The Dorrian Review and Juries in Rape Cases: Myths about Myths?
By Professors James Chalmers and Fiona Leverick (University of Glasgow) and Vanessa Munro (University of Warwick)
Today saw the publication of Improving the Management of Sexual Offence Cases, the report of a review chaired by Scotland’s second most senior judge, Lady Dorrian. The report is an extensive document which makes a wide range of recommendations designed, as the Review puts it, to improve the experience of complainers within the Scottish court system without compromising the rights of the accused.
One recommendation in particular is likely to dominate public debate: the suggestion that trial by judge alone should be considered in sexual offence cases. The Review was unable to reach a final consensus on this point, with “views being fairly evenly split between retaining and dispensing with a jury” (para 5.3), but it recommended a pilot scheme of judge-only trials, concluding (at para 5.52) that:
The traditional arguments in favour of juries are met by equally compelling arguments for trial by judge alone, which cannot be left unexamined and ignored. They are summarised or illustrated throughout this chapter. This is a question which ought to be examined in much greater depth, with further jury research, and a fully evaluated pilot scheme to support an evidence based approach, assessing any longer term changes to trial procedure introduced following this Review. This would enable the issues to be assessed in a practical rather than a theoretical way. The fact that a system has been sanctified by usage may make it difficult to change, but it should not make it exempt from thorough examination of its suitability.
Even in the form of a pilot, this is a radical suggestion. But it is important that the debate on this particular proposal does not crowd out discussion on the many other recommendations which the Review makes, just as it is important that this specific proposal be assessed holistically as part of wider recommendations that include a shift towards specialist courts.
We are in no position, on the day of its publication, to reach a concluded view on the various proposals put forward in the Review. We do, however, want to correct one misapprehension, which has already been expressed in media discussions of the proposal within hours of its publication.
It has been suggested that, to adopt the headline of an article by Joshua Rozenberg summarizing recent research by Professor Cheryl Thomas, “belief by juries in rape myths is itself a myth”. We – like many other researchers in this area – respectfully disagree.
Professor Thomas’s research is a valuable contribution, but it does not demonstrate this, as we show through a series of questions and answers below.
What did the research prior to Thomas’s work show? A substantial body of research exists, using scientifically validated ‘rape myth acceptance scales’ to measure endorsement of misconceptions regarding rape. This conclusively demonstrates that a person’s score on a rape myth acceptance scale affects verdict choices in concrete cases. This sits alongside a further body of scholarship which has utilised realistic trial reconstruction methods to explore the ways in which the sort of abstract attitudes captured in quantitative studies might translate into deliberations. These studies provide extensive evidence of participants expressing false and prejudicial beliefs during deliberations. One example of this – discussed in the Dorrian Review – is the Scottish Jury Research, which recorded the deliberations of 32 “mock juries” who watched a rape trial reconstruction. In these deliberations, jurors regularly expressed beliefs including that a lack of injuries to the complainer and/or accused is indicative of consent, that even a very short delay in reporting an alleged rape is suspicious and that false allegations of rape are frequently made and difficult to refute.
What did Thomas’s research show? Thomas published a paper in the Criminal Law Review ([2020] Crim LR 987) exploring the attitudes in relation to selected “rape myths” of 771 individuals who had just completed jury service (whether in a sexual offence trial or otherwise) across four court centres in England and Wales. She concluded that “previous claims of widespread ‘juror bias’ in sexual offences cases are not valid”, attributing this difference primarily to the fact that there are “fundamental differences between real jurors and volunteers”, with previous studies undermined by the self-selecting nature of participation, which she suggests ‘builds in’ bias. Against this background, she says very little about the research mentioned in the previous paragraph, instead concentrating on rebutting conclusions drawn from certain high-profile opinion polls.
But is it necessary to look at this work anyway? Aren’t jurors at court different from research participants? Jurors are not an alien species: they are drawn from the general public and represent the community. If jurors at court did have radically different views from the general public, that would undermine one of the key justifications for jury trial. The objection that research participants are “different” might be a valid reason for caution in evaluating, for example, the findings of a research project which sought out volunteers who specifically wanted to offer their views on sexual offence trials by advertising that opportunity. It is less clearly applicable to research such as the Scottish Jury Research which used a mixture of door to door and street recruitment to obtain a pool of participants, representative of the communities from which they were recruited and – crucially – all themselves eligible for jury service.
But Thomas’s research obtained different results with jurors at court, so doesn’t that prove there must be a difference? No. Thomas’s research appears to have differed from the existing body of research in at least three important ways: the participants (in her case, jurors at court), the questions asked and the responses which participants were able to choose from. To the extent that her research reports different results from prior studies, it is thus impossible to say whether this is due to the different participants, different questions, or different response options.
