Pre-recorded evidence and juror decision-making: evidence review

5. Using Pre-Recorded Evidence in Practice: Influence upon Jurors

This chapter turns attention to a small but important body of research which examines the impact upon jurors of differential ways of introducing pre-recorded evidence or live-link testimony in practice within criminal trials. In this respect, the length and format of police forensic interviews has been suggested to have a significant effect, but the existing evidence remains inconclusive. What is more clear, however, is that jurors are prone to be distracted by the poor audio and visual quality of the live-links and pre-recorded evidence relied upon in many courtrooms, and that factors such as the choice of camera perspective used may bear careful scrutiny for their potential to influence jurors’ assessments. This research is relevant for jurisdictions, including Scotland, who are considering making greater use of such modes of testimony, and are keen to identify best practice to ensure its introduction without unintended influence upon jurors’ evaluations of the evidence before them.

5.1 Technological Considerations and Concerns

A number of studies across different jurisdictions have expressed concern regarding the practical experience for jurors of observing video-recorded witness statements. In particular, it has been noted that such videos can be of poor visual quality and that the screens upon which a witness’ image appears (in pre-recorded and live-link conditions) are often too small or too remotely located to be viewed comfortably. The inadequacy and unreliability of audio-visual technology within the courtroom is a problem that impacts upon both adult (Stern, 2010; HMCPSI/ HMIC, 2007) as well as child witnesses (MoJ, 2016). Plotnikoff & Woolfson (2009) found, for example, that 40% of child witnesses who gave evidence in real trials over the course of their study experienced problems with technical equipment in the courtroom, or with playing their visually recorded statements. Meanwhile, in a study in Australia involving 277 actual jurors in child sexual assault trials, nearly one-third reported having had problems seeing or hearing witnesses’ pre-recorded statements because of poor audio or visual quality, with more than half reporting that this had distracted them from the evidence being presented (Cashmore & Trimboli, 2006).

In line with this, in respect of adult rape complainers, Taylor and Joudo suggest that “many police officers have difficulty operating complex recording equipment” and they report common issues such as poorly focused cameras, sub-standard sound quality, and extraneous noise or interruptions in interviews, which may distract jurors and make it more difficult for them to concentrate on the content of testimony (2005: 281). In its ‘Evidence and Procedure Review’, the Scottish Courts Service directly highlighted these concerns, suggesting that – in order to mitigate any ‘barrier’ effect associated with observing a witness’ evidence on screen – it would be important to “work with other partners to improve the quality and accessibility of video-link facilities” (2015: 34) and ensure that “the interaction between the questioner and the witness is as clear as possible” (2015: 33).

While these issues of audibility and visibility are undoubtedly important, other technical aspects, for example, in relation to the camera angle used for recording, may also merit reflection. In the context of criminal confessions, for example, camera perspective has been shown – in some studies – to influence observers’ assessments of the voluntariness of the confession and of the suspect’s guilt (Lassiter et al, 2002; Ratcliff et al, 2006). When the camera is focused on the subject only, confessions tend to be perceived as more voluntary and reliable than when the camera is focused on both the interrogator and the subject. Moreover, Landström & Granhag (2008) have shown that camera perspective can similarly impact on assessments of credibility in relation to child witnesses. When the camera was focused on the child alone, they found in their study that participants tended to assess the statement as more truthful, compared with when the focus was on both the child and the interviewer.

It is important to note, however, that this finding was not confirmed in Landström et al’s subsequent research, which also involved a child witness but found no significant effects as a consequence of camera perspective. The researchers hypothesised that the source of these differential outcomes may lie in the fact that in the 2008 study, the child was testifying about a more emotionally neutral event whereas in the 2015 version, the child was testifying about a bullying incident that they had encountered. Landström et al suggested that “it is possible that when taking part of an emotionally involving interview, observers direct their attention more towards the child, leaving less room for extra-legal factors, such as the camera perspective bias, to influence their judgments” (2015b: 379). This is clearly an issue that requires further investigation, but if such factors are apt to influence jurors, it is important that they are taken into account, and that this influence is mitigated against in the best practice that is adopted.

5.2 Length and Structure of Police Forensic Interviews

A further concern that has been expressed is that the forensic interviews relied upon in the courtroom as evidence-in-chief by witnesses are often unduly lengthy and convoluted (Westera et al, 2013b; Westera et al, 2011). Research has certainly established that witness’ responses in the context of police interviews are often longer and more detailed than those provided via live evidence-in-chief: indeed, as much as five times longer in respect of female adult rape complainers (Westera et al, 2013b; see also Edwards (2003); Konradi (1977)). Whilst, on the one hand, this reflects a success of special measures to the extent that it enables witnesses to give a more complete account than they may manage in the courtroom, the worry of many commentators is that jurors will disengage from, or be unable to process effectively, the content of such lengthy video interviews.

