[134] The above extracts amply demonstrate that the judge applied care and attention to the issue raised by defence counsel, applied the relevant law and, to our mind, reached a conclusion that was well within his discretionary remit with which this appellate court will not interfere.
[135] The remaining point raised in support of this ground of appeal has arisen after the event and relates to a ruling of the Crown Court in the case R v Rainey and others , relating to the much publicised the murder of Ian Ogle, concerning the imagery analyst in that case, Mr Mark Buxton. Mr Buxton, like Mr Stephens, had been employed in the forensic department of Diligence International Ltd. Mr Buxton began his employment in January 2020, succeeding Mr Stephens after he left in December 2019. The evidence of Mr Buxton was excluded in Rainey although no written ruling was made by the trial judge, McFarland J. It is agreed that the reason given was simply that he was "not a credible witness." We note that, although the Regulator's correspondence with Mr Buxton was a subject of enquiry before McFarland J, his ruling did not opine on the dispute between them. It is also plain that unlike the trial judge in this case McFarland J did not consider that he required the assistance of an expert to compare the relevant images in that case.
[136] Self-evidently we must consider the nuances of the Rainey trial in deciding whether there is merit in the comparative point now raised by the applicant. In Rainey case material was sought and obtained from the Forensic Science Regulator ("FSR") in relation to the methodology employed by Mr Buxton by way of third-party disclosure. That material has also been provided to the applicant in this appeal and now takes centre stage in support of this appeal point.
[137] The applicant now relies upon the correspondence between Mr Buxton and the FSR in late 2021. Summarising same, the FSR stated that they had received a complaint about a report of Mr Buxton in a case in Ipswich Crown Court which had not declared non‑compliance with the FSR Code of Practice, namely his lack of accreditation to ISO17025. The regulator further commented that, having seen the report, Mr Buxton's use of the FIAD scale as a means did not comply with the Regulator's requirements on the formulation of evaluative opinions.
[138] Following from the above the applicant makes the claim that Mr Stephens should have disclosed the issues raised about Mr Buxton in the instant proceedings. The applicant also contends that Mr Stephens wrongly presented the absence of differences between the respective images as providing support for them being the same. Therefore, the applicant submits, had the judge been aware of these matters he may have not permitted the scale of support to be used or may have declined to admit the evidence at all.
[139] This ground of appeal is plainly predicated on the basis that Mr Stephens "should have disclosed concerns with his approach and methodology expressed by the Forensic Science Regulator in England and Wales." It is said that this failure could have impacted the admissibility or weight of his evidence. A related claim is that the court should be left with a "significant sense of unease where relevant material was not disclosed to the defence."
[140] In reply to the defence arguments the prosecution accepts that an expert is under a duty to disclose any disciplinary finding or criticism by a professional body or regulator. However, the prosecution understandably makes the point that the criticism by the FSR did not concern Mr Stephens personally, nor was Mr Stephens under any duty to disclose something of which he was unaware. The prosecution places stress upon the fact that the FSR brought the issue to Mr Buxton's attention by letter of September 2021, almost two years after Mr Stephens had left Diligence. Thus, it is submitted that there is nothing to suggest that Mr Stephens was aware of the concerns raised by the FSR in advance of giving evidence.
[141] For the purpose of this appeal Mr Stephens was asked to comment on this alleged lack of disclosure on his part and replied as follows:
"In general, I am aware that there was correspondence between the FSR and Diligence (Jon Walklin and Mark Buxton) relating to a complaint. I left Diligence in January 2020. I did not have first-hand knowledge of the nature of the complaint and I don't recall ever being shown any documentation regarding the issue. I wasn't involved in Diligence's response to the FSR.
As for when I learned about this issue, even whether it was pre/post my oral evidence at Laganside in 2022, I cannot be certain. I've checked emails between myself, John Walklin (my former manager) and Mark Buxton (my replacement at Diligence) and the first email historically that alludes to Mark/FSR is dated 12/03/2024. Any personal WhatsApp correspondence with John Walklin only dates back as far as February 2023; there are no historical messages prior to 2023 even though we would have sent very occasional messages using WhatsApp in the past. There is a brief chain in September 2024 where we discuss the uncertainty around my attendance in the Lyra McKee matter; I show some confusion about Mark's involvement and attendance, and Jon responds to say that 'it was a different case Mark had issues with ... it wouldn't effect you.'
In summary, I think that it is unlikely I'd have known about the complaint about Mark Buxton when I gave evidence in the trial of Raymond O'Neill."