What do you mean by “different questions”? There are a number of validated scales which have been developed to measure the extent to which people hold rape myth supporting attitudes, with the most well-known being the Illinois Rape Myth Acceptance Scale (IRMAS), the Acceptance of Modern Myths About Sexual Aggression Scale (AMMSA) and the Subtle Rape Myth Acceptance Scale (SRMAS). The AMMSA and SRMAS in particular have been rigorously designed and calibrated specifically to minimise the possibility of social desirability bias, where participants may be influenced by their knowledge that certain answers would be regarded as socially unacceptable. They do so by language choice in the phrasing of prompts and by using multiple, differently worded questions to test for the same core attitude, which also assists in testing for consistency of views. Thomas’s research uses a very different set of questions. That may be defensible – this choice is not explained or discussed in Thomas’s article – but it means that her results cannot be directly compared with the existing research, and so does not show that jurors at court are different. It may just evidence the point which researchers have tried to address by developing these scales: that question sets have to be rigorously designed and calibrated to avoid bias.
What do you mean by “different response options”? As explained above, researchers in this field are conscious of the risk of social desirability bias. One of Thomas’s questions, for example, was to ask participants if they agreed with the statement “A woman who goes out alone at night, puts herself in a position to be raped”. This question has such an obviously socially acceptable answer that respondents may be reluctant to say “yes” to this, even if they believe it (although 4% of Thomas’s respondents did). Recognising this, researchers normally use a “Likert scale” of responses including alternatives such as “agree somewhat”, “agree” and “strongly agree”, allowing respondents to be tentative in the answers that they offer. Thomas appears to have used a blunt agree/disagree/unsure scale, which can be expected to have the effect of minimising the number of participants willing to express agreement with views which they realise are or may be socially unacceptable.
But there was an “unsure” option: wouldn’t that identify those jurors who were open to rape myths? It might. But Thomas’s conclusion rests on ignoring those jurors: she states in her Criminal Law Review article that jurors who subscribe to rape myths “amount to less than one person on a jury”. When “unsure” jurors are included, the picture is very different: 27% of Thomas’s participants reported that they either agreed, or were unsure whether they agreed, with the statement “it is difficult to believe rape allegations that were not reported immediately”; 17% either agreed, or were unsure if they agreed, with the statement that “a woman who wears provocative clothing puts herself in a position to be raped”; 15% agreed or were unsure whether “if a person doesn’t physically fight back, you can’t really say it was rape”; and 13% agreed or were unsure whether “a rape probably didn’t happen if the victim has no bruises or marks”. The fact that such a high proportion of jurors are open to these views is crucial, and even more so when these figures may be an underestimate if social desirability bias has nudged them towards giving the “right” answer.
But this is still a minority of jurors, isn’t it? Won’t their views be outweighed by the other jurors in the jury room? This is a limitation of Thomas’s research: it did not involve jurors deliberating on cases, and so it says nothing about what impact a small number of jurors expressing these beliefs in a deliberation may have. And there is a body of research on deliberations – which was not discussed in Thomas’s article, despite being widely cited and relied upon in academic and policy discussions elsewhere. That research, albeit mostly conducted with mock jurors, has demonstrated that even those who reject “rape myths'“ in the abstract may refer to and rely on them in the process of deliberation. Juries do not vote: they deliberate. No research which relies solely on polling individual jurors can provide answers about what is said and discussed behind the closed doors of the jury room.
So are you saying we should scrap juries? Not necessarily: the Dorrian Review is not doing that either (it proposes only a pilot project). But for the reasons we have given above, any claim that it has been shown that jurors simply do not subscribe to rape myths – that there is simply not a problem, or that the problem is something else – is not reliably borne out by the existing evidence and should be treated with considerable caution. This is, we would stress, only one part of the argument. It would be perfectly possible to argue – and some have - that rape myths are a real problem but that non-jury trials are not the answer and that other strategies should be adopted. But crucially, identifying the best solution to the problem requires us to be led by the weight of evidence in accepting that the problem is real, and to explore in a measured and reflective way a range of mechanisms for reform.
The authors (along with Ipsos MORI Scotland) were investigators on the Scottish Jury Research project, which reported in 2019. Fiona Leverick’s article “What do we know about rape myths and juror decision making” was published in the International Journal of Evidence and Proof in 2020, and the authors’ article examining mock jurors’ deliberations in rape cases, “The provenance of what is proven: exploring (mock) jury deliberation in Scottish rape trials”, will be published in the Journal of Law and Society later this year. Further detail on the Scottish Jury Research, along with working paper versions of both of these articles, can be found here.
Update (July 2021): The arguments made in this blog post are developed in more detail in the authors’ “Why the Jury Is, and Should Still be, Out on Rape Deliberation” [2021] Criminal Law Review 753-771. A short response by Professor Thomas appears in the same issue of the Review at 772-773.
Staff links
Photo credit: Scottish Court in Session by Mike McBey (Creative Commons Attribution 2.0 Generic Licence)
