Scottish jury research: findings from a mock jury study

4 What impact do the unique features of the Scottish jury system have on how juries reach decisions?

Key findings

• Juries asked to reach a unanimous verdict deliberated for substantially longer than those only required to reach a simple majority. There were no statistically significant differences in the length of time taken to reach a verdict between 12-person and 15-person juries or between juries which had two or three verdicts available.
• There was no statistically significant variation in the mean number of evidential issues raised within jury deliberations by jury size, the number of verdicts available, or the majority required.
• In general, there was no statistically significant variation in whether or not juries discussed specific legal issues between the three features examined. The only exception to this was that juries asked to choose between two verdicts (guilty and not guilty) were more likely than those who also had not proven as an option to discuss the meaning of reasonable doubt. Jurors that discussed the meaning of the reasonable doubt standard of proof, however, often defined this somewhat differently to the definition given by the judge.
• Jurors sometimes struggled to recollect legal tests accurately, although this did not vary with jury size, the number of verdicts available, or the majority required. Other legal misunderstandings that sometimes arose during deliberations included a belief that the accused is required to prove his innocence and a belief that self-defence is not a legitimate defence.
• 15-person juries were associated with somewhat lower levels of juror participation than 12-person juries across a number of measures. Jurors in 15-person juries were more likely to be seen wanting to contribute but being unable to do so than those in 12-person juries. 15-person juries also included both more 'dominant' and more 'minimally contributing' jurors than 12-person juries. Jurors in 15-person juries (compared with 12-person juries) and in simple majority juries (compared with unanimity juries) felt they had less influence on the collective verdict outcome.

4.1 Introduction

Chapter 3 indicated that each of the three unique features of the Scottish jury system do appear to incline jurors in particular directions in terms of their verdict choice. But what impact do they have on how jurors make those choices? In other words, what effect, if any, do they have on the collective process of deliberation? This chapter examines this question, drawing primarily on analysis of the recorded jury discussions (which examined factors including: length of discussion, scope of discussion, observed level of juror participation, and tone). It also draws on jurors' own perceptions of the deliberation process, based on their responses in the post-deliberation questionnaire.^[63] Our analysis focuses particularly on whether there is any evidence that the different features of interest - number of verdicts, jury size and majority required - may be associated with variations in the quality of jury deliberation and decision-making.

4.2 Length and scope of deliberations

The relationship between length and quality of deliberation is not clear cut. Longer deliberations might reflect repetitive discussion or inefficiencies in decision-making. Alternatively, they might be indicative of more and better substantive discussion.

In our study, juries were able to deliberate for up to 90 minutes before either returning a verdict or hanging (where a jury was required to reach near unanimity but failed to do so). On average, juries returned a verdict after 45 minutes (37 minutes for the assault trial and 54 for the rape trial).

There were no statistically significant differences in the length of time taken to reach a verdict between juries asked to choose between two and three verdicts, or between 12 and 15-person juries.^[64] The lack of difference in deliberation time by jury size contrasts with a meta-analysis of existing jury research, which suggests that, in general, larger juries tend to deliberate for longer.^[65] However, as discussed in Chapter 1, there is no existing research comparing 12 and 15-person juries specifically (comparisons tend to be between 12-person juries and smaller juries).

Jurors asked to reach a unanimous verdict did take substantially longer over deliberations than those required to reach a simple majority - an average of 54 minutes, compared with 37 minutes. However, this did not appear to be associated with unanimity juries having more wide-ranging discussions, at least in terms of the number of issues discussed. Each jury's deliberations were coded to record whether or not there was discussion of different aspects of the evidence (using an agreed coding frame^[66]). We found no statistically significant difference in the mean number of evidential issues raised between unanimity juries and simple majority juries.^[67]

We reviewed the 10 unanimity juries that deliberated for the longest period (all for 70 minutes or longer) to see if this provided any further insight into the reasons for, or nature of, these longer deliberations. We were unable, however, to discern any clear pattern. Sometimes these lengthy deliberations were relatively disorganised; sometimes they reflected a clear and entrenched difference in position from the outset between different groups of jurors. When unanimity juries did contain strong differences of opinion, jurors sometimes became frustrated at being required to deliberate for longer when they believed that no consensus could be reached. In other long deliberations, there was a gradual process of jurors changing their position until they were able to return a verdict.