[142] We can identify no basis for looking behind what Mr Stephens has stated above. Hence, we are not satisfied that a valid case has been made that he failed in his duties to disclose relevant information to the court or misled the court in some way.
[143] In addition, we note the prosecution submission that Mr Stephens' report, which included an "expert's statement" created for the purposes of proceedings in Northern Ireland, was not technically required to declare non‑compliance with the FSR Code either because of the absence of accreditation to ISO17025 or in its use of the FIAD scale. The prosecution goes on to say that if the applicant's complaint is that he was unaware that the FSR had recommended a different scale to that used by Mr Stephens, this information was publicly available in the Appendix to the FSR Code of Practice and Conduct: Development of Evaluative Opinions FSR-C-118, published in February 2021: see 8.5.13 to 8.5.15 and the scale suggested at 8.5.12.
[144] It may be said that the prosecution stance outlined above is overly procedural and deflects somewhat from the point. It is undoubtedly preferable for every expert witness to be fully candid with the court about prevailing scientific norms. Against that, we should not be over critical of Mr Stephens, as this does not appear to have been the usual practice at least in the Northern Ireland courts at the time of this trial.
[145] In any event, even if there has been material non-disclosure, it does not follow that a conviction is unsafe on that basis. Dealing specifically with this issue i n R v A [2017] NICA 68 , at [26] the Court adopted the two-question test as articulated in R v Hadley and others [2006] EWCA Crim 2544 :
"The first question is whether the material ought to have been disclosed as being material that would have undermined the case for the prosecution or assisted the case for the defence. The second question is whether the failure to disclose renders the convictions unsafe."
[146] Ordinarily, the remedy for material non-disclosure is that the disclosure test is applied once the information has come to light. If disclosure should have been made, it can be ordered on appeal and the fairness of the trial process ensured: R v Asiedu [2015] 2 Cr App R 8, per Lord Hughes, at para [27]. In Asiedu , at [55], the court concluded that, while the material ought to have been disclosed at trial, it did not, in fact, undermine the expert's conclusions on the topics on which there was an issue at trial.
[147] In its consideration of the effect of any non-disclosure in the instant case, the prosecution has postulated three matters that fall for examination as follows:
(a) The status of the FSR;
(b) The use of levels of support; and
(c) The impact of the absence of accreditation to ISO17025.
[148] As to the first matter in para [152](a), we note that the role of Forensic Science Regulator came into existence in 2008. The Regulator is entrusted by the Home Office with ensuring that quality standards are developed, implemented and used effectively in criminal justice. They are responsible for issuing Codes of Practice, Regulatory Notices and other guidance. The Regulator was put on a statutory footing by the Forensic Science Regulator Act 2021, which came into force in England and Wales on 2 October 2023. The Act creates a duty to publish a Code of Practice and empowers the Regulator to conduct investigations and issue compliance notices. However, the jurisdiction of the FSR is confined to the investigation of crime and criminal proceedings in England and Wales, irrespective of where the 'forensic science activity' takes place: sections 2 and 11 of the Act. It has no application in Northern Ireland so the Code of Practice and other guidance is of persuasive value only.
[149] On the second issue identified in para [152](b) it is accurate to state that the views of the Regulator and other forensic science bodies do not bind the court. In R v T [2011] Cr App R 9, Thomas LJ rejected the endorsement by the Regulator of the approach taken by a number of forensic scientists and examiners within the UK who employed a likelihood ratio in the interpretation of footwear mark evidence, despite the lack of a statistical database: see [52]–[53] and [60]–[61]. He went on at [92]–[96], to reiterate the admissibility of an evaluative opinion and scale as approved in the judgment of Hughes LJ in R v Atkins and Atkins [2010] 1 Cr App R 8 .
