---

The power of inculpatory DNA evidence is without equal.

Failure to realise the exculpatory nuances of DNA evidence can be catastrophic.

“ There is still a lot we don’t know about DNA.” [1]

The Australian criminal justice system relies upon inculpatory DNA expert evidence with great frequency and effect.

Since the first DNA conviction in 1988, the developmental trajectory of DNA within the criminal justice system has been nothing short of revolutionary. Initial trepidation of DNA has been replaced with a widespread level of familiarity and trust.

“The courts have approached DNA evidence with caution. However, that caution is naturally abating as experience with the use of such evidence has grown.” [1]

The proliferation of DNA forensic evidence has given rise to widespread practitioner and judicial participation. The legal community has taken up the mantle of comprehending DNA and continues to absorb knowledge on the subject at a fantastic pace.

The general understanding of DNA is an achievement and simultaneously is masking of the dangers.

“DNA’s apparent certainty can be deceptive”. [1]

It was “who” now we ask “how”

In forensic evidence, DNA can be broadly categorised into the areas of “identity” and “activity”. By far, the substantial basis of scientific evolution and our understanding relates to identity. Whilst there are still mistakes and areas of paucity within specific matters as they relate to identity, those errors relate substantially to human and subjective factors, rather than a failing of the science.

Regarding identity, the courts have justifiable confidence in the science.

Finding comfort with the information around “who”, the natural progression is for the courts to seek information as to “how” and “when” the DNA came to be where it is. This how and when is the concept of “Activity Level”.

The concept of Activity level is not new nor novel, our assessment however is at the early stages of scientific development.

“ the scope of DNA evidence has dramatically expanded, with transfer or ‘activity level’ information developing into an issue as critical to the courts as the question of identity “ [2]

Regarding activity, the courts should view the science with great caution.

Activity Level Reporting

DNA is highly transient and can be readily deposited, carried and subsequently redeposited. DNA can  be transferred by touch and contact, carried through the atmosphere, projected by talking, and cells containing DNA can be transferred during washing. These factors amongst many others, mean that any effort to assign a probative value or weighting at activity level, is highly scientific and contingent upon the exact and specific scenario.

Activity level reporting (ALR) is the quantification of the likelihood between competing action propositions, typically being a prosecution hypothesis and a defence hypothesis.

A DNA expert activity level opinion must evidence a numerical ratio. This likelihood ratio is much like you would see with identity (i.e. 700,000 times more likely).

Simply put, activity level reporting is not at all a mere case of saying something is ‘more likely’ or ‘probable’.

It is widely acknowledged within the forensic science community, that expert opinion as to likelihood in evidence shall not be speculative and shall only be based on an auditable body of knowledge. The use of verbal qualifiers cannot be used in isolation of a numerical qualifier.

Activity Level Reporting before the courts

With broad comfort around identity, we are seeing sustained jurisdictional pressure particularly at trial stage, for scientific input associated with activity.

Activity level has traditionally, and inappropriately, been dealt with at hearing with the insertion of unscientific and ad hoc probability assignments. Such DNA expert opinion will commonly be tendered verbally in chief or under cross examination, in support of the prosecution version being “more likely’ or “most likely” etc.

Our representative industry bodies have variously sought to clarify this specific subject, in that a numerical qualifier must be assigned in reliance upon an evidenced body of knowledge.

 The European Network of Forensic Science Institutes (ENFSI) Guideline clearly sets out the correct approach for experts;

“These probability assignments shall still be expressed by a number between 0 and 1 rather than by an undefined qualifier (such as frequent, rare, etc.). Such personal probability assignment is not arbitrary or speculative, but should be based on a body of knowledge that should be available for auditing and disclosure.”

The International Society of Forensic Genetics (ISFG) guideline also provides conclusivity on the matter;

 “The verbal scale is optional but cannot be used by itself. If it is used, then the numeric equivalents must also be available/provided. In

practise, one would provide first one’s likelihood ratio, then the verbal equivalent is applied afterwards”.

The application of unsupported verbal weighting before the courts is of particular concern, especially given the broad jurisdictional understanding as to the potential paucity of DNA transfer evidence. Many matters with DNA expert evidence before the courts are such that the question as to direct or indirect deposition is beyond our current scientific ability to evaluated in a robust way.

A jury may be readily persuaded as to a defendant’s culpability, when he is known to be in the area, has committed similar offences, and the prosecution DNA expert casually inserts at hearing, that the defendant “more likely” deposited the DNA directly.

In Australia, I have only seen 1 circumstance in which an activity level numerical ratio was evidenced at hearing. In this same period between 2009 and 2024, I have heard a multitude of unsupported verbal probabilities tendered orally in prosecution expert testimony.

At the time of writing, South Australia is the only laboratory with capability to calculate numerical likelihood ratios for complex activity level scenarios. This Hugin Expert software, utilises Bayesian Networks to apply mathematical algorithms in the production of numerical likelihood ratios. This

process requires the analyst to make extensive subjective determinations including a selection of robust data for calculation by the software.

South Australia began utilising this methodology to evidence in 2022.

As to comparable methodologies that the reader may be familiar, this software application to Bayesian Networks, is not at all analogous to the development or utilisation of STRmix.

This process of complex activity level numerical evaluation is in its scientific infancy.

Wethered et al stated in 2020 that “ It is submitted that the ranking of DNA transfer scenarios currently lacks robust scientific validation. If the central question of a case is ‘how did the DNA get there’, then courts need to be cautious” [2]

Notes for legal practitioners – 2024

Activity level is the next frontier of DNA expert evidence before the criminal justice system.

Stand-alone verbal weightings at activity level are contrary to guidelines, lack probative value, and should be rejected.

We expect to see increased occurrences of complex activity level reporting utilising Bayesian Networks before the courts.

The Justice system and practitioners will need to be provided with an understanding of the principles, which is in and of itself complicated by the reality that the process “currently lacks robust scientific validation” [2]

[1] Andrew Haesler (2019) – Dealing with DNA in court, its use and misuse

[2] Weathered, L., Wright, K., & Chaseling, J. (2020). Dealing with DNA Evidence in the Courtroom: A Plain English Review of Current Issues with Identification, Mixture and Activity Level Evidence. The Wrongful Conviction Law Review