https://hdl.handle.net/11436/7331 2026-06-29T00:15:36Z https://hdl.handle.net/11436/10967 Can a large language model judge a child’s statement?: a comparative analysis of ChatGPT and human experts in credibility assessment Karataş, Zeki Purpose: This study investigates the inter-rater reliability between human experts (a forensic psychologist and a social worker) and a large language model (LLM) in the assessment of child sexual abuse statements. The research aims to explore the potential, limitations, and consistency of this class of AI as an evaluation tool within the framework of Criteria-Based Content Analysis (CBCA), a widely used method for assessing statement credibility. Materials and methods: Sixty-five anonymized transcripts of forensic interviews with child sexual abuse victims (N = 65) were independently evaluated by three raters: a forensic psychologist, a social worker, and a large language model (ChatGPT, GPT-4o Plus). Each statement was coded using the 19-item CBCA framework. Inter-rater reliability was analyzed using Intraclass Correlation Coefficient (ICC), Cohen’s Kappa (κ), and other agreement statistics to compare the judgments between the human-human pairing and the human-AI pairings. Results: A high degree of inter-rater reliability was found between the two human experts, with the majority of criteria showing “good” to “excellent” agreement (15 of 19 criteria with ICC >.75). In stark contrast, a dramatic and significant decrease in reliability was observed when the AI model’s evaluations were compared with those of the human experts. The AI demonstrated systematic disagreement on criteria requiring nuanced, contextual judgment, with reliability coefficients frequently falling into “poor” or negative ranges (e.g. ICC = -.106 for “Logical structure”), indicating its evaluation logic fundamentally differs from expert reasoning. Discussion: The findings reveal a profound gap between the nuanced, contextual reasoning of human experts and the pattern-recognition capabilities of the LLM tested. The study concludes that this type of AI, in its current, prompt-engineered form, cannot reliably replicate expert judgment in the complex task of credibility assessment. While not a viable autonomous evaluator, it may hold potential as a “cognitive assistant” to support expert workflows. The assessment of child testimony credibility remains a task that deeply requires professional judgment and appears far beyond the current capabilities of such generative AI models. 2025-01-01T00:00:00Z https://hdl.handle.net/11436/10717 Punitive discipline in Turkish secondary schools: a qualitative exploration of students’, teachers’, and parents’ perspectives Selçuk, Ozan; Özkan, Yasemin Disciplinary practices play a critical role in shaping students’ academic, emotional, and social development. Research shows that use of punitive school disciplinary measures is associated with negative life outcomes, including post-traumatic stress, depression, aggressive behavior, poorer academic performance, and increased school dropout. This qualitative study, using interpretative phenomenological analysis, explores the experiences and perspectives of students, teachers, and parents regarding punitive disciplinary actions in Turkish secondary schools. Findings indicate that the lack of a clear definition of discipline and an emphasis on punishment leads to student alienation and erosion of school community. Differing perceptions of discipline between teachers and parents further create inconsistencies in disciplinary practices and strain relationships. This research underscores the urgent need for policy reforms aimed at integrating restorative justice practices to create a supportive educational environment that promotes student well-being and academic success. 2025-01-01T00:00:00Z https://hdl.handle.net/11436/9617 The mediating effect of rumination in the relationship between secondary traumatic stress and self-compassion in psychosocial support providers Karataş, Zeki; Tüccar, Esin The aim of this study is to determine whether rumination has a mediating role in the relationship between secondary traumatic stress and self-compassion levels of professionals who provide psychosocial assistance to survivors after the earthquake in Turkey on 6 February 2023. The research was organized in accordance with the relational screening model. Within the scope of Turkey Disaster Response Plan, 153 professionals including social workers, psychologists and psychological counselors who provide psychosocial assistance to earthquake survivors participated in the study. The research data were collected through an online survey link using the Secondary Traumatic Stress Scale, Self-Compassion Scale, Ruminative Reaction Scale and Personal Information Form. In the analysis of the data; descriptive statistics, Pearson correlation analysis and PROCESS Macro (Model 4) were used in the mediation test. According to the results of the study, the direct (β = −0.204) and indirect (β = −0.175) effects of self-compassion on secondary traumatic stress are statistically significant and negative. Therefore, it was determined that rumination played a significant mediating role in this interaction. Accordingly, it is understood that increasing self-compassion and regulating rumination reactions are important in reducing the secondary traumatic stress levels of professionals providing psychosocial assistance. There is a need for more research into the factors that influence the secondary traumatic stress of mental health workers who intervene with traumatized people. 2024-01-01T00:00:00Z https://hdl.handle.net/11436/9517 Untold struggles: unveiling employment challenges faced by asylum seekers in Türkiye Güzel, Bekir; Selçuk, Ozan; Aygüler, Edip In the 21st century, conflicts all over the world still cause people to leave their homes and/or countries and seek refuge in other countries. Due to Türkiye’s geographical location and history as a former empire, it has faced numerous requests for asylum since its founding. This research was conducted using the principles of the phenomenological approach and by interviewing 15 asylum seekers (6 of Afghanistan and 9 of Iraqi nationality) and 17 non-governmental organization (NGO) professionals working with asylum seekers in Ankara in order to delve into the challenges asylum seekers face in accessing employment and the problems in employment. Data was analysed using thematic analysis. The findings of this study are categorized into three themes: (1) Challenges faced by asylum seekers in the labour market; (2) Problems in assessing employment from a professional perspective; and (3) Common misconceptions and priority issues in the employment of asylum seekers. A total of 16 sub-themes were included under these themes. It has been determined that asylum seekers in Türkiye are forced to work in unregistered and temporary jobs, under severe conditions such as long working hours, with a salary lower than the minimum wage. It is also concluded that asylum seekers are inadequate in terms of language skills and lack of professional skills that facilitate access to the labour market; some asylum seekers cannot receive their wages from employers and they do not know any legal mechanisms that they can apply and claim their rights in such cases. 2024-01-01T00:00:00Z