4.3 Discussion of legal tests and issues

In addition to coding the number of substantive aspects of the evidence discussed, we also recorded whether or not key legal issues mentioned in the judge's closing directions to the jury were raised during deliberations. Specifically, we recorded whether there was:

• Any mention of (a) either the standard of proof or the meaning of reasonable doubt and (b) whether the evidence was believed to have met that standard of proof (beyond reasonable doubt).
• Any mention of (a) the meaning of corroboration and (b) whether the evidence was believed to provide corroboration for the complainer's account.
• In the rape trial only, any reference to each of the two legal tests for rape described in the judge's directions, namely (a) whether the evidence suggested the complainer had consented or not, and (b) whether the accused had reasonable belief in her consent.
• In the assault trial only, any reference to (a) the general plausibility of the self-defence argument (being put forward as a special defence) and (b) each of the three individual tests of self-defence, namely reasonable belief in imminent danger of attack, violence as a last resort, and use of no more than reasonable force to stop an attack.^[68]

(For brief definitions of these terms, see the 'key terms' section at the start of this report).

There was no statistically significant variation in whether or not juries discussed these legal issues by the number of verdicts available, jury size, or the majority required, with one exception: juries asked to choose between two verdicts were more likely than juries asked to choose between three verdicts to discuss the meaning of the reasonable doubt standard (but not whether or not this threshold had been reached).

4.3.1 Reasonable doubt

It is not obvious from the content of discussions about reasonable doubt why juries with only two verdict options should have been more likely to discuss its meaning. However, what is apparent is that where juries did discuss this, the definitions they reached were often more demanding than the threshold indicated by the judge.

We found no examples of jurors referring directly to the definition of 'beyond reasonable doubt' provided by the judge (taken from the Jury Manual), which states:

"…the Crown must establish guilt beyond reasonable doubt. That's a doubt, arising from the evidence, based on reason, not on sympathy or prejudice, or on some fanciful doubt or theoretical speculation. It's the sort of doubt that would make you pause or hesitate before taking an important decision in the practical conduct of your own lives. Proof beyond reasonable doubt is less than certainty, but it's more than a suspicion of guilt, and more than a probability of guilt. This doesn't mean that every fact has to be proved beyond reasonable doubt. What it means is that, looking at the evidence as a whole, you've to be satisfied of the guilt of the accused beyond reasonable doubt." ^[69]

While there were occasional references to the idea that proof beyond reasonable doubt is something less than complete certainty, there were repeated examples of other jurors referring to the perceived need to be "100%" confident or similar, which appears to run contrary to the judge's direction that proof beyond reasonable doubt "is less than certainty":

"You need to believe 100% both those (prosecution) witnesses"

(Juror, Assault trial / three-verdict / 12-person / unanimity)

"If you're 99 and three-quarters sure, you've got a quarter, you've got to go with that quarter."

(Juror, Assault trial / two-verdict / 15-person / unanimity)

Thus, the finding that jurors in two-verdict juries were more likely to discuss the meaning of reasonable doubt does not necessarily indicate a better understanding or application of the standard of proof.

4.3.2 Corroboration

In Scots law, corroboration is required in criminal cases. Mock jurors were directed on this based on the guidance used by judges in directing juries in real criminal trials. All jurors were first told:

"I must tell you about corroboration. The law lays down that nobody can be convicted on the evidence of one witness alone, no matter how strong that evidence may be. There must be evidence from at least two separate sources which you accept and which taken together point to the guilt of the accused. There are two essential matters that must be proved by corroborated evidence. These are that the crime charged was committed, and that the accused was responsible for committing it."

They were then told which facts needed to be proved by corroborated evidence (which varied between the two trials - assault and rape).

In general, jurors appeared to have fewer difficulties understanding the corroboration requirement than the standard of proof. However, there were occasional (erroneous) contributions in rape trial deliberations suggesting that the doctor's forensic evidence would have to unequivocally indicate rape before the jury could convict.^[70] We identified six rape trial juries (all of which ultimately returned verdicts of acquittal) in which jurors appeared to suggest this. In three of these, that view was expressly challenged; in a fourth it was not directly challenged but other jurors clearly took a different view; while in the remaining two juries the conversation simply moved on without challenge. Jurors who challenged this view suggested it was unrealistic to demand such a high standard of evidence. For example:

First Juror: "To me the doctor is the key person and the evidence she gave does not convict him."

