Reliability and relevance

Concealed within the requirement that evidence, to be admissible, must be relevant, is a threshold of reliability.

It is a fundamental principle that relevant evidence is - subject to exclusionary rules -  admissible, and that evidence that is not relevant is not admissible: see, for example, s 7 of the Evidence Act 2006[NZ]. “Relevant” means having a tendency to prove or disprove anything that is of consequence to the determination of the proceeding. Is any “tendency”, no matter how slight, sufficient to establish relevance? Logically, the tendency of evidence to prove something must include a requirement that it has some reliability. The need for reliability can be seen from the use of that term in the Act.

The reliability requirement is found as an express requirement in the Evidence Act 2006[NZ], as a qualification on the admissibility of particular sorts of evidence. For example, hearsay evidence requires, from the circumstances in which it was obtained, “reasonable assurance that the statement is reliable” (s 18(1)(a)), unreliable statements by defendants in criminal proceedings must be excluded (s 28(2)), previous consistent statements must have been made in circumstances that provide “reasonable assurance” that they are reliable (s 35(3)(a)), and there are reliability rules concerning identification evidence (visual identification: s 45(1) and (2), voice identification: s 46). These references to reliability are consistent with the view that reliability is a fundamental requirement for admissibility.

A concept akin to reliability is probative value, but they are not the same. Evidence cannot have probative value unless it has reliability, and increasing reliability may lend the evidence more probative value, but not necessarily. Although highly reliable, evidence might have relatively slight probative value. Reliability reflects accuracy, whereas probative value is a reflection of the importance of the evidence to the logic of the case.

In the terminology of the Act, probative value is something that a judge must weigh against unfairly prejudicial effect, and this applies generally (s 8). The expert opinion rule in s 25 is governed by this probative value determination, because of its general application, but neither probative value nor reliability are mentioned in s 25. This raises the question: does expert opinion evidence have to be reliable before it can be admissible?

Expert opinion is admissible if (s 25(1)) “the fact-finder is likely to obtain substantial help from the opinion in understanding other evidence in the proceeding or in ascertaining any fact that is of consequence to the determination of the proceeding.” Here, “substantial help” may well be contingent on reliability; this seems to be the only route by which the common law requirements, of general acceptance of the opinion among experts in the field, can come into play in s 25.

Yesterday, the Supreme Court of Canada split 6-3 on what was essentially a difference about whether a threshold of reliability is a fundamental requirement for the admissibility of evidence: R v Trochym [2007] SCC 6 (1 February 2007). Here, the evidence was that of a witness who claimed to have seen a person she said was the accused, leaving the victim’s apartment, some time after the victim was murdered. She gave two versions of exactly when she had seen this: her first version was that the accused had left the day after the murder, but later, under hypnosis, she changed this to the same day as the murder. Was the version she had given under hypnosis sufficiently reliable to be admitted in evidence, so that its weight (probative value) could be assessed by the jury? Or, should the first question be whether her hypnotised version had sufficient probative value, compared to its unfairly prejudicial effect, for the matter of its weight to be left with the jury?

The majority (McLachlin CJ, Binnie, LeBel, Deschamps, and Fish JJ, with Charron J concurring but differing slightly, particularly as to the position on the part of the witness’s evidence that was not the subject of hypnotic inquiry) held that reliability is the fundamental requirement for admissibility as far as “novel scientific evidence” is concerned. Whereas in many areas of scientific inquiry the reliability of expert evidence is established, science develops:

“31. Not all scientific evidence, or evidence that results from the use of a scientific technique, must be screened before being introduced into evidence. In some cases, the science in question is so well established that judges can rely on the fact that the admissibility of evidence based on it has been clearly recognized by the courts in the past. Other cases may not be so clear. Like the legal community, the scientific community continues to challenge and improve upon its existing base of knowledge. As a result, the admissibility of scientific evidence is not frozen in time.

“32. While some forms of scientific evidence become more reliable over time, others may become less so as further studies reveal concerns. Thus, a technique that was once admissible may subsequently be found to be inadmissible. An example of the first situation, where, upon further refinement and study, a scientific technique becomes sufficiently reliable to be used in criminal trials, is DNA matching evidence, which this Court recognized in R. v. Terceira, 1999 CanLII 645 (S.C.C.), [1999] 3 S.C.R. 866. An example of the second situation, where a technique that has been employed for some time comes to be questioned, is so-called “dock”, or in-court, identification evidence. In R. v. Hibbert, 2002 SCC 39 (CanLII), [2002] 2 S.C.R. 445, 2002 SCC 39, at para. 50, Arbour J., writing for the majority, stated that despite its long-standing use, dock identification is almost totally unreliable. Therefore, even if it has received judicial recognition in the past, a technique or science whose underlying assumptions are challenged should not be admitted in evidence without first confirming the validity of those assumptions.

“33. … Even though the use of expert testimony was not in itself at issue in the present case — this appeal concerns the application of a scientific technique to the testimony of a lay witness — the threshold reliability of the technique, and its impact on the testimony, remains crucial to the fairness of the trial.”

The emphasis is on the threshold of reliability and the fairness of the trial. The majority concluded, para 55, and 65-66, that “this technique and its impact on human memory are not understood well enough for post-hypnosis testimony to be sufficiently reliable to be used in a court of law.” That applies to the witness’s evidence on topics that have been the subject of hypnosis. As far as evidence from the witness about matters that have not been mentioned under hypnosis, (as held by the majority except Charron J) the judge determines admissibility by assessing whether the detrimental effects of hypnosis are outweighed by the probative value of the evidence (para 64). Charron J would not require the balancing exercise to automatically apply here (para 87), and she would also allow evidence of topics covered under hypnosis to be given if supported by a pre-hypnosis consistent statement (para 88).

The minority (Bastarache, Abella and Rothstein JJ) would have admitted the post-hypnotic evidence, preferring not to change the rule. They declined to classify this area as novel science, and would have confined the reliability test, in this context, to novel science. They considered that the majority had placed too high a standard for consensus in the expert community. In particular, they held that the majority’s approach, which took judicial notice of cases in the USA in which experts had expressed doubts on the reliability of testimony by witnesses who had been hypnotised, amounted to denying a party in the present case the opportunity to cross-examine those experts.

The Court in Trochym was unanimous on another aspect of the appeal: similar fact evidence had been wrongly admitted. A single incident (knocking on the victim’s door after a fight with her) did not disclose a pattern of behaviour, and was not unique conduct that identified the offender (paras 74-78, 83; 185, 187).

This case illustrates that there are times when the judge cannot simply say that the issue of admissibility depends on whether the probative value of the evidence outweighs its unfairly prejudicial effect. The first requirement is relevance, and relevance depends on whether the evidence has a tendency to prove or disprove a matter in issue, and that in turn depends on a threshold of reliability. Doubts about the probative value of evidence may extend beyond the balancing exercise, to the more fundamental question of relevance.