Gibson v R [2001] TASSC 59

Supreme Court of Tasmania|2001-05-25|Before: Evans JJ, Underwood J, Slicer J

CITATION: Gibson v R [2001] TASSC 59

Criminal Law - Evidence - Evidentiary matters relating to witnesses and accused persons - Identification evidence - Circumstantial evidence - DNA profiling.

Pantoja v R (1996) 88 A Crim R 554, considered.

Jeffrey v R [1991] TASSC 114; [1991] Tas R 336, referred to.

Judgment Number: [2001] TASSC 59

1 The appellant was found guilty of the crimes of aggravated assault, rape, indecent assault and aggravated sexual assault. He was convicted and sentenced to six years' imprisonment. He has appealed against those convictions.

2 The Crown case was that at about midnight, or shortly afterwards, on 25 April 1999, the complainant was walking up Henry Street in Launceston. She was then aged 15 years. She had been in town with friends. She met a man who was a stranger to her, walking in the opposite direction. She asked him if Ravenswood was the same as Waverley and he told her it was not and that she needed to go in the opposite direction. He said that he would walk with her. As they were walking along together, the man suddenly hit the complainant in the stomach and winded her. He then dragged her to a bank on the side of the road and sexually assaulted her. It is unnecessary to set out the details of the assault. It suffices to say that he committed the crimes charged in the indictment, one of which was rape by penetration of the complainant's mouth. The complainant's evidence was that this man ejaculated in her mouth. Shortly after this, the man let the complainant go. She ran to a nearby house and raised one of the occupants. The police attended.

3 The complainant gave the jury a description of her attacker. She said that he was aged "about 40 odd", with medium length, brown, curly hair. She said that he had a goatee beard and a moustache. She described his build and height as medium. The complainant said he had small, light coloured eyes. She also gave a brief description of the clothes that her attacker was wearing that night.

4 The appellant, who represented himself at the trial and on the appeal, did not challenge the evidence of the complainant. The issue at trial was the identity of the assailant.

5 The complainant was examined by a medical practitioner at 3am that morning. Amongst other things, the medical practitioner took a swab from inside the complainant's mouth. A forensic scientist detected two spermatozoa in the swab. They were sent for DNA testing, along with a sample of blood later taken from the appellant.

6 The Crown case was, in very large measure, dependent upon the DNA evidence. Apart from the description given by the complainant, the only other evidence of the identity of her assailant was as follows:

(1) Evidence from a witness who happened to drive past the scene of the crimes at about 12.15am. He said that he saw a man in the street in which the attack took place. He said that the man was hitchhiking in one direction, and some distance away, he saw a girl walking in the opposite direction. His description of this man was vague and, at its highest, could only be described as not inconsistent with the physical features of the accused.

(2) Three months after the attack, the police asked the complainant if there was a photograph of the man who attacked her amongst a number of photographs that they showed her. She identified a photograph of the appellant as being her assailant, but said that she was no more than 70 per cent sure that it was him.

7 Mr Szabo, forensic scientist, gave evidence of his DNA profiling. He explained to the jury that he made DNA profiles of a sample of blood taken from the appellant, from a sample of blood taken from the complainant and (inter alia) of the spermatozoa on the swab taken from the complainant's mouth.

8 In accordance with what Mr Szabo called "standard" procedures, he profiled the DNA at 10 different locations on the genetic chain. He described the procedure as, in effect, measuring the length of the DNA at these 10 locations and converting the results into numbers. The profiles were reduced to writing in the form of a graph and ultimately in the form of a table of numbers. The documents became exhibits.

9 Mr Szabo explained that the critical information was the presence or otherwise of two peaks at each location - one inherited from each parent. At one of the 10 locations, Mr Szabo was only able to identify one peak to which he attributed the number 8. He said that it was possible that there was another peak at 10, but there was insufficient material to be sure of that.

10 Mr Szabo said that a comparison of the DNA profile of the blood of the appellant with that of the swabbed material, showed that each was consistent with the other and that "the chance of a second person unrelated to [the appellant] having the same DNA profile is less than one in a hundred million".

11 As part of its case, but I infer at the request of the appellant, the Crown led evidence from a Ms P J Scott, who is also a forensic scientist. She tendered into evidence the written results of her DNA profiling of what the appellant admitted was his blood and his spermatozoa. The format of her table of results was exactly the same as the format of Mr Szabo's table. Apart from a single exception, Ms Scott's profiling of both the appellant's blood and spermatozoa at all 10 locations, was identical to Mr Szabo's profiling of the appellant's blood and the spermatozoa on the swab. The single exception was at a location identified as "D7". At that location, Ms Scott's profiling of the appellant's blood and spermatozoa is recorded as "8,10". At the same location, Mr Szabo's profile of the appellant's blood is also recorded as "8,10" but the profile for the swabbed spermatozoa at that location recorded only one peak, identified by the numeral "8".

12 The appellant cross-examined Mr Szabo at some length and with considerable skill, about this difference in an attempt to show that it raised a doubt about Mr Szabo's ultimate opinion. Mr Szabo explained that the only difference between his tests and those of Ms Scott's was that Ms Scott's tests amounted to a full DNA profile, but his tests on the oral swab amounted to a "close to full DNA profile". He went on to explain with respect to his tests that at location D7, there was insufficient material to clearly disclose one of the two peaks. He said that there was a suggestion of a peak at 10, but it was not strong enough to be accorded scientific recognition and, accordingly, it was discarded. As mentioned, he considered that the absence was due to insufficient DNA material. In all other respects, the tests were identical. Mr Szabo said that his findings and those of Ms Scott were identical at all the other locations, were not inconsistent with one another at location D7, and the absence of one peak at that location in his tests, due to insufficiency of DNA, did not alter his ultimate opinion.

13 This point was very clearly put to the jury in the following passages taken from the learned trial judge's summing up:

"[Mr Szabo's] evidence was that when these tests are conducted they are looking for two peaks. One, that we have inherited from our father and one we have inherited from our mother. Usually it would seem that the peaks occur at different points, like 14,18; 14,15 and things like that. But sometimes both peaks are at the same point because the mother's and the father's are the same, they both had blue eyes or something like this. But for D7 Mr Szabo said he could only get one peak 8, so he did not report a 8,8 he just reported a single 8 and it was his opinion that the reason for that was that it was not a full result. He probably had a low quantity of DNA at that point he said and of course the little peak at 10 that was there on the computer print-out was just not sufficient to be able to claim that it was reportable.