[150] In R v Atkins and Atkins , the Court of Appeal considered the admissibility of evidence of photographic comparison and the use of expressions of levels of support in the absence of a statistical database. Hughes LJ said at para [23]:
"... we do not agree that the absence of such a database means that no opinion can be expressed by the witness beyond rehearsing his examination of the photographs. An expert who spends years studying this kind of comparison can properly form a judgment as to the significance of what he has found in any particular case. It is a judgment based on his experience. A jury is entitled to be informed of his assessment. The alternative, of simply leaving the jury to make up its own mind about the similarities and dissimilarities, with no assistance at all about their significance, would be to give the jury raw material with no means of evaluating it. It would be as likely to result in over-valuation of the evidence as under‑valuation. It would be more, not less, likely to result in an unsafe conclusion than providing the jury with the expert's opinion, properly debated through cross‑examination and, if not shared by another expert, countered by contrary evidence." [emphasis added]
[151] Hughes LJ noted at para [25], that, if such evidence were inadmissible, it would also not available to the defence before noting at para [26], that scales of expression of opinion were common in a number of fields of comparison, such as fibre comparison evidence, glass fragments and footwear patterns. He went on to offer guidance as to how the evidence should be assessed by the tribunal of fact. He said, at para [29]:
"The absence of a statistical database is something which will undoubtedly be exposed in cross-examination. The witness may expect to be asked to explain how, if no-one knows how often ears or noses of the shape relied upon appear in the population at large, it is possible to say anything at all about the significance of the match; his answers may be satisfactory or unsatisfactory but will be there to be evaluated by the jury, which will have been reminded by the judge that any expert's expression of opinion is that and no more and does not mean that he is necessarily right. Similarly, the expert may be expected to be tested upon the extent to which he has not only looked for similarities but has actively sought out dissimilarities. Those are but the simplest of the questions which plainly need to be asked of anyone offering evidence of this kind. Cross-examination will also be informed by the fullest disclosure of his method, generally, and of his working notes in the particular case being tried." [emphasis added]
[152] At para [31], Hughes LJ continued:
"We conclude that where a photographic comparison expert gives evidence, properly based upon study and experience, of similarities and/or dissimilarities between a questioned photograph and a known person (including a applicant) the expert is not disabled either by authority or principle from expressing his conclusion as to the significance of his findings, and that he may do so by use of conventional expressions, arranged in a hierarchy, such as those used by the witness in this case and set out in [8] above ..."
[153] Atkins was approved by Thomas LJ in R v T and, in this jurisdiction was relied upon by the Northern Ireland Court of Appeal, in R v McDaid [2014] NICA 1 , in the context of facial mapping. The expert in that case used the FIAD scale to express his conclusions: see para. [5]. Coghlin LJ said at para [10]:
"Such a witness may give evidence of facial similarities without being able to make a positive identification and, provided that the factual tribunal is aware that his views are not based upon a statistical database recording the incidence of the features compared as they appear in the population at large, such a witness is entitled to make use of the assessment framework employed in this case."
[154] The prosecution rightly recognises that the material produced by the FSR in respect of Mr Buxton reveals a tension between imagery analysts and the Regulator in that guidance was sought from the Regulator who was reluctant to suggest or approve an alternative scale. The correspondence from Mr Buxton also shows that he had indicated to the Regulator:
" We are completely happy to change this for a more suitable scale and would welcome the opportunity to discuss this with you in detail, face to face ." The reply from FSR effectively left it to him as it was expressed in the following terms: " you are free to develop and validate an alternative that does comply with the Regulator's requirements on formulation of evaluative opinions."
[155] Furthermore, the FSR guidance, as indicated above, takes the form of an Appendix to the FSR Code of Practice and Conduct: Development of Evaluative Opinions FSR-C-118, published in February 2021: see 8.5.13 to 8.5.15. The alternative scale compares two propositions with each other, eg the proposition that the coat is of the same of the same tone, make and mass-produced model , against the proposition that it is not. It is detailed at 8.5.12 as follows:
- No more probable (that proposition A rather than B is true).
- Slightly more probable.
- More probable.
- Much more probable.
[156] Standing back and comparing this to the FIAD scale we pause to observe that the differences may, in fact, be of limited significance in the overall scheme of opinion evidence of this nature. Furthermore, it should be noted that section 4.1.2 of the Code states:
"A staged approach will be taken to compliance, which will be published by a Regulatory Notice and/or Regulator's Codes of Practice and Conduct. The aim, however, will be for all work within the scope of this document to be compliant by October 2026."
[157] Thus, whether the Code strictly applies, either presently or at the time of trial, it is clear that verbal scales continue to be used in the criminal justice system. In R v Abdi [2022] EWCA Crim 315 , at paras [7] to [15], the use of the FIAD scale was not the subject of any criticism by the Court of Appeal where an imagery expert concluded that there was strong support for the contention that K was the man on the CCTV. The FIAD scale also finds support in the 2025 update to the Crown Court Compendium. At 15-5, para 9 it states:
" If E expresses their conclusions in relative terms (eg "no support, limited support, moderate support, support, strong support, powerful support") it may help the jury to explain to them that these terms are no more than labels which E has applied to their opinion of the significance of their findings and that, because such opinion is entirely subjective, different experts may not attach the same label to the same degree of comparability."
