Online counselling services have seen increased use in recent years, providing critical emergency mental health support. These interactions are typically long, complex, and varied in the dialogue between help seekers and counsellors. The lack of domain-specific models, especially in low-resource languages, poses a significant challenge for the automatic detection of suicide risk in online chat services for mental health support. To address this challenge, our approach adapts a general-purpose large language model (LLM) to the suicide prediction task that employs a two-stage classification architecture to deal with sparse and imbalanced data. It extends the state of the art by: (1) incorporating psychological theory into model training and (2) capturing key aspects of conversation structure in counselling sessions. We evaluate the performance of the proposed LLM against the state-of-the-art LLMs for suicide detection on thousands of conversations in the Hebrew language from a leading national online counselling service in Israel. Results show that the proposed LLM outperformed existing state-of-the-art approaches in detecting suicide risk, as measured by relevant literature metrics. Moreover, the LLM outperforms other approaches even in the early stages of a conversation, which is crucial for real-time detection in practice. We also discuss the ethical implications of combining LLMs in counselling services. The contributions of this work are (1) extending existing LLM architectures to incorporate domain-specific information; (2) evaluating LLM technologies in the context of socially relevant problems; and (3) introducing novel LLM tools for resource-constrained languages.

Team science is central to clinical and translational research; however, systematic evaluation of collaborative efforts remains inconsistent and underdeveloped. To better understand current team science evaluation practices within clinical and translational science programs, we conducted a structured cross-sectional survey of team science and evaluation professionals. We analyzed quantitative data using descriptive statistics and qualitative responses through thematic analysis. Among participating organizations, the majority have implemented team science evaluations, predominantly using mixed-methods approaches combining quantitative metrics and qualitative assessments. Evaluation findings were primarily used to inform programing, improve team functioning, and secure funding. Reported challenges fell into four key areas: methodology; implementation; data analysis; and organization. Facilitators included: methodological enhancements, organizational support, collaborative approaches, and infrastructure elements. Participants emphasized using interim measures (e.g., team dynamics and satisfaction) that move beyond traditional outcome measures so that evaluations better reflect how teams interact, adapt, and progress as they develop. While team science evaluation adoption is substantial among leading translational research institutions, significant methodological gaps persist. Future directions focus on developing standardized frameworks with local flexibility, creating validated instruments, utilizing interim process measures, and demonstrating return on investment (ROI) to advance both evaluation science and translational outcomes.

Depuis plus d’une décennie, le Mali est confronté à une crise sécuritaire complexe ayant conduit au déploiement de plusieurs missions internationales de paix, dont la Minusma. Si cette dernière avait pour objectif de soutenir la stabilisation du pays, elle a progressivement fait l’objet de critiques croissantes au sein de la population.

Cet article propose d’analyser les attentes des Maliens à l’égard de la Minusma et d’examiner comment leurs perceptions ont influencé son acceptabilité ainsi que l’évaluation de son efficacité. En mobilisant l’approche de l’évaluation qualitative des politiques publiques, l’étude considère cette mission comme une politique publique internationale de paix. L’analyse repose sur une enquête qualitative réalisée à Bamako en 2020 et 2021. Elle révèle un décalage entre les attentes locales et le cadre de référence de la mission ainsi que l’importance de la légitimité locale dans la conduite, voire la poursuite, des opérations de paix.

The 2025 Evaluation Special Interest Group (SIG) meeting at the Association for Clinical and Translational Science conference brought together clinical and translational science (CTS) professionals to address evolving challenges in translational science evaluation. The meeting presentations and discussions addressed concept mapping for commonly used metrics, continuous quality improvement (CQI) practices, translational science impact evaluation, and evaluator toolkit development. Key themes and lessons learned included the tension between institution-specific and network-wide evaluation goals, the need for standardized yet flexible evaluation frameworks, and persistent barriers including limited staffing capacity and data ownership challenges. Facilitators identified included diverse CQI approaches, the evolving frameworks, and collaborative evaluation practices. Convened during a time of increasing research funding uncertainty and accountability, the meeting underscored the urgency of strengthening evaluation capacity to sustain the impact of CTS, highlighting both the enduring value of heterogeneous evaluation approaches and the critical need for coordinated CTS evaluation strategies to demonstrate impact and secure continued funding support.

Narrators like to highlight important events in their stories. In some languages, they may shift to first- or second-person pronouns to refer to third-person referents in order to do so. Such pronoun shifts show functional parallels with tense shifts like the historical present, as both highlight events through shifts in deictic categories. Longacre (1983:138-39) discusses the parallels between person and tense shifts in his account of narrative peak, that is, the formal marking of important narrative events. Labov (1972) analyzes similar strategies as internal evaluations. Person shifts constitute a phenomenon of the discourse-syntax interface and present a clear case of discourse structure influencing grammar. Both person shifts themselves and their motivation in narrative structure have been little investigated. The article reviews person shifts in a number of languages reported in the literature and analyzes in detail the characteristics of this discourse strategy in Saliba-Logea, an Oceanic language of Papua New Guinea. The study contributes to the growing body of research on pronouns and person markers, and on referring expressions more generally, by adding a new angle of investigation. Previous studies have tended to focus on the morphosyntactic choices of referring expressions and their motivations, that is, on the choices between lexical nouns, free vs. bound pronouns, and so forth. The present study focuses on the paradigmatic choices between different person forms within one and the same morphosyntactic expression type. In doing so it offers a new perspective on pronoun choice and the factors influencing it crosslinguistically. While some types of person shift appear to be rare, overall, the strategy of person shift at narrative peak seems to constitute a solid crosslinguistic phenomenon.

Grassroots organizations have a central place in service provision within many urban communities, particularly as they work with young people. Enhancing the potential of youth serving organizations is vital as many of these small groups lack infrastructure and resources. One organization has gone beyond grant making to organizations to include providing technical assistance, then moving to a model of capacity building to improve their ability to fulfill their mission. This paper describes the efforts of New Detroit to improve the infrastructure of youth serving grassroots organizations through a multi-faceted model and the outcomes evidenced from the first cohort. In addition we explore the contradictions inherent in working with small groups. How do you improve effectiveness without destroying the “homegrown” character these groups possess? The model presented here shows the value of empowering small grassroots groups to build capacity.

The ERC has been a pivotal innovation in the set of funding instruments that the European Commission has established for fostering research and innovation in the European Union. With more than 8000 projects funded so far, it is worth asking about empirical evidence regarding the ERC’s specific impact on the social sciences. This article provides some basic data, along with descriptive statistics, on the social scientists who have been sitting on ERC evaluation panels, and on ERC-funded research projects from the social sciences. The article ends with a discussion of the data and poses questions for further investigations.

Non-profit organisations are under increasing pressure to demonstrate their social impact. This paper examines the experience and behaviour of non-profit organisations in the UK in relation to a demand for social impact evaluations. External resource providers request organisations to present evidence on how resources are used and what organisations have achieved. While most organisations are willing to comply and accept this control, they can also resist through using their discretion in deciding what to measure, how to measure and what to report. Non-profit organisations can proactively and voluntarily use social impact measurement for learning and promotional purposes, and as a way of exerting control over their environment. The analysis develops the concept of strategic decoupling to explain the differences observed between what organisations are asked to do, what they plan to do and what they are doing in practice.