I asked him what difference would it have made if it has been an 8, 10 there rather than just an 8, you reported an 8 what if you had actually found the other peak which was reportable at 10 also. He had given evidence that in his opinion the chances of somebody unrelated to the accused having the same DNA profile as he found on that spermatozoa I think he said less than one in one hundred million. And I said well what difference would it have made if you had found the 10, and he said well no difference, he would have reported the same thing. The chances of it would be less than one in one hundred million, they never report anything, they do not go to less than one in one hundred billion, it is only one in one hundred billion is the furthest that they have ever gone. So that is the evidence that was given about that."

14 By his notice of appeal, the appellant seeks an order that the conviction be quashed upon the following grounds:

"1) Said Judge failed to instruct jury that the statistic Mr Szabo arrived at has no legal precedent in Australia. [original emphasis]

Said Judge erred in instructing jury that said method is accepted thru Tasmania Australia and other countries.

Said Judge erred when instructing jury ID is not obtained in courtroom. Most witnesses have been asked to ID accused if in court.

Denied opportunity to produce evidence to dispute witness testimony."

15 No error occurred in the failure of the learned trial judge to instruct the jury that the statistic Mr Szabo arrived at had no legal precedent in Australia. There was no evidence to that effect. The evidence of Mr Szabo with respect to how he arrived at his opinion that the chance of the spermatozoa on the swab having come from a second person unrelated to the appellant, was less than one in one hundred million, was correctly and completely summarised by the learned trial judge in the following passages in his summing up:

"The first question for you when you come to consider this evidence is whether you are satisfied that the evidence establishes beyond reasonable doubt that there was a match between the DNA profile of the spermatozoa and the DNA profile of the accused's blood. Now it seems to me that if you are not satisfied beyond reasonable that the two DNA profiles matched then you would have to exclude this evidence. The most important aspect of DNA evidence is that it can exclude a person. If the profiles do not match then the accused is excluded. It is not possible that he could have been assailant if the profiles do not match. Okay. Now if the profiles match it does not necessarily mean he must have been the assailant. If you accept the evidence of the scientist well the chances of somebody else though unrelated to the accused is less than one in one hundred million. But nevertheless it is not an absolute certainty. So understand that. The scientist does not come in here and say I've got this result therefore it must have been the accused who produced that spermatozoa. He is not saying that. The scientist is saying, and Mr Szabo is saying in my opinion they match, the chances of it not being the accused spermatozoa is less than one in one hundred million so far as somebody unrelated to him being the source of it. Okay so that is all he is saying.

And, of course, to support the likelihood that it was the accused you have got the other evidence of the photoboard identification and the descriptions of the man coming down the hill and things of that nature.

Now Mr Szabo's evidence was then that there was a less than a one hundred million chance that the spermatozoa came from someone unrelated to the accused. Why does he say that? Well he says he says that because they have on data base 200 samples of other people's blood and what they have done is they have DNA profiled those 200 samples. And so they have carried out the first one, the D3 test with respect to those 200 samples they have got. And they have looked at where the peaks are and they have been able to see well how many peak at 14, how many peak at 18 and then they have conducted the next test the VWA one apparently and they look for where the peaks are. And they find out by doing 200 how many people's coincide at the same points. And then using that they are able to come up and calculate this less, in this case less than one in a hundred million chance that he talked about. He also gave evidence that there are data bases throughout Australia and throughout the world and basically they all come up with the same result so far as white Caucasians are concerned.

So it is not just the Tasmanian laboratory that comes up with these kind of results, they are consistent he said throughout Australia and the world so far as white Caucasians are concerned. So I suppose the evidence of Mr Szabo therefore suggests that in Australia, which has a population of less than 20 million, it is unlikely that anyone else unrelated to the accused would have a DNA profile identical to him. But not absolutely certain, unlikely, but it is possible that somebody else might. And the scientist does not exclude the possibility that there might be. He can only give what the scientific likelihood or unlikelihood of it is.

There was also some evidence from him which went to the question well might it have been somebody related to him. And the chances there are different. Identical twins would be identical. Well there is no suggestion that Mr Gibson has got an identical twin, but their profile would be the same. If it was your father or your son the chances were not said to be less than one in a hundred million, they are one in one million. If it is a son and a father as I understood his evidence. If it was your brother or sister, well brother, it would not be sister because it would show up the sex, but if it was your brother the chances are of one in six thousand. Well there is no suggestion that Mr Gibson has got a brother and he might have been around at that time. But that was the evidence. So your DNA profile with your brother is you are a male or with your sister if you are a female there is about a one in six thousand chance, or less than a one in six thousand chance that your DNA profiles would be the same, unless you were identical twins."

16 It was a question of fact for the jury to determine. As the learned trial judge put to the jury, it was entirely up to them whether they accepted the evidence of Mr Szabo or not. He told them that they were to assess his evidence in the same way as they assessed the evidence given by any other witness.

17 During the hearing of the appeal, the appellant pressed on the Court two letters written by Dr Atchison, a forensic scientist at the Victorian Institute of Forensic Medicine. Although one of these letters is undated, it was clearly written prior to the trial. The other was obviously written after the trial, as it is dated 8 January 2001, and expressed to be in response to the appellant's letter to him dated 4 January 2001.

18 Although it is clear that mere presentation of letters to an appeal court is not evidence, and that there are strict rules with respect to the reception of fresh evidence after the trial has concluded, the Court had regard to the letters with the consent of counsel for the Crown as the appellant did not have the benefit of counsel.

19 In the first letter, Dr Atchison stated that he does not dispute the DNA results obtained by Mr Szabo, and then wrote:

"On the assumption that the DNA results are accepted, the usual question posed by a layperson is 'How many males are there is [sic] a population (eg, Tasmania) who have the same DNA type as the accused?' In fact, this is a difficult question to answer. It is important that a jury not confuse the figure presented by Mr Szabo with this question. I enclose a detailed explanation of the different ways the 'statistics' can be approached, but I would make the following specific points."

