breaking news

The verdict on reasonable doubt

By Samantha Lundrigan, PhD Anglia Ruskin University

Much has been said about the ability of jurors to understand legal direction in criminal trials.  In particular, legal direction regarding the criminal standard of proof, beyond reasonable doubt (BRD) has come under close scrutiny with questions raised as to the extent to which jurors’ correctly understand and apply the standard. Although reasonable doubt is embedded in legal parlance and the general public are familiar with the phrase, it does not necessarily follow that jurors truly understand what it means. Indeed, there is plenty of research evidence to suggest that interpretations of BRD can vary considerably between jurors, and inappropriate application of the standard is a basis of appeal upon conviction. This is potentially problematic as in the eyes of the law BRD is a static concept that should be uniformly applied.

            Over the last few years, I have been working in collaboration with Prof. Mandeep K. Dhami from the University of Middlesex and Dr. Katrin Mueller-Johnson from the University of Cambridge on an Arts and Humanities Research Council funded project that empirically investigated the public’s interpretations of reasonable doubt. We attempted to identify some of the extra-evidential factors that may account for differences in jurors’ understanding and application of this conviction threshold. Gaining a better understanding of how people interpret standards of proof can help legal professionals appreciate how it is used by jurors, and how they may be assisted to use it appropriately.

         The starting point for our research programme was to better understand how people quantitively interpret BRD. We wanted to find out whether jurors’ interpretations were generally in line with the perceived wisdom of a probability of .90 or, to borrow Blackstone’s adage, “it is better that ten guilty persons escape than one innocent suffer” (which some translate as a conviction criterion of .91). Beyond this, we were interested in how peoples’ interpretations of BRD were influenced by the wording of judicial instructions such as the JBS Bench Book recommended ‘sure’ direction. We were also interested in examining how jurors’ perceptions of the potential ‘costs’ of a case such as potential penalty and consequences of a custodial sentence for a defendant might affect their interpretations of BRD. Finally, we wanted to explore the relationship between people’s criminal justice-related attitudes and their conviction thresholds. Since the Contempt of Court Act 1981 prohibits the study of real jurors, we employed a method common in juror decision-making research i.e., surveys of individual mock jurors at the pre-deliberation stage of a simulated trial. Altogether, we surveyed 515 members of the public and focused on cases involving burglary, assault and attempted murder.


What does ‘beyond reasonable doubt’ mean to jurors?

            Across the studies conducted, we observed the lowest average interpretation of BRD  was found in a study examining the effect of  the ‘sure’ instruction. Here, the average interpretation of BRD was .92 (which is potentially acceptable). The highest average interpretation was found in a study examining consequences of a custodial sentence for a defendant. Here, the average BRD interpretation was .97 and would be considered to be too stringent.

We were also able to examine the variability in interpretations of BRD across jurors. Across the studies we conducted, variability in interpretations ranged from a worryingly low BRD interpretation of .35 to an equally concerning high interpretation of 1.00 (or absolute certainty). Finally, we also found evidence of considerable intra-individual variability implying that the reasonable doubt represents a fuzzy concept in the minds of individual jurors.


Does the ‘sure’ instruction help jurors understand ‘reasonable doubt’?

         A possible source of the variation in interpretations of BRD lies in the imprecision of the standard. In an effort to reduce inter-individual variability in interpretations of BRD as well as to bring interpretations closer to that intended, there have been attempts to define the standard for jurors.  A departure from reasonable doubt language in England and Wales over the last 25 years towards the JSB Bench Book specimen direction “of being sure of guilt” is one such attempt. 

         In one study we examined the effect of the ‘sure’ instruction on peoples’ quantitative interpretations of BRD as well as on their understanding of the concept and their confidence in its correct use. We found that under the sure direction, the average interpretation of BRD was .92. People also felt that the ‘sure’ instruction helped them to understand the standard and it gave them confidence in using the standard appropriately. Overall, the sure instruction was found to be significantly easier to understand than when the standard is left undefined. In addition, the sure instruction appears to have the desired impact across different segments of the jury eligible public.


Do jurors’ concerns over a defendant’s welfare influence their conviction thresholds?

