Solicitors:
Director of Public Prosecutions (Crown)
Kidman Legal (Accused)
File Number(s): 2011/147183
[2]
Judgment
HER HONOUR: After delivering judgment on 28 July 2015 (R v Xie (No 18) [2015] NSWSC 2129), the Crown advised that when Mr Goetz was recalled to continue his evidence in chief before the jury he proposed to ask him the following question:
In relation to the similarities that you observed between the three crime scene samples and the sample from the garage floor of the accused's home, apart from the number of alleles, what else did you notice about them that was similar?
The Crown indicated that he anticipated that Mr Goetz would say in answer to that question that, in addition to the number of alleles and their respective peak heights, the relative proportions of the alleles in the mixed contributor DNA profiles in Items 550 and 616 at each of the 20 loci generated by application of PowerPlex 21 were also similar, and the "most similar" as between Item 550 and the various crime scene samples he reviewed.
The Crown also advised that he also proposed to ask Mr Goetz the following further question:
What do you say about the similarity between three profiles from the crime scene compared to the sample from the garage floor? What do you say about the similarity between them compared to the crime scene samples that you have observed, in your experience in the past, from the one crime scene?
Mr Turnbull's primary submission was that the further evidence the Crown proposed to call from Mr Goetz was covered by my ruling.
In support of that submission, Mr Turnbull placed particular emphasis on the emails that were exchanged between Mr Goetz and the ODPP, also extracted in the judgment, where Mr Goetz was asked by the Crown to address the very questions he proposed to ask when Mr Goetz's evidence resumed (set out above).
When addressing that question by return email, Mr Goetz said as follows:
To provide an answer for the questions requires some sort of analysis of casework far more than my general observations about casework samples. All I can state is that there is great deal of variability in mixtures that can occur within casework samples from the same case ranging from ones with similar appearance to ones where the contributor proportions of alleles to the mixtures are very different.
After reviewing my judgment in light of the further questions the Crown proposed to ask Mr Goetz, it seemed to me that, quite apart from whether the further evidence infringed my previous ruling, unless Mr Goetz had undertaken casework analysis sufficient to express an opinion about comparative similarity, the evidence the Crown sought to elicit from him would not satisfy the test of admissibility under s 79 of the Evidence Act 1995 (NSW), and would be excluded for that reason.
The Crown submitted that Mr Goetz's email had been misinterpreted and that what Mr Goetz was endeavouring to convey, as confirmed in conference with the Crown following receipt of the email, was that a DNA analyst would expect to see different proportions of DNA in mixed contributor samples found within the one crime scene because of the dynamics of what is commonly encountered in a bloodshed event, but that the degree of similarity he observed between Items 550 and 616 was more than what an analyst would normally expect to see where multiple samples from the same contributors at the one crime scene were analysed.
On the assumption that Mr Goetz's statement of opinion was based on relevant training, study or experience, the Crown submitted his evidence as to the similarity in the relative proportion of alleles as a feature of the similarity he observed was probative not only for the reasons advanced and discussed in the earlier judgment, but because the Crown proposed to submit that the degree of similarity from Item 616 to Item 550 referable to that additional feature adds further weight to proof of the fact that the DNA in the blood of the deceased was transferred from the crime scene to the garage, the children being, on the Crown case, the last to be killed.
Although the Crown proposed to confine Mr Goetz's evidence to his observation of this additional fact or feature of similarity between the two items as an "objective" fact or feature of the appearance of the profiles or their pattern, without him venturing any expression of opinion that the degree of similarity was "unusual", "significant" or "high", Mr Turnbull submitted that it was inherent in Dr Goetz's observation that there was a "marked" degree of similarity in the relative contributor proportions which he had not previously encountered but which he could not scientifically quantify, and which I have previously held was not an opinion that either Mr Goetz or Dr Walsh were entitled to express for that reason.
Ultimately it was agreed that the appropriate way to test whether Mr Goetz could give this evidence in a way that did not infringe my earlier ruling (and in a way that properly met the test for admission as opinion evidence under s 79) was to permit the Crown to call evidence from him on the voir dire.