There follow two pages of material which could have provided the appellant with material for the cross-examination of Mr Szabo and which, I suspect from reading the transcript of evidence, did, in fact, provide material for such cross-examination. The letter concludes with a view that Mr Szabo's ultimate opinion can be accepted, provided the assumptions he made, eg, the Tasmanian database, is representative of the ethnic/racial background of the appellant and that a close relative of the appellant was not involved.

20 The learned trial judge dealt with these matters in his charge to the jury as can be seen from the last paragraph of the passages taken from the summing up and set out earlier. There is nothing in the first letter to support the proposition that some error occurred at the trial which should lead to the conviction being quashed.

21 The other letter does not assist the appellant either. Dr Atchison commenced with the disclaimer that without seeing the test results, he cannot comment on specific discrepancies between different cases. The letter then provides a short discourse on how differences between DNA profiling can arise, such as mix-up of samples, error in reading the results, and so on. The letter does not purport to assert that any error occurred in this case and discloses no basis for upsetting the convictions.

22 Ground 2 of the notice of appeal asserts that error occurred in instructing the jury that Mr Szabo's method is accepted "thru Tasmania Australia and other countries".

23 I have already set out the learned trial judge's instruction to the jury with respect to that. It is true that the learned trial judge said to the jury:

"So it is not just the Tasmanian laboratory that comes up with these kind of results, they are consistent he said throughout Australia and the world so far as white Caucasians are concerned."

24 However, when that sentence is read in its context, it is perfectly clear that his Honour was, quite correctly, summarising the effect of Mr Szabo's evidence. He did not instruct the jury that any method is accepted in any place. He reminded the jury that Mr Szabo gave evidence that there are databases throughout Australia and the world which come up with the same results so far as white Caucasians are concerned. There is no substance in ground 2.

25 Ground 3 asserts error in instructing the jury that identification is not obtained in the courtroom. Even if it were made out, this ground would not result in the convictions being quashed. However, it is not made out. It arises out of something that the learned trial judge said to the jury in the course of giving them a thorough and complete direction with respect to the danger of accepting the complainant's identity evidence based on the photographs that she saw. He told them that ordinarily complainants are not allowed to give dock identifications because there is a very real risk that such evidence may be erroneous. That was a correct statement of the law. No dock identification was given in this case. There is no substance in this ground.

26 The last ground asserts that the appellant was denied the opportunity to produce evidence to dispute witness testimony. There are two aspects to this ground. The first relates to the material contained in Dr Atchison's letter, or rather, more accurately, his first letter, for the second did not exist at the time of trial. At the conclusion of the Crown case and in the absence of the jury, the learned trial judge gave the appellant a very careful and detailed exposition with respect to his right to give evidence or not, as he chose, and his right to call witnesses or not, as he chose. Following that exposition, there was some exchange between the appellant and the learned trial judge which makes it abundantly clear that the appellant understood that if he wished to do so, he could call any evidence he liked that was relevant to the issues. The exchange concluded with the appellant telling the learned trial judge that he understood his position. At that stage, the trial was adjourned to the next morning. When the trial resumed the following day, the appellant said "No" to the learned trial judge's question, "Do you have any other witnesses you wish to call?"

27 The other aspect of this ground arises out of some evidence-in-chief given by the appellant. During the course of that evidence, he referred to the evidence of the complainant that at one stage during the sexual attack, the complainant's assailant asked her to rub his penis. The complainant said that she complied with this demand and at his insistence, used two hands to do so. The appellant's evidence was that this was an impossibility because he was "not adequately endowed". The appellant then asked if he could produce his penis to the Court to prove that fact. There followed an exchange between the learned trial judge, senior counsel for the Crown and the appellant, during the course of which counsel for the Crown said he would not take any issue with the claim that the appellant made. There is no substance in ground 4.

29 The appellant was convicted of the crimes of rape, aggravated and indecent assault. He represented himself, both at trial and on the hearing of this appeal.

30 In the early hours of 25 April 1999, the complainant, a young woman aged 15, approached a man in Henry Street, Launceston and asked for directions. The man offered to walk with her and told her he was going to Mayfield to see his former girlfriend. Shortly thereafter, the complainant was sexually assaulted and raped by that person. Afterwards she ran to a nearby house from which the police were called. There was no issue that the conduct perpetrated on her constituted the crimes of rape, indecent and aggravated assault.

31 The issue at trial was the identity of the assailant. The evidence of identification can be summarised as:

(1) The description by a motorist who saw a male in Henry Street at the relevant time who:

"was about 40-odd years of age or a little bit older, about five foot nine, stocky build, receding hairline"

The complainant was wearing a mid-length coat with a fur collar at the time of the assault. The learned trial judge advised the jury that they might use the evidence:

"... for two purposes. One, I suppose to support the girl's evidence that she did meet a man as she was walking up the hill, and also you might use it for the purposes of identification. It is a critical question of identification in this case. Well there is a description of a man given by Mr Foran, you may use it as you like, I mean it is not that the accused is definitely that man, but you may say well Mr Foran has described somebody, it might be the accused, it might be consistent with his description. So that is a piece of evidence you can use, it is obviously not conclusive, but something you can use."

(2) The observation of a police officer that on a subsequent occasion he had seen the appellant walking in Henry Street.

(3) The description of the assailant given by the complainant at trial which in the summary given by the learned trial judge was:

"... dark brown hair, I think, was the way she expressed it. Medium length, curly, goatee and moustache, widish face, eyes small, lightish eyes I think she said, 40 years old, or 40 odd years old, medium height and build."

(4) Evidence of the complainant and a police officer of the identification of the appellant as the assailant by means of a photo-board process, some three months after the attack. The jury were given thorough and unexceptional directions in relation to the permitted use and danger of identification evidence, especially in the circumstances of this case.

(5) Evidence of the comparison of body samples taken from the appellant with bodily fluids located on the body of the complainant (DNA testing) and opinion evidence as to the implications of that comparison.

(6) Evidence given by the appellant at trial that he was neither the assailant nor present near the scene of the attack at the relevant time.

