Search

This paper presents an analysis of 88 criminal convictions that have since been quashed on the basis of error of fact (wrongful convictions) that have occurred in England and Wales since 2007, in the context of wider set of 389 wrongful convictions that have occurred in England and Wales since 1970. Based on this analysis, three key contributors of concern are identified as having been influential in leading to wrongful convictions recently - digital evidence, guilty pleas, and misleading testimony. Cases involving each of these factors are discussed, including cases from the Post Office Scandal, which make up many of the identified wrongful convictions during this period. In considering each factor, failings in the criminal justice system that leave defendants vulnerable to wrongful conviction are discussed. The paper concludes with brief initial suggestions for reform to provide greater protection against highlighted vulnerabilities.

We recently proposed, and found support for, a new conceptual model of plea decision making (informed by Fuzzy Trace Theory; Reyna, 2012), that explains the plea decisions of non-detained, factually guilty (mock) defendants who are choosing between a guilty plea or trial, where both the plea and potential trial sentence entails incarceration (Zottoli et al., under review). Our model expects plea decisions to be affected by meaningful, categorical changes in perceived conviction probability (e.g., low-to-moderate; moderate-to-high) and by the presence and magnitude of categorical distinctions between plea offers and potential trial sentences (e.g., small-to-medium; medium-to-large), rather than fine-grained differences between individual offers. In this study we seek to clarify the prior findings with respect to the form of the function relating plea-discount and perceived probability of conviction to plea acceptance for guilty-mock defendants, and extend the model to factually innocent defendants. We will use a modified crime scenario to replicate the Zottoli et al., (under review) plea discount conditions (0.1 to 0.9) across two of their three potential trial sentence conditions (9 year, 18 year), and will add a factually innocent condition.

These raw data include images that were used in our analysis and processed images formJMicroVision v1.3.2.&nbsp

Authoritarianism has been shown to affect jury decision making in predictable ways across different types of case (Narby et al., 1993), with mock jurors who score highly on authoritarianism measures displaying a greater tendency to convict. Whilst such authoritarian pre-dispositions are relatively stable (see, for example, Stenner, 2005), the expression of authoritarianism has been found to vary under certain circumstances. For example, under threat. Research suggests that threat can lead to shifts towards authoritarianism (Bonanno & Jost, 2006; Nail et al., 2009; Toorn et al., 2017), but can also lead to increased polarisation between groups based on political attitudes, with liberals shifting away from authoritarianism (Stevens & Banducci, 2022). Research in political science suggests that the effect that threat has is dependent on the type of threat experienced. On the one hand, normative threats, potentially including dissatisfaction with existing political institutions or feelings of threat to society generally, have been shown to generate divergent responses from authoritarians and libertarians (this also includes sociotropic threats, such as the threat of terrorism to society; Arikan, 2022). On the other hand, personal threat can produce the opposite effect. Personal threats, such as the threat of physical harm to friends and family, generate convergent responses towards authoritarianism by both libertarians and authoritarians (Hetherington & Suhay, 2011). This research seeks to develop and test predictions as to the effects of different types of threat on verdicts in the context of juror decision-making. Specifically, it will examine whether the content of a case (terrorism) can invoke threat in different ways. It further seeks to determine the extent to which participants’ experience of personal and normative threats may interact differently with political persuasion in influencing verdict through triggering the relationships discussed above.

This project will examine a specific type of metacognition that is relevant to evaluations of testimony in criminal and civil justice systems – judgments of what another person would remember had they experienced a stimulus, i.e. social metamemory judgments. This study will examine the potential for inaccuracy to occur in social metamemory judgments as a result of generalisations from one’s own beliefs about memory and experience of a stimulus to others, even where the stimulus is presented in the same way to the assessor and the person whose memory is being assessed. In the experiment we will examine the accuracy participants’ judgments of what they themselves would remember after viewing a stimulus (judgments of learning; JOLs) and the accuracy of participants’ judgments of what they would expect others to remember after viewing the same stimulus (social metamemory judgments). Broadly, we predict that JOLs will be more accurate than social metamemory judgments, and that this is partly the result of people making generalisations from their memory to the memory of others who may differ from them in meaningful ways. We will test these predictions by comparing the accuracy of JOLs and social metamemory judgments, and by examining whether the accuracy of social metamemory judgments improve when participants make judgments about others who share a meaningful characteristic with them, making generalisation more appropriate.

The proposed work is the second in a set of studies examining social metamemory judgments – judgments of what another person would remember had they experienced a stimulus. The work examines one potential source of error in these judgments – error resulting from the fact that relevant stimuli are presented to jurors in a different way from how they have been experienced by the person whose memory the jurors are assessing. For example, while a witness may have seen a person or crime scene for a short time with many distractions, a juror may see the person or crime scene for a longer period with no distractions. However, because the influence of experience of a stimulus can influence judgments about the memorability of that stimulus unconsciously, the assessor may not realise that their expectations about memorability have been influenced by their own experience, which is different to the experience of the person whose memory they are assessing. Thus, their own experience may subconsciously influence their social metamemory judgments, despite this experience having no probative value. For example, if a juror is able to examine a picture of a crime scene for a longer period they may think it is a more memorable scene than a juror who has only been able to examine the picture of the crime scene for a shorter period, but the length of time the jurors have had to view the stimulus has no effect on how memorable the scene would have been for the witness. In this experiment, we will examine the potential source of error described above through altering the experience of a stimulus for assessors by manipulating stimulus size and length of exposure to the stimulus (two experience-based cues shown to influence JOLs in the context of words), but keeping the described experience of the person whose memory is being assessed constant. We will test the prediction that changes in how the assessor experiences the stimulus will impact their social metamemory judgments for the person they are assessing, despite having no probative value in those social metamemory judgments.

This study looks at how people evaluate identification memory of others, and how this evaluation might change or become polarised as a result of case context.

Previous research on visual comparison tasks has shown that low non-match base-rates impair non-match detection and result in more a decision-making bias (C) in face-matching and fingerprint-matching tasks (known as the “low prevalence” effect, or LPE). This study is a follow-up to one which found that completing feature ratings (e.g., rating the similarity of different fingerprint regions between two prints) reduced the LPE in fingerprint comparisons, compared to completing no ratings. This study will investigate this same question with an improved control group that will complete image quality ratings rather than no ratings). Further, no research has investigated the LPE in fingerprint comparison in non-novice samples. This project will also recruit a group of forensic science trainees that will act as a comparison group to novices to investigate whether feature-comparison may also be an effective training tool to reduce bias in individuals with forensic training.

This study examines a mechanism that may underlie polarisation in testimony evaluation and associated verdicts - populations differing in their evaluations of base-rate estimates that speak to the likely credibility of a witness. For example, a person who believes that false allegations are relatively common compared to the frequency of true complaints, may have an existing expectation that an allegation being made by a complainant is likely to be false (as they believe it is common that such allegations are false). This prior expectation is likely to influence their perceived likelihood that testimony will be accurate, and consequently their assessments of testimony and associated verdicts, as a result of confirmatory biases. In this way, base-rate expectations may at least in part explain the relationship between other factors such as gender and political opinion and juror evaluations of testimony and juror verdicts.