Mr Turnbull sought advance notice of what further evidence Mr Goetz might give on the voir dire in explanation or elaboration of the evidence that was comprehensively tested in the interlocutory hearing before Johnson J, being the basis upon which I had also been invited to rule on the admissibility of his evidence in my earlier judgment. The voir dire was adjourned to allow that to occur.
Mr Goetz ultimately gave evidence on the voir dire that whilst mixed contributor samples are frequently encountered in a crime scene, an analyst does not "routinely look at them and say they're different or similar". The analyst's awareness of differences between samples is solely for the purpose of determining which of the profiles is identifiable and suitable for uploading to the database as a single source sample. He said "you cannot load mixtures to the database". He said he has probably seen mixtures of two, three, four or five contributors but could say nothing further than having encountered them in practice, having always assumed that differences in the appearance of mixed contributor samples was accounted for by the differing amounts of blood or other biological fluids shed by either the victim or victims and any suspect or suspects at a crime scene.
In his dealing with the mixed contributor samples in this case, Mr Goetz made it clear in his evidence on the voir dire that he only focused on Item 616 (being one crime scene sample) which he compared with the profile generated from Item 550. (For that reason alone the questions the Crown proposed to ask him set out above were inadmissible. So much was acknowledged by the Crown.)
Mr Goetz also gave evidence that in undertaking that comparative exercise, although Item 616 was a slightly weaker sample than the sample from the garage, all of the alleles that were present in Item 616 were also present in Item 550. That variability was explained in the following way:
A: ... The ones that were weaker in the sample from the garage weren't present in the sample from the mattress, so I thought that's possible, that it could have been diluted out and, therefore, you wouldn't see them. That's one possibility, of course.
Q: Can I just ask you to clarify that. Are you saying that the weaker parts of the profile in item 550 were the ones that were missing from item 616?
A: I'm saying ‑ I'm saying that the alleles that were not present in the sample from the mattress could have been not present because they were diluted out; in other words, a weaker sample, the weaker alleles may not be visible.
He then gave the following evidence:
Q: And, Mr Goetz, why was it that you chose to focus on 616?
A: 616 not only contained the alleles, but what I found interesting about it was the proportion of the alleles were similar at each locus between the sample from 616 and 550.
Q: Are those the peak heights that you're referring to?
A: I'm talking about the proportion of the peak heights. Generally, if you looked at that profile, anybody in the court could see it, the big peaks in ‑ at one locus were the big peaks in the other sample. The small peaks in one locus ‑ at one locus were also the small peaks at the other ‑ in the other sample at that same locus ‑ not exactly but‑‑
Q: So, there is a correspondence at the level of peak heights; is that a fair way of putting it?
A: There was a proportionality between them; because one was weaker than the other, you wouldn't expect the peak heights to be the same, but the proportion of contribution was the same. So, in other words, if you had three peaks present and, say, in one sample, there would be three peaks labelled "A", "B", "C", A was smaller, B was bigger, C was smaller, in the other sample the same was present.
…
Q: All right. Now, what do you say about the similarity that you observed between 616 and 550 compared to other cases where you have done DNA analysis on samples from the one crime scene?
A: This is probably why I found this interesting, in that I found two samples that, if they did not come from the crime scene, were unusually ‑ were similar. They exhibited the same mixture proportion and they exhibited the same alleles that were there, other than a few that were missing; but, of the stronger alleles, they were all the same and the proportions were very similar - were similar. Were they exact? No, they weren't, and I wouldn't expect them to be exact, but they, in the majority of loci, gave very similar proportions.
Q: What do you say about those similar proportions compared to testing that you have done in other cases where there have been multiple contributors but at the one crime scene?
A: Usually within the one case itself you don't get samples that will exhibit a high degree of similarity, so that the proportions are similar. You will normally get variations. These two were unusual, in that there was that degree of similarity.
Q: So, are you saying that there is a higher degree of similarity in this case than you generally get in other crime scenes where there are samples taken from the one crime scene?
A: I'm saying this is an unusual event. What we've got here is a sample taken from the garage that, if it did not come from the crime scene, they're matched by chance. The chance match is made more unusual because of the proportions than simply the alleles being present there.