32 The evidence of DNA testing is the main issue raised on this appeal. The grounds of appeal claim error in that the learned trial judge:

"(1) ... failed to instruct the jury that the statistic(s) Mr Szabo arrived at has no legal precedent in Australia.

(2) ... erred in instructing jury that said method is accepted thru [sic] Tas Australia and other countries.

(3) ... erred when instructing jury ID is not obtained in courtroom. Most witnesses have been asked to ID accused if in court.

(4) Denied opportunity to produce evidence to dispute witness testimony."

33 No attempt was made by the prosecution to lead evidence of a dock identification. Under the circumstances of this case, any attempt to do so would have been highly prejudicial to the appellant. Dock identification is fraught with danger unless the witness knows the accused or the circumstances of identification such as prolonged presence or apprehension at the scene render the evidence safe (R v Wright A102/1991). The appellant had declined to accompany police officers to a police station for interview and no "line-up" was conducted. In his charge to the jury, the learned trial judge advised:

"And he [the appellant] also said that he was not identified by (S) in the sense that she looked him in the eye and identified him. Well the fact of the matter is, of course, that there was not a line-up and she did not do that and I have already made some mention to you about that. The law prefers line-ups to photoboard identification. But I should perhaps just add that it would not have been proper if she had have been asked here to come into Court and look at him and say is that the man because the law regards that as a most dangerous form of identification. What else would you expect her to say but yes. Once you get in a Courtroom situation she has picked him out in the photographs, she is hardly likely to say no is she, so we do not normally allow a witness to make an identification in the Courtroom. We call it a dock identification. And it is regarded as a highly dangerous means of identifying.

The effect of the warnings given by the learned trial judge can be summarised in his advice to the jury:

"So I request that you take care when considering her identification from the photoboard and in fact I say to you that I consider that you should not find Mr Gibson guilty upon the basis of that identification and that you would, if you found him guilty, you would have to rest your verdict largely on the DNA evidence. That is the critical evidence, it seems to me, well one of the most important pieces of evidence in this case and you could not find him guilty simply on the basis of that photoboard identification."

34 The direction was appropriate and the absence of any attempt to obtain a dock identification was advantageous to the appellant. The ground is not made out.

35 The appellant represented himself at trial. At the conclusion of the Crown case, the formal question of election was put to him which was followed by the exchange with the learned trial judge:

"HIS HONOUR: ... Do you want to give evidence yourself?

HIS HONOUR: Do you have any other witnesses you wish to call?

36 The appellant was not precluded from adducing evidence. His complaint is that he was inhibited from making use of a document which contained commentary and opinion on the validity of DNA testing in his cross-examination of a forensic witness, Dr Szabo. The document was shown to the witness and questions asked about its contents, but was not tendered. In his closing address to the jury, the appellant attempted to make a statement in relation to the reliability of DNA evidence but, as the following exchange with the learned trial judge indicates, was prevented from so doing:

"MR GIBSON: ...DNA has to be 100 per cent. It's not in this. There's too many irregularities. Tasmania has only got a data base of 201. The DNA Mr Szabo gave was in some case guesswork, an assumption as he says in here on page 155. Have to be extremely dangerous to convict on guesswork. I've read and watched TV over the last six months on DNA and whatever and the head of Scotland's Forensic Department, Superintendent Naper, has done similar DNA and has been successful -

HIS HONOUR: Well excuse me a moment I can't allow you to give evidence.

HIS HONOUR: All right. You've got to put your arguments on the basis of the evidence the jury has before it in Court.

MR GIBSON: I just started to deal with DNA data bases and whatnot, sir.

HIS HONOUR: Yes well I am afraid you can't give evidence from where you are. I mean if you had some material it should have been put to the scientists who have given evidence or you could have called a witness to give evidence for you. But you can't give evidence from where you are now, I am sorry.

37 In order to explain the basis of the appellant's complaint, this Court was shown, without objection by Crown counsel, two documents which the appellant claimed showed both the material he was attempting to use and which supported his argument in relation to grounds 1 and 2 of the notice of appeal. He sought to tender the documents as fresh evidence. Given the conclusion which I have reached in relation to those grounds, it is not necessary to determine whether they should be so received. I will accept that the two documents replicate in general terms the material contained in the document shown to the forensic witness and, as such, form a basis for criticism of the conclusion reached by that witness at trial. They will be given full effect on the basis that they comprised the basis of cross-examination of a witness in order to determine whether the material substantiates a claim that the opinion evidence led to a miscarriage of justice.

38 Insofar as the material relates to this ground, it has no import. The appellant was not prevented from using the material in his cross-examination of Mr Szabo, the forensic witness, as the following exchange indicates:

"MR GIBSON: ... I am going to read you something, you can disagree with it, it's from Mr Bentleigh, whatever his name is, from Monash, one of the professors over there done another check:

'After DNA testing, a scientist may conclude that the DNA profile of a crime scene sample is the same as that of the accused. This in itself has little evidentiary value unless a jury can place some weight on the evidence of a matching profile. For example, he - Professor whatever-his-name- is-Bentleigh, I forget his last name.

WITNESS: It's Dr Bentleigh Atcheson who works at the Victorian Institute of Forensic Medicine.

MR GIBSON: (Resuming) 'For example, he can do hundreds of DNA tests and always obtain a match.'

Do you agree with that? ... Sorry, can you read that again?

For example he can do hundreds of DNA tests and always obtain a match? ... I'm sorry I don't know understand what that means in the context. Who's he?

HIS HONOUR: Excuse me, would you like him to have a look at it?

MR GIBSON: Where it's underlined, Dr whatever from the Monash University Forensic thing. First paragraph? ... Okay what you didn't read out was the sentence after that which is also underlined, this is because everyone in the population has the DNA type for these tests.

HIS HONOUR: No, no please, don't interrupt him? ... What Dr Atcheson is saying here and it's a little dangerous paraphrasing what someone else - my guess is that what his intent is in this particular paragraph from what Mr Gibson has read out and from what I can see underlined is, he's saying what I said to you before, that just because DNA profiles match, that in itself, is not very useful information unless the weight to give to that match, so that's the evidentiary value, so the value of that depends on how common the DNA profile is in the population and what I believe Dr Atcheson is saying is that if DNA tests aren't particularly good, don't discriminate very well between individuals, it's a bit like blood group O which is present in half the population, if that's all the DNA test could do, then you could get lots of people matching by accident and this is this point of, for example I could do hundreds of DNA tests and always obtain a match in profiles if the DNA test results are very common.