Currently, there is little evidence to support this concern. Westera et al (2015), in a study that involved over 400 members of the public as volunteers, tested the impact that different lengths of testimony have on third party observers. They concluded that neither the overall length of the testimony nor the length of individual responses per se had any significant effect on ratings of complainer credibility or in relation to the perceived guilt of the accused. These testimonies were not, however, situated in the broader context of a simulated trial where their relative length may have had a different significance. What is more, they took the form of computer-generated audio recordings, which reduced the influence of other factors associated with juror engagement for experimental purposes, but left open the question of this influence in the context of ‘real’ trials.

The most obvious remedy to any lingering concern about length would, of course, be for prosecutors to edit video interviews before they are shown to the jury. While this comes at some risk to narrative flow, in its guidance in relation to child abuse cases, the CPS in England and Wales clearly supports this approach, stating that “it is particularly important to edit irrelevant material from long, rambling interviews” lest jurors struggle to identify and concentrate effectively on the key issues (2009: Reviewing of Videos, para 3). One of the sections of the interview most likely to meet with the cutting room floor in this context is the section at the start, where the interviewer explains the basis upon which the questioning will proceed, and asks basic questions to create a rapport and, particularly in the case of child witnesses, establish their levels of understanding. In respect of adult rape complainers, Westera et al’s recent work with New Zealand prosecutors would strongly support that removal. Here, prosecutors suggested that “the preamble is detrimental to the perception of the victim” because “what is the first thing the jury see of our victim? Her sitting there for 10 minutes being lectured on how she’s to respond” (2017: 257).

At the same time, however, Krahenbuhl has raised questions about the wisdom of this editing, at least in relation to child witnesses. In her study, 323 participants were recruited from the local and university community and asked to read an edited version of a real investigative interview transcript involving a child witness (who was aged either 4 or 6 years old depending on trial condition). Though the study’s reliability may be somewhat undermined by its use of a written transcript rather than an observation of videotaped testimony in a trial simulation, Krahenbuhl found that removing the rapport stage had an overall negative effect on jurors’ assessments. While it has been suggested that the rapport stage is superfluous at best, and prejudicial towards the witness at worst, her findings indicated that “when [it] was presented, participants rated the child as providing more complete and clear information, as being more accurate, honest, credible and confident than when the rapport stage presentation was omitted. The child was also perceived by participants to understand the functions of the interview more fully and the interview was perceived to be of a higher quality” (2012: 855).

Related to this point about its length in and of itself, concerns have also been expressed regarding the adequacy of police interviewing techniques relied upon in the video interview (Stern, 2010). In McMillan and Thomas’ research on rape prosecution, for example, practitioners reported that, in addition to “inadequate stylistic / technical abilities on the part of the police”, which meant that complainers were filmed at a distance that made it impossible for jurors to see their facial expressions, they felt that the ability of the prosecution to present an effective case was often undermined by “the attention to detail and the rigorous and at times repetitive questioning required by the police to further their investigation” (2011:275-6). Similarly, in a study involving interviews with prosecutors in New Zealand, Westera et al reported that respondents consistently described the police forensic interview as incoherent and “not unfolding like a normal human interaction” (2017: 256). Whilst the use of open questions is core to best practice under ‘Achieving Best Evidence’ guidelines in England and Wales (and its counterpart in other jurisdictions where police undertake forensic interviews), respondents expressed frustration at the tendency for such questions to provoke narrative responses in which too much irrelevant detail was provided, creating the prospect of minor inconsistencies that could taint a witness’ overall credibility.

This speaks, of course, to a broader tension which lies at the heart of the use of such interviews for dual purposes within the criminal justice process. The aim of the police interview is not, after all, just to gather evidence for future use in the courtroom, it is also a fundamental part of the investigative process, conducted at a point in time in which the interviewer cannot know or predict which factual specifics (the colour of the curtains, the weather on the night in question, etc.) may prove to be relevant. While using the police investigative interview may be cost- and time-efficient, and ensures that the witness needs only give testimony (or at least evidence-in-chief) on one occasion, there are undoubtedly challenges associated with navigating its dual purpose in a way that avoids the interview becoming too comprehensive and unwieldy.

5.3 Future Research

As with the question of the overall impact of mode of evidence delivery (whether by child or adult witnesses) upon jurors, the evidence base in respect of these more operational aspects of introducing pre-recorded testimony into the criminal courtroom is limited. Nonetheless, it is clear that, to the extent that there is evidence of some level of influence upon jurors - in respect of some types of witness in certain types of cases – it is important to understand better what exactly produces this effect. Further research about the impact of camera perspective, the consequences of high and low quality AV equipment, and the challenges and opportunities associated with the use of police forensic interview is needed. This will allow us to more confidently identify, and take steps to ameliorate, any negative impacts upon jurors’ credibility assessments associated with the introduction of pre-recorded testimony into trial proceedings.