Second Juror: "But, I don't think… I would be amazed if there is a crime that happens anywhere in the world, that physical evidence is provided for, that the defence can't turn round and say, 'but could something else not have caused that injury?' I don't think there's any injury that you can't say that for."

(Rape trial / three-verdict / 12-person / simple majority)

4.3.3 Trial-specific legal tests

In all types of Scots criminal cases, juries should be convinced that the relevant legal tests around reasonable doubt and corroboration have been met in order to convict. In addition, given this study's focus on rape and assault, our jurors were also required to apply legal tests linked to those specific offences.

In the rape trial, jurors were required to assess whether the evidence suggested the complainer had consented, and whether it suggested the accused had a reasonable belief in her consent. The deliberations revealed considerable debate on the former, some of which reflected differing social norms about what would constitute evidence of free agreement to sexual intercourse.^[71] However, the extent or nature of this debate did not vary between two-verdict and three-verdict juries, 12-person and 15-person juries, or between unanimity and simple majority juries. Discussion typically focused on the notion that the complainer had not suffered extensive injury or had not done enough to physically resist the accused (having said in her evidence that she had tried to push the accused off but he was too heavy) or on the fact she had not called for help (no evidence was presented to indicate whether or not there was any person who might have been able to hear such a call).^[72] Many jurors said they understood a "freeze" reaction (a term used repeatedly) to be common among rape victims, while others insisted this was not how they would have reacted.^[73] This is illustrated by the following exchange:

First juror: "So, think of it this way, think of it this way. Imagine a Saturday night and somebody is out on the town and some people decide to try and assault them. Okay, some people might fight back, some people literally can live in a bubble and don't do anything about it because they're in shock and they can't fight back. The same could apply in that case."

Second juror: "Well, I'm just saying this is how I'm feeling, I'm not saying I'm right, but you're asking my opinion. I'm just saying, if you're being attacked and it's a serious assault...then to me you would scratch, you would scream, you would try and do anything possible to get him off."

(Rape trial / three-verdict / 12-person / simple majority)

In the assault trial, jurors were required to consider the legal tests for self-defence. These tests were not always mentioned (of the 32 assault trial juries, 13 did not specifically reference all three of the tests^[74] ). Moreover, even when they were referenced, jurors often struggled to recollect each element clearly. In one jury, jurors suggested it would have been helpful to have been directed on self-defence before the evidence had been led, rather than after:

First juror: "See if the judge had said that at the beginning, then you would be looking out for this wouldn't you, as opposed to saying at the end, what did he actually say? If he said to you 'self-defence means you need to be convinced it was a last resort and it was their way of doing it and it was a minimum use of force' and you can look..."

Second juror: "I think we all pretty much agree with that".

(Assault trial / three-verdict / 15-person / simple majority)

However, there was no difference in the likelihood of the self-defence tests being discussed by any of the features of the jury system examined in this research (number of verdicts available, jury size, or majority required).

4.3.4 Supporting juror understanding

Overall, then, these findings suggest that the number of verdicts, jury size, and majority required have no clear impact on the extent or accuracy of jury discussion of legal issues. However, they do raise important questions generally about how to support jurors' understanding of those legal issues.

This latter point is also supported by analysis of the frequency with which legal misunderstandings occurred (and were or were not corrected) within the deliberations. In addition to issues around interpreting the standard of proof and corroboration, discussed above, the researchers also identified two further potential legal misunderstandings that were articulated on multiple occasions across the 64 juries.^[75] These were:

• A belief that the accused was required to prove his innocence. This was directly contradicted in the judge's directions, which stated, "throughout the trial every accused is presumed innocent unless proved guilty. The accused is not required to prove his innocence." Nonetheless, this belief was expressed in 14 jury deliberations, though it was challenged, with varying degrees of effectiveness, in eight of these.
• A belief that self-defence did not operate as a defence to assault (that is, that the accused in the assault trial was automatically guilty simply because he had stabbed the complainer). This belief was expressed in 14 assault trial deliberations - for example, "he has actually admitted that he stabbed him, so he's guilty" - and was challenged in 10 of these.^[76]