         Two extra-evidential factors that may alter a juror’s conviction threshold relate to the perceived ‘costs’ associated with a guilty verdict for the defendant. These include the potential penalty associated with an offence if the defendant is convicted, and the consequences of a potential custodial sentence for the defendant. It is possible that jurors may be less likely to convict when the defendant’s alleged offence is more serious, and consequently his/her potential penalty is more severe. This so-called ‘severity-leniency’ effect suggests that as penalty severity increases jurors’ are less willing to risk making a type I error (i.e. convicting an innocent person) over a type II error (i.e. acquitting a guilty person). It is also conceivable that jurors might be influenced by the consequences of the penalty for a defendant. A custodial penalty can have many negative ramifications, and some defendants may have more to lose by a custodial sentence than others. For example, defendants with primary responsibility for childcare, household finances and employment may be deemed to have ‘more to lose’ than a defendant with no family, financial or employment responsibilities. Consequently, jurors may adjust or increase their conviction threshold so they are less likely to return a guilty verdict. This is problematic because in the eyes of the law, the conviction threshold should not fluctuate according to characteristics of a case.

         In two studies we asked people to imagine that they were serving on a jury in a criminal trial. They were asked to read a trial summary of a fictitious case. In one study we varied the seriousness of the offence the defendant was charged with and in the other the severity of the potential consequences of a custodial sentence for the defendant. Overall, we found that although these two extra-evidential factors did not influence jurors’ interpretations of BRD, it did affect other important aspects of their decision-making.  Defendants charged with more serious offences (that had more severe penalties attached to them) and defendants for whom there were more adverse consequences of a potential custodial sentence were all judged less likely to have committed the crime. In addition, jurors rendered fewer guilty verdicts on defendants who faced more adverse consequences of a potential custodial sentence. Thus, the influence of such extra-evidential variables arises at early stages of a trial when jurors are considering the evidence rather than at later stages when they are applying the conviction threshold. From a pragmatic standpoint, the fact that jurors may be influenced by extra-evidential factors further underscores the need to instruct them on the application of legal (probative) factors alone, irrespective of the potential ‘cost’ to the defendant. Both of the extra-legal factors studied here are legally relevant at a later stage in criminal proceedings (i.e. sentencing) but they should not be informing earlier decisions made by jurors.


The biasing effect of jurors’ pre-trial attitudes

            It is well known that jurors’ criminal justice-related attitudes can lead to a pro-prosecution bias. In another study we examined how jurors’ interpretations of BRD and their verdicts were biased by their pre-trial attitudes. We predicted that a juror with a pro-prosecution bias would have a lower conviction threshold than a juror with a pro-defence bias, and consequently the former would be more likely to return a guilty verdict compared with the latter. The results of our study clearly demonstrated that jurors' interpretations of BRD and their verdicts are associated with their pre-trial attitudes. The more biased jurors were towards conviction, the lower their quantitative interpretation of BRD (i.e., lower conviction threshold). In addition, jurors with a pro-prosecution bias were more likely, given the same evidence, to find a defendant guilty than those with a pro-defence attitude. Finally, pre-trial attitudes combined with interpretations of BRD were shown to be an important predictor of how a juror will vote regardless of the evidence.




            The right to trial by a fair and impartial jury is one of the pillars of our justice system. In addition, the due process model provides legal safeguards for defendants. Studies have shown public confidence in the jury system to be consistently high and by and large, it seems that they have every right to be confident. The jury system works well and most of the time the verdict is the appropriate one. However, there is no doubt that the decision-making task faced by jurors is a difficult one. Therefore, identifying those parts of the process where jurors may need help is important. The findings of our research programme suggest that initiatives might be needed to deal with potential variability in jurors’ interpretations of BRD.  Jurors’ ability to understand legal directions and operate with as little influence from extra-evidential factors is a crucial element in the proper functioning of the jury decision-making process. Given that the legal system confers great responsibility on jurors to make decisions that may have severe consequences for the defendant’s liberty and for public safety, the system is responsible for setting out clearly what it asks of the juror, so that jurors can, and are confident that they can, accomplish this task.


Samantha Lundrigan, PhD

Anglia Ruskin University


selected articles
Copyright © Barrister Magazine 2010