…
A: I can't put a number ‑ I can't put a number on this similarity. All I can say is, from my experience, you tend not to see similar patterned profiles from the one case. Here we have potentially two different samples from two different scenes that makes it, to my way of thinking, an unusual event.
[Emphasis added]
After Mr Goetz had given evidence in chief, Mr Turnbull sought the opportunity to reflect upon his evidence and take advice from those advising him before he cross-examined. Mr Goetz returned the following day.
In advance of Mr Goetz returning to be cross-examined, the Crown further refined the questions he wished to ask Mr Goetz in the trial by focusing on the similarity Mr Goetz noted in the "proportions" of DNA in the profiles generated from Items 616 and 550, with a view to having him elaborate on what the Crown described as an objective fact or feature of similarity.
In essence, the Crown indicated he wished to elicit from Mr Goetz that there was not simply a similarity in the number of alleles and their relative peak heights (evidence he and Mr Walton have already given before the jury without objection), but a "correspondence" or a "similarity" in proportions of alleles found at each loci in each sample as an additional and discrete feature of the objective fact of similarity between the two mixed contributor profiles generated from the two items. The Crown submitted that this was not to elicit from him an opinion as to extent or degree of similarity between the two samples in any qualitative sense, as would offend my ruling, but a further feature of the objective fact of similarity in the same way that allele numbers and their peak heights are objectively similar.
To exemplify that observation, the Crown had Mr Goetz prepare a histographic representation in percentile terms of the proportions of alleles at each locus in each of the two samples. (Item 616B, being the more complete profile that was generated by Mr Walton, generating 63 alleles, was relied upon for comparative purposes as distinct from Item 616A, which generated 61 alleles.) The Crown intends to tender the histograph to further or better exemplify the proposition that the vast preponderance of alleles are in proportion with each other at each of the 20 loci.
When he was recalled, Mr Goetz gave further evidence in chief on the voir dire that his preparation of the histograph was based on the results of Mr Walton's DNA testing and, in particular, the relative allele peak heights in two samples. He said that there was no subjective analysis involved in his reading of the relative peak heights or assessing the relative proportions of the different alleles at each locus, or in the production of the histographs. That said, he accepted in cross-examination that what is rendered pictorially in the histograph is the "degree of similarity between the two profiles", which he has repeatedly described as "unusual". He went on to say that he regarded the degree of similarity as "dramatic" and, that being the case, he accepted it necessarily raises the question as to why that degree of similarity presented. He also agreed in cross-examination that the pictorial representation or the diagrammatic representation reflects his interpretation of both similarity and, to an extent, some features of dissimilarity, between the two samples.
In further elaborating on the extent to which the histograph was informed by his subjective perspective, he said in answer to a question I asked:
Q: The pictorial representation is informed by your subjectivity, do you say, and how is that so?
A: What we are dealing with here is something unusual, your Honour. It was picked up in a laboratory, which probably normally wouldn't have been picked up in most situations because you normally don't compare two samples like we did in this case. For whatever reason, we found it and I suppose extrapolated that there was a similarity between these two samples that needs to be explained. So because I cannot put a number on the degree of dissimilarity, like I can for a billion, for a profile, I believe I am no longer giving an objective result, I am more using my experience and therefore there is more subjectivity in what I am reporting.
He went on to say in the balance of that particular answer:
...
A: However, when you put the numbers in there, certainly I think you can diagrammatically see there's similarity there. I don't [sic] believe anybody can see that. What that actually means and what number I can put on it, I can't do that.
In cross-examination Mr Turnbull invited Mr Goetz to restate, in accordance with his evidence before Johnson J, that he is unable to quantify the significance of the similarities between the two profiles or, and in that exercise, to attach any mathematical weighting to the dissimilarities that were revealed. He also agreed that he had not, on any other occasion, been invited to, nor had he expressed an opinion about comparative similarities (or dissimilarities) between DNA mixtures involving multiple contributors, and that the normal course when called upon to express an opinion of that kind is to test a given hypothesis in a validated and disciplined scientific paradigm. I then asked the following question:
Q: Mr Goetz, do you consider, with those concessions that you have just made, that the pictorial representation is meaningful or meaningless ‑ perhaps I should extend the question ‑ in representing pictorially the fact of there being a similarity in the proportion of alleles found in the two samples?