There was no statistically significant variation in the frequency with which these misunderstandings either arose or were challenged by other jurors between two-verdict and three-verdict juries, 12-person and 15-person juries, or unanimity and simple majority juries. Moreover, while we cannot know whether such misunderstandings influenced individual jurors' views, there were no cases where the discussion and verdict expressly proceeded on the basis of such a legal misunderstanding. However, the fact that jurors struggled to remember legal tests and that legal misunderstandings arose relatively frequently across the 64 juries suggests a need to consider whether jurors require additional guidance to aid their discussion.

Another strand of this research involved an extensive evidence review on ways of improving communication with juries.^[77] This found that written directions and structured decision aids can be effective in improving memory and/or understanding. Written directions involve giving jurors a written copy of the judge's directions in the case. There is a substantial body of evidence to suggest that this helps jurors to remember and understand legal tests. Such improvements in understanding tend, however, to be limited to improvements in simple comprehension. In other words, they help jurors to remember and accurately re-state legal directions. Structured decision aids (sometimes called 'routes to verdict') are documents that contain a series of primarily factual questions - which might be presented as written questions or in diagrammatic or flowchart form - that gradually lead jurors to a legally justified verdict. Evidence suggests that these are particularly effective at improving 'applied' comprehension - jurors' ability to correctly apply legal tests to the evidence. The evidence review found that the use of structured decision aids is well-established in other jurisdictions, including England and Wales, Canada, New Zealand and some Australian states and territories.

4.4 Levels of individual juror participation

Juries are intended to proceed on the basis of discussion between jurors about the evidence they have heard. The ideal is that all members of a jury are able to contribute freely and make their views known. The research team recorded various measures of juror participation in deliberations, ^[78] including:

• How often individual jurors were observed apparently wanting to contribute but being unable to do so.
• The number of 'dominant jurors' in each jury (defined as "any juror who contributes very obviously more substantially than most other jurors").
• The number of 'minimally contributing jurors' in each jury (defined as jurors who made fewer than three contributions, excluding non-verbal contributions (e.g. nodding), simple agreement (e.g. 'yes, I agree' with no expansion), or very short contributions made only as part of 'going around the table' to establish each juror's view on what verdict they should return).

These are imperfect measures. The concept of a 'dominant juror', for example, is often used to imply an inappropriate dominance of discussion at the expense of others, but is used more neutrally here, simply to identify those who spoke very substantially more than others. Similarly, we could only record instances where jurors visibly appeared to be trying to speak but were interrupted or spoken over by others - we could not capture occasions where jurors wanted to contribute but kept silent. Nonetheless, when taken together, and in combination with jurors' own views on the deliberation process (captured in the questionnaires they completed after deliberating), they provide an insight into patterns of participation.

Across a number of these measures, there was evidence of jurors being less able to participate in 15-person juries compared with 12-person juries:

• Jurors were more likely to be observed wanting to contribute, but being unable to do so, in 15-person juries (13 out of 32 juries) compared with 12-person juries (2 out of 32 juries).
• There were more dominant jurors, on average, in 15-person juries (1.8 jurors) than in 12-person juries (1.0 jurors). Jurors in 15-person juries were also more likely to agree that "some members of the jury talked too much" (41% of jurors in 15-person juries, compared with 28% in 12-person juries).
• There were more minimally contributing jurors, on average, in 15-person juries (2.5 jurors) than in 12-person juries (1.6 jurors).
• Jurors in 15-person juries also gave lower ratings of their own influence over the verdict in comparison with jurors in 12-person juries. When asked to rate their influence on a scale from 1 (none at all) to 7 (a great deal), jurors in 15-person juries had an average rating of 4.1, compared with 4.5 among those in 12-person juries.^[79]

The fact that 15-person juries contain more dominant and minimally contributing jurors might be expected simply by virtue of the larger group size. However, in combination these findings suggest that bigger juries may be associated with more jurors struggling to participate fully, and consequently with jurors feeling they have less influence on the final decision. It is also worth noting that, when reviewing 15-person juries, it was common for the research team to observe side conversations running concurrently within discussions. It was also common to note high levels of interruption and speaking over one another in 15-person juries, leading to a decreased sense of order within the deliberations overall.

