校园欺凌案件中证词评估的全球评估系统测试。
完整摘要 · 中文译文
背景/目的：学龄儿童和青少年中校园欺凌受害的流行率估计在5%（采用严格的流行率分类标准）至33%（采用宽松的流行率评估标准）之间。欺凌受害涉及行政（学校环境）、民事（受害人赔偿）和刑事（刑事责任）层面的决策，这些决策必须基于对指控真实性的确信。本研究设计了一项高保真模拟研究，旨在检验全球评估系统（GES）在欺凌指控案件中对证词的评价。方法：共154名13至18岁的青少年（86名男孩）被随机分配到真实生活负面经历记忆的证词条件或编造的欺凌事件记忆条件。对陈述记录（证词）进行内容分析，使用GES的证据有效性类别和真实生活经历记忆的内容类别。结果：结果显示GES内容类别系统具有极好的信度，α = .874，95% CI [.843, .901]。研究结果还支持GES内容类别与真实生活经历记忆之间的关联。相反，编造的欺凌事件记忆与GES内容类别之间未发现显著关联，且在此条件下观察到信度不足，α = .476，95% CI [.289, .632]。最后，计算了严格的决策标准（基于证据的效标效度），以保证符合规定。 符合无罪推定原则（任何无辜者不应被定罪；I类错误/假阳性 = 0）：在完整模型中编码≥10个内容类别，在较高流行率模型中编码≥8个（与真实生活经历记忆显著相关的内容类别），在稳健模型中编码≥5个（与真实生活经历记忆显著相关且效应量为中等或更高的内容类别）。结论：这些发现支持GES内容分析分类系统在校园欺凌受害真实证言验证（法医任务）中的效度和信度，为严格遵循无罪推定原则提供了科学证据（标准效度），但不适用于虚构记忆的分类（调查任务）。
完整摘要 · 英文原文
Background/aim: The prevalence of school bullying victimization has been estimated between 5% (strict criteria for prevalence classification) and 33% (lenient criteria for prevalence assessment) of children and adolescents. Bullying victimization entails decision-making at the administrative (school setting), civil (victim compensation), and criminal (criminal responsibility) levels, which must be based on certainty about the truthfulness of the allegation. A high-fidelity simulation study was designed with the aim of testing the evaluation of testimony of the Global Evaluation System (GES) in cases of bullying allegations. Method: A total of 154 adolescents aged from 13 to 18 years, 86 boys, were randomly assigned to the condition of testimony of a memory of real-life negative experiences or a fabricated memory of a bullying event. The protocols (statements) were subjected to a content analysis using the categories of validity of the evidence and the content categories of memories of real-life experiences of the GES. Results: The results showed excellent reliability, α = .874, 95% CI [.843, .901], for the GES content category system. The findings also supported the association between the GES content categories and memories of real-life experiences. Conversely, no significant association was found between fabricated memories of a bullying event and the GES content categories, while a lack of reliability was observed, α = .476, 95% CI [.289, .632], in this condition. Finally, strict decision criteria were computed (evidence-based criterial validity) that guarantees compliance with the principle of presumption of innocence (no innocent person should be convicted; type I error/false positives = 0): coding of ≥ 10 content categories in the complete model, ≥ 8 in the higher-prevalence model (content categories associated significantly with memories of real-life experiences), and ≥ 5 in the robust model (content categories associated significantly with memories of real-life experiences with a moderate or higher than moderate effect size). Conclusions: These findings support the validity and reliability of the GES content analysis categorical system for the validation of truthful testimonies of school bullying victimization (forensic task), providing scientific evidence (criterial validity) to ensure the strict compliance of the principle of presumption of innocence, but not for the classification of fabricated memories (investigative task).