MR GIBSON: (Resuming) So with the limited number of people that you've got, with that one and the (S) one, you've got a match, you can conclude what Dr Atcheson said? It could have been anybody? ... Well the chance of it being someone else, the chance of it being - what we have is a situation here where we have a DNA profile on this oral swab that's not from (S) and there's male DNA there and we're surmising that it's from the seminal material on this oral swab so we have an oral swab from a woman, it has male DNA in it, it has the DNA profile that we've got, that DNA profile matches the DNA profile of Mr Gibson's so there are a couple of explanations for this. One, that it's his DNA or it's not his DNA at all, it's someone else's DNA. Now that one in 100,000,000 - (new tape) - and one in a hundred million is the chance that it is a second person, unrelated to Mr Gibson who also has a DNA Profile that would match that profile. So if you like it is the chance of going out onto the street of Launceston or Hobart or anywhere in Tasmania and plucking someone out who also has that same DNA Profile, charging them with this offence and putting them on trial and, you know, would they have the same DNA Profile - the chance of that happening, them having that same DNA Profile as that on the oral swabs is less than one is a hundred million. That is for someone unrelated to Mr Gibson. And we have talked about brothers and relatives.

MR GIBSON: You said earlier that you diluted some of the material, used too much of the dilution and lost some, would that have an effect on the outcome as well on the specimen that you made, the one you done? ... I was asked a question about why it was that maybe I didn't get all of the DNA information from the - from Item 7M and my guess is that because it was below the reporting threshold that is an indication that there were low DNA levels at that point. So there were, you know, relatively low amounts of DNA. So what probably happened - this is my guess based on my experience within a laboratory - is that in taking quite a reasonable amount of DNA and diluting it down, taking a smaller amount of that and testing it, I perhaps went a little bit too far and so one of the pieces of information from the oral swab fell off, if you like. Didn't reach the reporting threshold. So that is my explanation for why it is that we didn't have complete information. And, again, that is quite common with DNA tests. I have got, well, the amount of DNA that you will have in each test will vary. So sometimes you have too much, sometimes you have less than you would like, sometimes you have got no control over it and you just have very, very little DNA and you get very, very fragmented information.

And you say that is your guess in this part? ... If I am asked to explain why I didn't get a full result from the oral swab -

It is a very dangerous guess, sir? ... It is an intelligent, sir. It is based on experience. I am not just plucking it out of thin air, it is -

Well, you just said it was a guess? ... Well, I can't -

General assumption, very - ? ... I can't say why it happened, but that is my explanation.

That is dangerous, sir. And extremely dangerous when you have got two different read-outs, regarding two different sperm samples. Three different sperm read-outs. Your D5, your D7's, in both cases. And this is just on a guess. I have got no -

HIS HONOUR: Well, is it a guess, was that a question?

39 The appellant was not precluded from making use of the material by way of cross-examination. He was, quite properly, prevented from giving evidence during his closing speech and could not have given opinion evidence about DNA in his own evidence-in-chief since he was not qualified to do so. Ground 4 cannot be sustained.

40 On 25 April 1999, eight samples were taken from the body of the complainant and analysed at the laboratory operated by the Forensic Science Service, Tasmania. On 30 March 2000, a sample of blood was taken from the appellant and submitted for analysis. Spermatozoa was identified from a single swab taken from the complainant and its genetic characteristics compared with those possessed by the appellant. Dr Szabo had explained to the jury the basis of a DNA profile and the process of matching that profile with that of a particular individual and a representative body of the population at large. A copy of the DNA printout was tendered at the trial. That printout showed a correspondence of 16 peaks from the sample taken from the complainant with those from the blood taken from the appellant. A seventeenth peak was excluded from calculation because it:

"... was below our reporting threshold, so we didn't report that."

41 The results formed cogent evidence in the case against the appellant. Mr Szabo then interpreted the conclusions which could be drawn from the matching of genetic characteristics or markers and it is this evidence which is the subject of the appeal. Mr Szabo told the jury that he then compared the result with that obtained from a database of 200 individuals in Tasmania which, in turn, was checked with other statistical studies conducted within the Australian population as a whole. It is the interpretation which is the subject of challenge by the appellant The interpretation by, and opinion of, Mr Szabo, were stated in the following terms:

"So, for instance, lets say you had the suspect matching that DNA Profile on the t-shirt, if one in ten people in Launceston would also match that DNA Profile of the blood, well, okay you have got someone who matches that and maybe that is the wrong person. I mean, one in ten, quite a large number of people in Launceston would also have that same set of characteristics. If on the other hand the chance of going down the street and finding someone with that same DNA Profile is much less, like less than one in a hundred million, then that information becomes potentially much more useful in terms of going towards identification. So in this particular instance, we, based on the database that we have got and the knowledge of how common these various characteristics are and applying some standards statistical principles, they are in common use for other laboratories as well, basically I can say that the DNA Profile from the seminal material from the oral swab matched the DNA Profile of Michael Gibson and also that the chance of a person, a second person unrelated to Mr Gibson, having that same DNA Profile is less than one in a hundred million.

So if you are saying, well, okay, we have got this DNA Profile from the oral swab, we have got the DNA Profile from Mr Gibson and they match and the chance of a second person matching is less than one in a hundred million, well, what about the chance that it was a brother of Mr Gibson? There the calculation would be quite different. So instead of less than one in one hundred million, it is actually not quite as unlikely that someone would have the same DNA Profile, if it was a member of Mr Gibson's family. And I have actually got the calculations somewhere for the brother.

So the chance of a sibling, a brother of Mr Gibson having that same DNA profile is less than one in 6000, okay, so it's quite different to the one in 100,000,000 figure because Mr Gibson and any brothers that he may have are much more likely to have the same DNA profile.

And obviously if he was an identical twin then they'd be identical? ... Yes.