There were no statistically significant differences in observed levels of juror contribution between unanimity juries and simple majority juries. However, there were some differences in jurors' own perceptions of their participation. Jurors who were asked to come to a simple majority verdict:

• Were more likely to disagree that they had felt able to say as much as they wanted (10% vs 5%).
• Were less likely to agree that they felt fully involved in the jury's decision (86% vs 92%).
• Gave lower ratings, on average, of their own influence over the verdict (4.1 among jurors asked to reach a simple majority vs 4.5 among jurors asked to reach a unanimous verdict).

In combination with the longer average deliberation time for unanimity juries, these findings suggest that requiring unanimity may provide more opportunity for everyone to feel they have been able to put their views across. This is consistent with previous research. For example, the Inside the Jury study compared 12-person juries that were required to reach either unanimity, 10 votes out of 12 or eight votes out of 12. The study found that, while there were no statistically significant differences in the frequency of jurors' participation, the relatively shorter deliberation time for juries deliberating under the 8/12 and 10/12 conditions meant that minimally participating jurors appeared much more frequently in those juries than in unanimous juries.^[80]

There were no clear differences in observed or self-assessed juror participation and influence between two-verdict and three-verdict juries.^[81]

4.5 Observed tone and juror feelings about deliberations

The function of a jury is to come to a verdict based on a clear assessment of the evidence. Whether this process is calm or fraught, and whether jurors enjoy the experience are not primary considerations. However, in the context of attempting to understand how different features of the jury system might impact on the process of decision-making, the overall tone of discussions is potentially relevant, as is jurors' own level of confidence in the verdict returned.

The research team coded 27 of the 64 deliberations as 'almost always completely calm', 22 as 'occasionally heated or animated' and 15 as 'frequently heated or animated'. Based on this data, there was no clear, statistically significant difference in tone of deliberations between two-verdict and three-verdict juries. However, 15-person juries, and juries asked to reach unanimity, were each marginally more likely to become frequently heated or animated.^[82]

Of course, designating a tone for the jury as a whole can be misleading, as the temperature of a discussion can fluctuate. Further analysis of researchers' notes on deliberation videos suggests that such fluctuations were particularly apparent in juries asked to reach a unanimous verdict, but where deeply entrenched differences between the jurors meant that the jury was struggling to agree. In these juries, exchanges tended to become more animated in the later stages of deliberation, when time was running out, even amongst those coded as only 'occasionally more heated' overall.

Overall, 81% of jurors said they were 'very' or 'fairly' confident that their jury had reached the right verdict. Unsurprisingly, juror confidence varied substantially depending on whether or not individual jurors personally agreed with the collective verdict returned. 93% of those who personally agreed with the jury's final verdict^[83] were confident it was the right decision, compared with just 38% of those who did not. A related finding is that jurors required to reach unanimity were somewhat more likely to express confidence in the final verdict (87%, compared with 76% of jurors asked to reach a simple majority). Again, this is unsurprising since requiring unanimity means that jurors are more likely to agree with the verdict returned at the end of deliberations. Jurors asked to choose between three verdicts were also a little more likely to express confidence in their jury's decision (83%, compared with 78% of jurors asked to choose between two verdicts).^[84] There was no statistically significant variation in juror confidence in the verdict between 12-person and 15-person juries.^[85]

Finally, jurors were asked (after deliberating) how satisfied they felt with the experience of being a (mock) juror. Overall, levels of satisfaction were very high - 88% were very or fairly satisfied with the experience. Of course, the voluntary (and moderately remunerated) nature of this mock jury research means that this finding is unlikely to transfer to the 'real' courtroom. What is of relevance is whether there were any differences by jury size, number of verdicts, or majority required. The only significant difference in this respect was that jurors asked to choose between two verdicts were slightly more likely to indicate that they had been dissatisfied with the experience than were jurors asked to choose between three verdicts (6% vs 3%).