[158] In Atkins at [31], Hughes LJ said of the scale of support:
"... We think it preferable that the expressions should not be allocated numbers, as they were in the boxes used in the written report in this case, lest that run any small risk of leading the jury to think that they represent an established numerical, that is to say measurable, scale. The expressions ought to remain simply what they are, namely forms of words used. They need to be in an ascending order if they are to mean anything at all, and if a relatively firm opinion is to be contrasted with one which is not so firm. They are, however, expressions of subjective opinion, and this must be made crystal clear to the jury charged with evaluating them."
[159] Thus, how a subjective conclusion, such as that arising out of a comparative analysis of imagery, is expressed is of less importance, as long as the meaning of the evidence and its relative strength is conveyed to the jury. As has been the case for many years, this is adequately accomplished by evaluative scales such as the FIAD scale.
[160] The subject of the complaint to the FSR, raised by the Regulator in his correspondence with Mr Buxton was that his expert report did not declare non‑compliance with the Code of Practice. The need for such a declaration is itself contained within the Code. There is no requirement for a report served in proceedings in Northern Ireland to contain such a declaration.
[161] Delving more deeply into this issue and engaging with the third issue identified at para [152](c) of accreditation we find merit in the prosecution analysis for the following reasons. First, non-compliance arises principally from the fact that the Code states that the Regulator 'expects' accreditation to ISO17025. An "expectation" and a mandatory requirement are two different things. Thus, it is wrong for the applicant to claim that the Regulator took issue with Mr Buxton's "approach or methodology." The same standard applies across the forensic spectrum, including laboratories dealing with such disciplines as DNA extraction, toxicology and firearms discharge residue. It is plainly wide-ranging in its application. We are told that accreditation to ISO17025 is a costly process, beyond the financial capabilities of many of the relatively small firms who provide expert forensic imagery analysis. To the prosecution's knowledge, no imagery analyst in the United Kingdom has yet been accredited to that standard, some years on from the introduction of the requirement in the FSR Codes. That has not prevented the continued reliance on such evidence by the courts.
[162] In any event, whilst we can see that issues arose with Mr Buxton's evidence in the Rainey case that does not automatically translate into issues raised in other cases where this species of video analysis evidence is utilised. Within modern criminal justice this type of evidence will likely have value. However, that assessment is ultimately a matter for a trial judge to determine.
[163] The lesson to be taken from the foregoing analysis is that in future we consider that it would be preferable, as a matter of good practice, for experts instructed in Northern Ireland in this area to provide as full as account as possible of the methodology employed, to include use of the FIAD scale or otherwise, the FSR position and the relevance of accreditation.
[164] In the instant case, we consider that the evidence of Mr Stephens was correctly admitted with all the caveats expressed by the judge. He was entitled to give evidence using the FIAD scale. It is unclear how the lack of accreditation impacts, in any way, upon the reliability of the conclusions reached by Mr Stephens. Furthermore, the applicant does not point to any defect in his analysis and did not seek to rely on any other expert evidence in that regard.
[165] In any event, it is critically important not to lose sight of the fact that the evidence of Mr Stephens provided low or limited support to the prosecution case. This was only one aspect of a multi - layered circumstantial case. Added to that is the telling fact that no complaint has been raised as to the judge's charge on this or any other issue. Thus, there is no question of a judicial omission affecting the safety of this conviction. When properly analysed this ground of appeal satisfies the leave threshold but ultimately it must fail on its merits and so is dismissed.
Ground 8: The conviction was against the weight of the evidence
[166] This is a classic makeweight and omnibus ground of appeal. It is entirely non‑specific and unparticularised. It adds nothing of substance to the applicant's appeal. Correctly, it was not canvassed at the hearing before this court in oral submissions. While nothing further is required of this court, we would nonetheless add the following.