What about a father or a son? ... Okay the chance of - so if you were saying it wasn't Mr Gibson, what's the chance of Mr Gibson's father having left that DNA profile, well having that same DNA profile rather, the chance is less than one in 100,000,000, no, one in 1,000,000 rather, so again, it's not one in 100,000,000.

42 The material supplied to this Court comprised two letters from Dr Bentleigh Atchison, the Manager, Molecular Biology of the Victorian Institute of Forensic Medicine, one undated but presumably written before the date of trial, and in all probability the same as that used in cross-examination, and the second dated 8 January 2001. The letters contain the following observations and reservations about the use of DNA evidence:

(1) That on the material provided to him, the procedures followed by Dr Szabo were in accordance with acceptable techniques, were clear and subjected to appropriate controls.

(2) Differences between DNA profiles can arise from:

(iii) incorrect typing due to artefacts in the testing system;

(3) Probability affected by sampling error can depend on the amount of material available for sampling. Differences in the figures provided by Dr Szabo could be explained by the method used, but the disparity was explicable by reference to the method used. However:

"... to say that DNA testing has to be 100% accurate would be impossible to achieve."

(4) Differences of opinion as to the validity of DNA figures presented to courts exist within the Australian scientific community. In July 1999, a resolution was agreed:

"... that there are three ways of presenting such data:

a Multiplication of allele frequencies without population structure and relatives.

b A conditional probability estimate using a correction for sub-population but assuming a close relative is not involved.

c A conditional sub-population estimate based on a profile being seen in a brother. This is considered to be a conservative estimate, and is presented even though it may not be known if a full sibling exists."

"...there is no general agreement as to which of the methods is correct."

(5) That there was difficulty with the underlying assumption adopted by Mr Szabo, namely:

(a) that the database was representative of the ethnic/racial background of the accused which, in turn, required evidence of his genetic make-up;

(b) the statistical calculation was derived independently of the sub-population which might not be valid for the Tasmanian population, whilst the calculation of conditional probability was derived from a sub-population, rather than a general database. In the opinion of Dr Atchison:

"Testing of a database drawn from the whole of the population is therefore irrelevant to the question being posed.

The sub-population approach has no legal precedent in Australia where Courts have ruled that a general population database must be used. If it cannot be assumed that the alleles are independent in the sub-population, then the figure cited is incorrect.

There is no logical reason for alleles to always be independent in a sub-population and there is no scientific way of testing this. This may be considered to be an unacceptable hypothesis under scientific methodology."

Presumably it is from this passage that the appellant derived the formulation of his first ground of appeal.

(c) the assumption that a close relative was not involved in that the appropriate calculation would be reduced from a conditional estimate of 33 billion to 6172 if a full sibling was considered. A figure of 1 in 100 million is not conservative if a relative has been excluded from the probability calculation.

(6) Absence of any calculation to take into account testing/error rate of, say, 5 per cent.

43 The propositions advanced by Dr Atchison will be accepted for the purpose of the determination of this appeal, although evidence in support of those propositions was not advanced at trial.

(7) The method adopted depended on a number of assumptions, some of which were either subject to disagreement or required explanation. An example of the complexity of the methodology was Dr Atchison's disagreement with a statement by Dr Szabo in his report that:

"The chance of a second person, unrelated to Michael John GIBSON having the same DNA profile as that of the seminal material on the swab is less than one in 100 million."

"The probability of producing the DNA profile of the seminal material by random mating in a sub-population surrounding the accused and conditional on the fact that the accused has the profile, is less than 1 in 100 million."

(8) One of the assumptions made was that a close relative had not been involved.

(9) The main critique concerned the use of the Tasmanian database which assumed that it was representative of the ethnic/racial data background of Mr Gibson. Since the use of the conditional probability method requires consideration of a sub-population, the use of a general population database is irrelevant. In his opinion:

"If it cannot be assumed that the alleles are independent in the sub-population, then the figure cited is incorrect.

44 Propositions (1), (2), (3), (5)(c), (6) and (8) can be readily dealt with. The evidence was that the tests had been properly conducted and appropriate controls employed. Three forensic officers had been involved in the analysis. There was sufficient material for analysis and certain results excluded because it was below the reporting threshold and the witness conceded the appropriate reservations during the course of cross-examination. Acceptance of a possible error rate would not have dramatically reduced the calculation and in any event Dr Szabo conceded that part of his opinion evidence was dependent on judgment. No evidence was given by the appellant as to the existence or whereabouts of a close relative. Dr Szabo provided the jury with statistical calculations dependent on the existence of a father or brother. There was no evidence which might have required the jury to consider the lower set of probability statistics.

45 Propositions (4), (5)(a) and (b), (7) and (9) require further consideration. Two issues are raised, namely the use of conditional probability as an appropriate methodology and the use of the database identified by Dr Szabo. The implications of the use of a general database were considered by the Court of Criminal Appeal in the Northern Territory in Latcha v R [1998] NTCCA 120; (1998) 104 A Crim R 390. In that case, the court was dealing with a prosecution in which the complainant had identified the perpetrator and in relation to the issue of statistical probability, stated, at 395:

"In a case such as this, where the prosecutrix's evidence has identified the accused as the perpetrator, the relevant question is: What is the probability of obtaining a matching analysis of the crime scene sample if someone else left it? In order to answer that question, the evidence as to probability must be based on the whole population, and not on a limited part of it."

46 The formulation of the question is dependent upon whether a selective or general database is used. In this case, Mr Szabo stated that whilst he used, as a commencing point, data obtained within Tasmania, the correlations made with the wider Australian figures shows consistency between the two and lessens the possibility of error. Nevertheless, care should be taken by forensic officers in formulating the correct question before providing an opinion to a jury. The distinction might be clear to a scientist, but the expert should remember that the task is to assist jurors chosen at random from within the community, to understand the evidence, not to present an irrefutable conclusion.

47 The forensic witness explained the terms of probability calculated in accordance with statistical principles. In explaining those calculations, he told the jury:

"If on the other hand the chance of going down the street and finding someone with that same DNA Profile is much less, like less than one in a hundred million, then that information becomes potentially much more useful in terms of going towards identification. So in this particular instance, we, based on the database that we have got and the knowledge of how common these various characteristics are and applying some standards statistical principles, they are in common use for other laboratories as well, basically I can say that the DNA Profile from the seminal material from the oral swab matched the DNA Profile of Michael Gibson and also that the chance of a person, a second person unrelated to Mr Gibson, having that same DNA Profile is less than one in a hundred million."

