Whether expert evidence is “reliable”, “valid”, or just “expert” is an important consideration, write Emeritus Professor Brynn Hibbert AM and Professor Gary Edmond, in conjunction with Unisearch.
In criminal prosecutions and civil disputes, expert opinions can be determinative. Think DNA tying a person to the scene of a crime, chemical analysis of contaminated river water, or analysis of trace impurities in a multimillion-dollar shipment of lithium ore.
Courts set great store by the “expertise” of persons giving opinion evidence, as per Honeysett v The Queen (2014).
The venerable professor is clearly an expert (in something) but is his/her/their evidence “reliable”? Does it provide the court, specifically the fact-finder, with accurate information? Are the underlying methods valid? How does a court know whether the evidence is appropriate (or representative), or determine whether it is sufficient to draw a legal conclusion? Analytical chemist Professor Hibbert’s position is that chemistry may have its problems — but even worse is the identification of evidence from fingerprints, CCT images, handwriting, voice recordings, and the like.
Expertise is certainly worth having, but it is often conflated with experience or treated expansively. Professor Hibbert has appeared in many drug manufacturing cases where the opposing expert, often called by the Crown, states that he/she/they have testified in numerous proceedings. For example, on the question of how much drug could have been made by a clandestine laboratory: “many times over the past 20 years”. However, no one asks: “but is the method used to determine quantity sound and how might we evaluate the opinion?” No one asks — “how do we know they can actually do it (analyse the drug, match the fingerprint)?”
“Reliability” is a concept that addresses whether a scientific method and the evidence that comes from it, gives results that are sufficiently near ground truth to be useful.
Validation is the process used by scientists to acquire the necessary information to assess the ability of a procedure to obtain the correct result, determine the conditions under which such results can be obtained, and indicate the limitations of the procedure. The validation process identifies the critical aspects of a procedure that must be carefully controlled and monitored, as noted in Tuite v Queen . Metrology (the science of measurement, of which all of this kind of evidence forms part) stresses the need for validation to be demonstrated, not just asserted.
Before a scientist switches on the GCMS instrument or starts a feature comparison, he/she/they must be able to demonstrate that the method to be applied gives results that are valid and reliable. A valid method will be:
- Repeatable: the scientist will obtain the same results in repeated tests on the same sample;
- Reproducible: different scientists will obtain the same results, within measurement uncertainty, from the same sample; and
- Accurate: with known probabilities, the scientist will obtain a positive result from a truly positive sample, and a negative result from a truly negative sample, according to the US-based President’s Council of Advisors on Science and Technology.
The methods used must be fit for purposes of the litigation (and the facts in issue) and the resulting opinions capable of being understood and evaluated by fact-finders.
It is important that lawyers understand the need to question scientific evidence and be aware that, on many occasions, what is presented as solid science may be untested or speculative opinion.
Australian barristers and courts have often been less interested in reliability and validity than seems desirable for a putatively rational legal system, as per R v Tang (2006). However, there are occasions where courts have understood the need for validity and reliability, and lawyers can learn from them.
In 2020, the Supreme Court of Queensland excluded evidence of lead isotopes used to “match” bullets removed from a murder victim with those in a box in possession of a defendant. The trial judge was “not satisfied that the methodology employed satisfies the criteria which have been set out as necessary to establish a foundational validity for a forensic feature comparison method” (Pentland v The Queen ). “Clearly, this evidence, given it does not meet the threshold test for scientific validity, if it were to be admitted would clearly be more prejudicial than probative because if it is not scientifically valid, it has low probative value.”
The judgment highlighted both the lack of any documented validation behind the analytical chemistry procedure along with the lack of a statistical basis for attaching significance to the match were the procedure to be accepted. (How many other boxes of bullets have similar lead isotope contents?)
What does this mean for lawyers?
- Critically review evidence. The legal team should critically review expert evidence, using independent and/or multiple experts, if necessary;
- Experts must come from the appropriate field. It is a waste of resources to ask a quantum physicist to offer advice on ergonomics;
- Accept expert evidence. Be prepared for a genuine expert, respecting his/her/their duty to the court, giving you bad news. It is better to find out before trial than in the middle of it.
- Utilise the experts’ broader skill set. Apart from a report, an expert might advise on cross-examination and aspects of the case that ought to be probed;
- Ask the tough questions. You should always ask: “can the putative expert do what they claim?” Plus: “Where is the evidence – of validity and reliability – supporting the claimed ability?”;
- Define expertise. Experience, or doing something over and over, does not necessarily confer expertise. Methods and experts should be rigorously tested to determine, respectively their validity and abilities (including the conditions in which they are reliable and/or other limitations); and
- Favour impartiality. Some traits of an impartial expert include being transparent and proactively disclosing information about validity, uncertainty, and limitations.
Emeritus Professor Brynn Hibbert AM was the chair of analytical chemistry at the University of New South Wales, and he is a Unisearch expert. Professor Gary Edmond is a law professor in the School of Law at the University of New South Wales, where he directs the program in expertise, evidence and law.