A: Your Honour, as I've probably said a number of times, what we've got here is something unusual. How do you explain something unusual that doesn't fit within a box, a nice confined box? That's what we've got here. We've got a mixture that was found in a garage that, if it had been found at the murder scene, nobody would have cared about because it looked like it came from the murder scene. Now, we've got something in the garage that, if it doesn't come from the murder scene, it matches by chance.*
Now, the difficulty is trying to equate that chance. I've done it by showing that there is a similarity between the proportions of the different alleles at one scene to the mattress ‑ sorry, the garage to the mattress.
Does it show something? In my opinion it does show something. To what level I would go? I don't know, your Honour.
In identifying the "something" that is revealed in the comparative exercise Mr Goetz was invited to address, he went on to say there is a similarity, the degree to which he cannot expand upon and an explanation for which he is unaware.
In the result, he said "the most I would go to is that there is similarity between the two samples". He was at all times concerned to emphasise that he does not mean to suggest the profiles are "the same" and also concerned, given what he understood to be the constraints to which he was subject by reason of my ruling, to steer clear of utilising any comparative expressions such as "match", or "match by chance", or qualitative expressions of the degree of similarity.
Mr Turnbull submitted that although it can be said that the histograph renders, in a pictorial way, the relative proportions of the different alleles at each locus in each sample, and even accepting that did not involve any subjective assessment or interpretation of the results of Mr Walton's DNA profiling, the histograph does not simply "describe" the similarity between the two profiles and, in that sense, it is not "value-free". He submitted both the graph and Mr Goetz's evidence, even if only led from him vive voce, necessarily calls for an explanation of features of dissimilarity, which in turn brings into play questions of degree. Mr Turnbull submitted that it is the existence of a tension between similarity and dissimilarity which is revealed in the histograph itself, a tension which Mr Goetz has said (and repeatedly) he cannot resolve or describe "scientifically", that breaches my previous ruling, even if it does not do so in terms.
Whilst there may be nothing more than a linguistic nuance between the objective fact of a similarity in appearance or pattern between the DNA profiles generated from the two items, and the degree to which they are similar or dissimilar involving qualitative reasoning, I am satisfied that to admit the further evidence of Mr Goetz runs the real risk, against which directions cannot protect, of a jury misusing both Mr Goetz's evidence of a similarity in proportions of alleles and a representation of that feature of similarity pictorially in a way which will be productive of unfairness to the accused. The risk of a lay jury being unable to discern the difference between similarity of this highly esoteric kind and the DNA profiles from Items 550 and 616 being for all practical purposes "the same", is amplified by imposing an unfair forensic challenge on Mr Turnbull as the cross-examiner to confine the testing of Mr Goetz to an admissible opinion as to the fact of similarity in circumstances where even he is at a loss to explain it without introducing his view that it was the degree of similarity that was "unusual" which is, as I see it, simply another way of remarking upon its significance.
Although the additional feature of similarity which has been the subject of further testing on the voir dire has probative value (perhaps even considerable probative value if it were subjected to a statistical weighting) being, on one view of it, an objective feature of similarity in the two profiles along with allele numbers and peak proportions, it is not evidence in that same category. Even imposing upon Mr Goetz the discipline of not descending into a commentary on the extent or degree of similarity, or a range of possible explanations for it, there is, in my view, a danger of unfair prejudice of the kind that I have identified and, accordingly, the further evidence the Crown proposes to call from Mr Goetz is disallowed.
[3]
DISCLAIMER - Every effort has been made to comply with suppression orders or statutory provisions prohibiting publication that may apply to this judgment or decision. The onus remains on any person using material in the judgment or decision to ensure that the intended use of that material does not breach any such order or provision. Further enquiries may be directed to the Registry of the Court or Tribunal in which it was generated.
Decision last updated: 01 March 2017