"So if you are saying, well, okay, we have got this DNA Profile from he oral swab, we have got the DNA Profile from Mr Gibson and they match and the chance of a second person matching is less than one in a hundred million, well, what about the chance that it was a brother of Mr Gibson? There the calculation would be quite different. So instead of less than one in one hundred million, it is actually not quite as unlikely that someone would have the same DNA Profile, if it was a member of Mr Gibson's family. And I have actually go the calculations somewhere for the brother. Just bear with me.

Yes, take your time ... So the chance of a sibling, a brother of Mr Gibson having that same DNA profile is less than one in 6000, okay, so it's quite different to the one in 100,000,000 figure because Mr Gibson and any brothers that he may have are much more likely to have the same DNA profile.

And obviously if he was an identical twin then they'd be identical? ... Yes.

48 The dangers of non-mathematically trained persons becoming beguiled by scientific material suggestive of probability have long been recognised. It has been the subject of analysis by mathematicians, (Can Jurors Understand Probabilistic Evidence, Kay and Koehler, (1991) JRSS(A) 154:75-81), and members of the other scientific disciplines (The Evolving Role of Statistical Assessments as Evidence in the Courts 1989 Ed, Expert Evidence 1991 Wisconsin Law Review Gross, Interpreting Evidence: Evaluating Forensic Science in the Courtroom, Robertson and Vignaux, 1995, DNA Evidence: Probability, Population Genetics, and the Courts, Kaye 7 Harv JL & Tech 101 (Fall 1993)) and contributions to legal journals (DNA on Appeal Robertson and Vignaux NZLJ, June 1997, 210, DNA on Appeal - II NZLJ, July 1997, 247).

49 One of the many problems associated with the use of statistical calculation and probability is the application of Bayes Theorem. The correct approach for an expert testifying in a criminal case is to provide the jury with a likelihood ratio which represents:

That expression, in turn, is fraught with difficulty. It should not be expressed in terms of guilt or innocence, but in terms of presence or absence at the scene of the crime: the fact that the defendant left a source of DNA at the scene of the crime does not mean that she or he is guilty. However, the terms "guilt" and "innocence" are used for simplicity. (Redmayne Doubts and Burdens: DNA Evidence, Probability and the Courts [1995] Crim LR 464 at 467 n23). The role of DNA evidence can be depicted in a form using Bayes Theorem:

In evaluating DNA evidence, two distinct questions arise:

(1) what was the probability that an individual would match the DNA profile from the crime sample, given that he was innocent?

(2) what was the probability that an individual was innocent, given that he matched the DNA profile from the crime sample?

The giving of the answer to the first question as an answer to the second has been described by the English Court of Appeal as "the prosecution fallacy" (R v Deen CA 21 December 1993, The Times, 10 January 1994, and has been noted subsequently in R v Gordon CA 9 June 1998, (see generally Redmayne (supra)). In R v Doheny (CA 31 July 1996), the English Court of Appeal revisited the area. It warned that the following line of reasoning was flawed:

(1) only one person in a million will have a DNA profile which matches that of the crime stain;

(2) the defendant has a DNA profile which matches the crime stain;

(3) ergo, there is a million to one probability that the defendant left the crime stain and is guilty of the crime.

50 As observed by Robertson and Vignaux, DNA on Appeal - II (supra) at 248:

"Some argue that the methodology runs counter to the legal doctrine of presumption of innocence. If that presumption renders the prior odds at 0, then the equation is rendered meaningless."

51 However, those prior odds can be assessed on the basis of the existence of other evidence before the DNA evidence is introduced. That was done here. The complainant had made a photo-fit identification, the offender was a resident of the area, the general description given by the complainant was corroborated in general terms by an independent witness, and the sighting of the appellant in the same area on a different occasion. The appellant disputed presence at any time relevant to the commission of the crime. Thus, there was evidence capable of establishing a basis for prior odds and, more significantly, the focus of the DNA evidence was on presence at the scene, not guilt. Acceptance of a match between the genetic material retrieved from the body of the complainant and the denial by the appellant of any presence inhibited any wider speculation. The statement of probability might have been too simplistic, but was confined to presence.

52 The appellant made reference to the database and claimed that it was numerically deficient and made no differentiation between different races or ethnic sub-populations. Some scientists contend that "the occurrence of a higher frequency and non-interdependence between alles will prejudice defendants who happen to be from those sub-populations", (Redmayne, (supra) at 477, cf Weir, Population Genetics in the Forensic DNA Debate, (1992) 89 Proc Natl Acad Sci USA, 11654 at 11657). It would appear that the database relied upon by the forensic witness differed from that used by the prosecution in the trial leading to the appeal in Jeffrey v R [1991] TASSC 114; [1991] Tas R 336. In his evidence, the witness stated that the characteristics of the Tasmanian database corresponded with other standardised databases. As the witness stated:

"... we've got a data base of just over 200 individuals where we've got the DNA profiles of people from Tasmania. We've actually had a look and if you take that first area I've been talking about with the 14-18 result, we know how common it is to have a peak at the 14 location and how common it is to have a peak at the 18 location. So we can work out how common particular DNA patterns are in the population, and as a general rule with each of these DNA tests, roughly ten percent of the people, just as a rule of thumb, would have a particular DNA profile, so, without looking it up, at a guess, I would say that with that first DNA test, approximately ten percent of the population would have that DNA profile, so there would probably be, you know, two, three people in this room that would have that particular pattern of peaks if we were to test everyone in this room.

Right, now, the 200 plus lots of blood you analysed, you've analysed and used as your data base, from what sort of sources do you get those? ... We got those samples from parentage testing samples so people submitting from parentage testing to the laboratory several years ago. From some of the forensic staff within the laboratory and from a few samples submitted in the course of criminal investigations from various individuals.

Now from your research and communication with the international scientific world, are there similar data bases in other parts of the world? ... Yes there are.

And in other parts of Australia as well? ... Yes. Everyone in Australia runs this same set of DNA profiling tests that we run here so it's become standardised and everyone around Australia has standard data bases where they've had a look to see how common these characteristics are in their own population.

