor to keep in mind, the average age of the children being interviewed was 6.
"We know they offer less information and often overtly require prompting and questions."
Blackwell acknowledged there were "many" direct and suggestive questions in the interviews which was "undesirable" but the salient point was context.
"It is relevant to state these transcripts were coded in 2004, 2005 by Haynes' masters' student – the videos were not watched, and the tone of the voice of the interviews gave clues as to whether the questions were leading or establishing rapport."
The use of props, such as dolls, toys and drawings were also debated in the court.
Goodman told the court the interviews did use more props – but that doesn't mean they were not up to standard.
"Research at the time showed re-enactment increased very young children's recall. Interview techniques using props can illicit more answers than verbal questioning alone."
She says a lot of the interviewers were doing their best – and often began with neutral questions like "what have you come here to talk to me about today?" before taking the children's lead.
Professor Frederick Seymour for the Crown agreed, saying the use of props was mostly used when children had already disclosed sexual abuse and the dolls were just for clarity.
"These targeted interactions took place after the substantial abuse report had already been made by the child. The presentation of these dolls was in keeping with protocol of time following the child already giving a credible description of abuse."
He says dolls did not lead to any reports of new abuse, and it is his opinion they did not have a significant effect on any of the children in the Ellis case.
But Haynes says it is her opinion there was nothing that could undo the damage of contamination.
"As a memory expert it is my strong opinion, that I believe is shared by every other expert in this case, that was what happening in this case, there is a clear risk their accounts were contaminated prior to the interviews. I am not aware of any technique or court procedure that could undo that contamination effect."
The appeal continues next week.