[167] The case against the applicant was based on many strands of circumstantial evidence which the prosecution helpfully summarises in its written argument as follows:
(a) the applicant had been in Jennifer Dornan's company earlier that evening;
(b) he knew where she lived;
(c) he was again in her company at David Quinn's home at 71 Lagmore Avenue in the early hours of the morning up until she left;
(d) he had obtained a bottle of champagne which she was drinking while he drank beer;
(e) Ms Dornan arrived home at 02:52 hrs so would have left around 02:47 hrs;
(f) he left 71 Lagmore Avenue shortly afterwards, on foot, 5 or 10 minutes after 02:58 hrs;
(g) the murderer arrived at her home at 03:11 hrs;
(h) the applicant was wearing a cream coat with pockets on the front lower part when he left David Quinn's house;
(i) the murderer wore a jacket with a pocket on the front lower part similar to that worn by the applicant that night;
(j) he displayed forensic awareness, pulling the coat up over his face when confronted by cameras, staying close to the fence as he made his way to enter by the back door;
(k) the applicant is a recidivist burglar;
(l) on leaving 19 Hazel View, at 04:18 hrs, the murderer walked off to the left before a figure was seen in that direction walking towards Lagmore View;
(m) a knife with blood on it and bearing a partial DNA profile matching that of the deceased, was found in a garden situated next to Lagmore View;
(n) a man was seen by Megan Cunningham, just after 04:21 hrs, walking into White Glen from Lagmore View, towards Nos. 45 to 60 White Glen;
(o) a man carrying a light-toned item of clothing under his arm walked past a CCTV camera at 56 White Glen, at 04:23 hrs;
(p) no other person passed that camera between 0204hrs and 0504hrs;
(q) the man reached the camera at 56 White Glen 4 mins 50 secs after the murderer left 19 Hazel View; the police walked the same route in 3 mins 38 seconds;
(r) the man at White Glen showed forensic awareness as he neared the camera, bowing his head so that light shone first on his forehead and then on his crown;
(s) the applicant has a receding hairline and bald patch at his crown (although the light could also be explained by saturation);
(t) gait analysis provided limited support for the proposition that the man at Hazel View and White Glen was the applicant;
(u) the applicant arrived at Eileen McIlvenney's home at 109 Laurelbank at 04:46 hrs, approximately 28 minutes after the murderer left 19 Hazel View. Police walked a route from there to 109 Laurelbank via White Glen and Teeling Avenue (where the applicant had earlier told David Quinn he was going) in 28 mins 20 secs;
(v) when he arrived at 109 Laurelbank, he was not in possession of his coat despite:
(i) needing a coat at all other times he was seen on CCTV that day; and
(ii) the fact that it had been raining in the early hours of that morning so that he was soaking wet;
(w) his hand was covered in blood when he arrived
(x) he behaved oddly, giving Ms McIlvenney his scapular medals and talking about God;
(y) he visited his nephew at 12 Devonshire Close, waking his family, at 05:30 hrs, despite having no need to because the taxi was taking him to 'Georges' for drink;
(z) his nephew Shane O'Neill was prepared to help him at that hour and in those circumstances;
(aa) he was heard by Suzanne Hazley admitting killing someone when he arrived at their house. He and Shane O'Neill left, likely to dispose of evidence by burning;
(bb) he behaved oddly with Anne Marie Smith, rocking backward and forward as if he was disturbed. He was talking about having the first drink in ten years and his children;
(cc) on Sunday evening, the day after the murder, he asked his nephew to drive into Lagmore. The explanation for this was implausible. They drove up to the vehicular dead end at the top of White Glen where the suspect had walked;
(dd) he washed and dried his clothes and had packed a towel and underwear; he advised Paul Smith to wash the Barcelona shirt he had borrowed;
(ee) he stayed away from his bedsit at 89 Amcomri Street, staying at two different addresses on Sunday night and Monday night;
(ff) he used false names whenever booking taxis;
(gg) when Ms. Dornan's death was commented upon by Jemma Tierney in his presence, he said nothing despite having been in her company that night;
(hh) he left the jurisdiction on the evening of Monday 3 August, travelling to Donegal by a circuitous route via Dublin;
(ii) he knew the police were looking for him; and
(jj) he gave lying and inconsistent evidence that he had lost his memory which the jury plainly did not accept.
[168] On an overall rational view there was more than sufficient evidence upon which a reasonable jury, properly directed, could properly convict. As the single judge also observed, in refusing leave, at [65]:
"[65] Whilst the applicant spills much ink in identifying the alleged weaknesses in the Crown case, all of these matters were put to the jury in the course of closing speeches. The defence made the case that key prosecution witnesses ought not to be believed. Moreover, the trial judge summed up the evidence to the jury over the course of three days and, in doing so, he highlighted the various issues and inconsistencies in the evidence. There is no claim that the judge misdirected the jury in any material respect."
[169] Put simply where there is sufficient evidence to go before a jury, as in this case, it follows that the verdict could not be said to be against the weight of the evidence. This purported ground of appeal is unarguable and also fails.
Overall conclusion
[170] Applying the test in R v Pollock [2004] NICA 34 , the simple question for this court is whether the applicant's conviction is safe. We are not satisfied that any safety issues arise in relation to the jury's verdict. Leave to appeal is granted only in respect of ground 1, and ground 7 and refused on all other grounds. The appeal is dismissed on its merits for the reasons given.
[171] Finally, we thank counsel (and instructing solicitors) for their considerable assistance in this case. We also wish to commend the judge for providing a series of excellent rulings on complex legal issues at short notice during this trial. His handling of this difficult case was exemplary.