Right. Now, have you from time to time looked at your data base figures compared with those for other parts of the country? ... Yes we have.

And have you found any noticeable differences? ... No, not if you compare our data bases to other white Caucasian DNA populations. So what you find with these DNA profiling results, is that you get certain DNA peaks that are very common in the0 Tasmanian population, that will also be very common in other populations where the population is Caucasian or white. So it's the people of British European background essentially. If you compare our data base to people of different racial backgrounds, people of Asian background or people of Aboriginal descent, then our data base is rather different. So if you're looking to say - if you're saying to yourself, well is the Tasmanian population greatly unusual in terms of how DNA profiles present themselves in terms of what's common and what isn't, we seem to be pretty similar to other white or Caucasian populations around Australia and around the world."

53 Any question which arises from that evidence can only be the subject of evidence which, if disputed, can be tested by contrary opinion evidence. Here there is no contrary opinion evidence and any future discourse ought be the province of the scientific community conducted in such a way that the process of judicial review renders it amenable to consideration by a jury. As the Court of Appeal stated in Doheny, a "jury should not be instructed in the details of Bayesian methods", (DNA on Appeal - II (supra) at 249). But the methodology and resulting opinion ought be expressed in terms consistent with scientific rigour but susceptible to understanding by a juror.

54 The limits of that requirement were considered in Pantoja v R (1996) 88 A Crim R 554. In that case, the New South Wales Court of Criminal Appeal determined that:

(1) the trial direction should state that opinion evidence based on DNA testing is capable of establishing possibility, rather than factual certainty;

(2) competing views or opinions remain issues for a jury;

(3) care must be taken in the use of statistical calculations and their methodology;

(4) the database must be referable to the particular characteristics of an accused.

55 In relation to the fourth proposition, Hunt CJ accepts at 563:

"... that the Crown was entitled to use the databases derived from the general population"

"It is simply not practicable to devise a database which took all the circumstances peculiar to this case into account"

without discounting the possibility that in special cases, dependent on evidence that any probability opinion expressed by a witness should be subject to specific direction. In this case no such direction was required by the evidence.

56 The issues raised by this appeal are evidentiary. They raise questions of the necessity for special directions, rather than admissibility. They are primarily the responsibility of the scientific community and ought remain susceptible to the ordinary processes of scientific challenge and review. But the translation from scientific rigour to ordinary members of the community who comprise a jury requires caution, lest that discipline ignore its own constraints.

57 The opinion of the forensic witness was expressed in general terms unaccompanied in the reservations expected within the scientific community. That opinion was expressed in an endeavour to assist a jury in a manner not intended to prejudice the accused. The evidence does not disclose the methodology used by the witness in his use of statistical analysis. Nor does the evidence disclose the details of the composition of the database. No criticism is intended by the preceding statements. No coherent propositions were put challenging the opinions and no contrary evidence called. I have no idea as to whether the witness used Bayesian methodology in his formulation of the opinions provided to the jury. But an analysis of the arguments advanced by an unrepresented appellant have warranted an examination of the propositions discussed. However, that examination has caused me to reach the conclusion that there was no error in the reception of the evidence or the directions given in relation to the evidence nor that there has been a miscarriage of justice.

59 I have had the advantage of reading the reasons for judgment prepared by Slicer J and agree with his reasons for dismissing grounds 3 (identification) and 4 (procedural fairness).

60 In order to deal with grounds 1 and 2, I do not find it necessary to enter into a detailed analysis of DNA expert evidence as has been undertaken by Slicer J .

61 Ground 1 is that the learned trial judge failed to instruct the jury "that the statistic Mr Szabo arrived at has no legal precedent in Australia". This ground confuses the question of the admissibility of expert DNA evidence with the question of the directions that are appropriate when such evidence has been admitted. There is ample legal authority in Australia and elsewhere for the admission of DNA evidence. The first Tasmanian authority I am aware of is R v Brown A22/1990. In that case, Wright J, in the course of rejecting a submission that expert evidence of DNA tests and their results was inadmissible, commented that whilst it is true that the procedure of DNA testing is relatively recent, having first been developed in the mid-1980's, it can no longer be regarded as a novel or experimental system, nor does it lack acceptance within the scientific community. In that case, a challenge was also mounted to the statistical basis for the expert's opinion on the improbability of the chance matching of DNA profiles. In allowing the evidence his Honour said it was satisfactorily explained and demonstrated by the mathematical equations proffered and the statistical material referred to. Shortly prior to the decision in R v Brown, it was held in New South Wales that DNA profiling was now considered to be an acceptable scientific technique for the identification of the source of bodily tissues (R v Elliott, an unreported decision of Hunt J dated 6 April 1990). Over the ensuing decade DNA evidence has been accepted in courts throughout Australia and elsewhere on numerous occasions. The learned trial judge did not err in allowing the Crown to lead evidence of the DNA profiles obtained and evidence comparing those profiles which involved the calculation of the probability of coincidental matches between profiles. That evidence having been admitted, the directions which the learned trial judge was required to give in relation to it did not include a direction that Mr Szabo's calculations of the probability of profiles matching had no legal precedent in Australia. Calculations of this nature are frequently made in relation to DNA profiles and evidence of them is commonly given. This ground of appeal fails.

62 Ground 2 is to the effect that the learned trial judge erred in instructing the jury that the method adopted by Mr Szabo is accepted throughout Tasmania, Australia and other countries. In the course of his evidence Mr Szabo said that the database of Tasmanian DNA profiles was similar to databases in other parts of the world and that, throughout Australia, DNA databases and profile tests had been standardised. He said the DNA profiles of Tasmanians were pretty similar to those of other white Caucasions across Australia and throughout the world. In the course of his summing up the learned trial judge did not, as this ground of appeal asserts, instruct the jury that Mr Szabo's method was accepted throughout Tasmania, Australia and other countries. His Honour accurately and properly summarised the gist of the evidence of Mr Szabo to which I have referred. No error arose from him doing so. This ground of appeal cannot be sustained.

At a glance

Court

Decision date

Before

Catchwords

Source

Catchwords

Judgment (184 paragraphs)

Parties