Sexsomnia is a parasomnia that has been successfully used in court as a legal defense for sexual assault (Grøndahl et al., 2017). This diagnosis is unique in that it can serve as a double-edge sword; working to displace guilt from the defendant, yet also as an aggravating factor when it comes to punishment. In our study (N = 282), participants read about a sexual assault and were asked to make legal decisions about the case. Our participants attributed the least amount of responsibility to the defendant in the case involving a Sexsomnia defense compared to a Voluntary Intoxication or Consent defense strategy. However, participants also recommended the defendant go to prison more often when the defense was Sexsomnia compared to the other two defenses. Our results suggest a paradoxical effect of mental illness as perceived in the courtroom. Qualitative data to support these legal decisions will also be discussed.

In the courtroom, jurors are often asked to weigh up two conflicting accounts of an event (e.g. in a so-called “he said, she said” contested sexual encounter). In such cases, memory experts are regularly called upon by the defence to offer testimony on the unreliable nature of memory. This may encourage the unsupported assumption that it is only accusers who suffer from flawed recall, not the accused. In this talk, we will share findings from a series of well-powered, pre-registered studies investigating this issue. Our findings suggest that both accused and accusers are equally susceptible to post-event misinformation and memory distortion. We also draw from our mock-jury studies to provide specific recommendations for how these findings can be implemented in practice, by both legal professionals and memory scientists serving as expert witnesses.

Introduction
Victim statements plays a crucial role in sex offense cases, because forensic evidence is often lacking. In order to avoid false convictions, this brings about the complex task for judges to evaluate the statement on its credibility, its evidentiary value, and whether it is supported by corroborative evidence. We examined which indicators are named in Dutch rulings of victim statements of sexual abuse. Prior research on this topic is limited in the number of cases (n = 33-79), the lack of standardization in scoring, and selection bias (i.e., selecting cases based on credibility assessment criteria).
Method
We scraped all Dutch court rulings from 2023 on rape and/or sexual assault in which a credibility assessment was made and included 526 rulings for analyses. All cases were scored with a coding scheme, developed on the credibility assessment literature and piloted on cases from 2024. For reliability purposed, 10 percent of cases were double coded. Inter-rater reliability was moderate to high for the variables used in the analyses.
Results
Suspects were predominantly male (98%). Victims were predominantly female (87%). Victims and suspects were often acquainted before the crime (78%). Most statements were deemed credible (87.50%), with judges expressing some (6.80%) to sincere (4.70%) doubt on a minority of the statements. The most often named criteria in the credibility assessment were: Consistency (89%; often specifying ‘on core aspects’), detailedness (63%), authentic impression (32%), emotionality during statement (13%), signs of trauma (12%), and completeness (9%).
Discussion
In evaluating the credibility of victim statements of rape and/or sexual assault, Dutch judges often name criteria that are evidence-based (consistency on core aspects; detailedness, completeness). However, reliance on impressions of authenticity, emotionality, and trauma have no diagnostic value for determining credibility. Of note, while the Dutch supreme court ruled that statements should be evaluated on Consistency, Completeness (arguably this may also entail Detailedness) and Accuracy (i.e., Statement-fact consistency), only 3% of court rulings named all 3 indicators. Judicial decision-making may benefit from a more systematic use of valid indicators.

A lack of physical resistance during sexual assault can negatively impact the credibility of the complainant’s account, potentially leading to a not guilty verdict in criminal trials. In Australia, judges can provide jurors with instructions at any point during the trial to counteract victim-blaming attitudes. This study examined whether the timing of judicial instructions influences verdicts, as well as the perceived credibility of the complainant and defendant among mock jurors. A total of N = 215 Australian community members were randomly assigned to one of five conditions: receiving educational judicial instructions on the role of physical resistance in sexual assault either before, during, after, both before and after cross-examination, or receiving no judicial instructions. Participants read a trial transcript depicting an alleged acquaintance rape.

*Study is currently ongoing with results and conclusion pending*