First, there was the danger that Dr. Smith’s evidence
would usurp the jury’s function of deciding credibility. As noted
by the Supreme Court in R. v. Marquard,  4 S.C.R. 223,
 S.C.J. No. 119, at para. 49, “credibility is a notoriously difficult problem, and the expert’s opinion may be all too readily
accepted by a frustrated jury as a convenient basis upon which to
resolve its difficulties”.
 Second, the jury had before it, in Dr. Smith’s unedited and
unredacted reports, a great deal of material that was irrelevant to
issues in the proceedings, but that contained tremendous oath-helping potential. For example, Dr. Smith once referred to an
incident where Mr. Imeson admitted to having taken another
inmate’s medication. Mr. Imeson’s explanation, according to
Dr. Smith’s report, was that he came forward with the information because “he had committed himself to honesty”.
 To her credit, the trial judge acknowledged that the oath-helping concern was genuine, but determined that it could be
addressed through her mid-trial jury instruction and in her final
instructions. She did not, however, weigh all the risks I have identified against the possible benefits of admitting the evidence, or
consider how the risks might be mitigated other than through the
hearsay caution. In my view, the hearsay caution was insufficient
to address the various risks present in this case, including the
very real risk of oath-helping.
 As this court noted in Llorenz, oath-helping evidence
should not be admitted simply because it is led for another purpose. A court must still weigh the probative value of the evidence
against its prejudicial effect: “In cases which turn on the question
of which one of two witnesses is telling the truth, there is the
danger that a jury may attach significant weight to the oath-helping aspect of the evidence of an expert, even if instructed to
do otherwise”: Llorenz, at para. 32.
 In conclusion, Dr. Smith’s opinion evidence going to the
issues of liability and causation not only exceeded the scope of
proper opinions to be offered by a participant expert, but also
failed to satisfy the Mohan/White Burgess test for admissibility.
(3) Effect of the trial judge’s error
 The final issue concerns the effect of the trial judge’s error in
admitting Dr. Smith’s evidence on the issues of whether the assaults
occurred and whether Mr. Imeson suffered any harm as a result.
 The appellant contends that the admission of Dr. Smith’s
opinion evidence on those issues was a fatal error that would
require a new trial. The respondent does not really argue to