Data quality is a key input in efforts to link individuals across census records. We examine the extreme case of low data quality by identifying US census enumerators who fabricated entire families. We provide clear evidence of fake people included in the 1920 US Census for Homestead, Pennsylvania. We use the features of this case study to identify other places where information in the census may have been falsified. We develop an automated approach that identifies census sheets that have much lower match rates to other census records than would be expected, given the characteristics of the people recorded on each sheet. We perform a hand-check on the suspicious sheets using standard genealogy tools and identify at least 90 sheets where the entire census sheet appears to have been fabricated.

Integrated pest management (IPM) is the dominant pest management paradigm in agriculture, and adoption of IPM is a policy goal at various levels of government. However, with over 67 definitions, what is considered IPM varies, and some implementations of IPM fail to achieve desired results (e.g., reduced pesticide use). The natural complexity of agriculture and pest management leads growers to rely on professionals, usually certified crop advisors (CCAs), to help make management decisions. Though communication with crop advisors is one way to improve IPM, this effort requires an understanding of CCA education, knowledge, and information sources. Previous surveys in North America found most crop advisors were very experienced (>20 years), which may present a concern for the adequate supply of crop advisors in the future, since they will very likely retire in the coming decades. In this survey, nearly 95% of CCAs earned a bachelor’s level degree (or higher). Independent crop advisors reported spending about 43% less time communicating with growers than CCAs employed by consultancies. Most crop advisors gave appropriate, but incomplete, definitions of IPM, and adoption of specific guidelines (i.e., scouting recommendations for red sunflower seed weevil, Smicronyx fulvus) was also low. Some CCAs expressed an opinion that universities (and the federal government) were less valued than other sources of information (e.g., their own CCA network). Collectively, survey responses show room for improvement to IPM through CCA education, but because crop advisors are most likely to be influenced through their network of peers, outreach might be best accomplished through targeting early adopters of IPM practices among the CCA population.

Le présent article propose une caractérisation du phénomène linguistique de la distanciation, qui permet au locuteur de se désengager d’un premier contenu discursif, à partir de l’examen d’un certain nombre de propriétés constitutives. Nous prendrons comme témoins trois marqueurs de discours du français contemporain, formés sur le verbe dire: c’est vite dit, c’est beaucoup dire, c’est toi qui le dis. Notre conviction est que l’étude en parallèle des caractéristiques spécifiques permettant de distinguer une relation discursive, d’une part, et, de l’autre, du fonctionnement sémantico-pragmatique des marqueurs de discours, apportera un éclairage fructueux sur la question.

Sentiment analysis and stance detection are key tasks in text analysis, with applications ranging from understanding political opinions to tracking policy positions. Recent advances in large language models (LLMs) offer significant potential to enhance sentiment analysis techniques and to evolve them into the more nuanced task of detecting stances expressed toward specific subjects. In this study, we evaluate lexicon-based models, supervised models, and LLMs for stance detection using two corpuses of social media data—a large corpus of tweets posted by members of the U.S. Congress on Twitter and a smaller sample of tweets from general users—which both focus on opinions concerning presidential candidates during the 2020 election. We consider several fine-tuning strategies to improve performance—including cross-target tuning using an assumption of congressmembers’ stance based on party affiliation—and strategies for fine-tuning LLMs, including few shot and chain-of-thought prompting. Our findings demonstrate that: 1) LLMs can distinguish stance on a specific target even when multiple subjects are mentioned, 2) tuning leads to notable improvements over pretrained models, 3) cross-target tuning can provide a viable alternative to in-target tuning in some settings, and 4) complex prompting strategies lead to improvements over pretrained models but underperform tuning approaches.

Mental health legislation in the Middle East varies significantly in terms of scope, content and implementation. Some countries have rights-based laws aligning with international standards, but others experience significant gaps in legal protections, regulatory oversight and the integration of community-based care. In conflict-affected regions, the disparity between legislative goals and the reality of functional mental health systems is even more pronounced. This paper examines the existing legal frameworks in Middle Eastern countries, emphasising the urgent need for harmonized, enforceable, and contextually relevant legal frameworks to promote equitable and rights-based psychiatric care throughout the region.

Deep generative modeling is a powerful framework in modern machine learning, renowned for its ability to use latent representations to predict and generate complex high-dimensional data. Its advantages have also been recognized in psychometrics. In this article, we substantially extend the deep cognitive diagnostic models (DeepCDMs) in Gu (Psychometrika, 89:118–150, 2024) to challenging exploratory scenarios with deeper structures and all $\mathbf {Q}$-matrices unknown. The exploratory DeepCDMs can be viewed as an adaptation of deep generative models (DGMs) toward diagnostic purposes. Compared to classic DGMs, exploratory DeepCDMs enjoy critical advantages, including identifiability, interpretability, parsimony, and sparsity, which are all necessary for diagnostic modeling. We propose a novel layer-wise expectation–maximization (EM) algorithm for parameter estimation, incorporating layer-wise nonlinear spectral initialization and $L_1$ penalty terms to promote sparsity. From a parameter estimation standpoint, this algorithm reduces sensitivity to initial values and mitigates estimation bias that commonly affects classical approaches for deep latent variable models. Meanwhile, from an algorithmic perspective, our method presents an original layer-wise EM framework, inspired by modular training in DGMs but uniquely designed for the structural and interpretability demands of diagnostic modeling. Extensive simulation studies and real data applications illustrate the effectiveness and efficiency of the proposed method.

Experts conducted a free statewide series on antibiotic stewardship for hospital, outpatient, and long-term care settings. In total, 366 participants from 244 sites represented 66% of counties in the state. Furthermore, 62% worked in nonmetropolitan counties, and 55% were from counties with medium-to-high social vulnerability, demonstrating reach into diverse sites.

This paper investigates the extent and modalities of Latin and Greek teaching in early medieval British monastic communities by examining the indirect evidence offered by the manuscript known as the Liber Commonei, part of the composite manuscript Oxford Bodleian Auct. F.4.32. Using the patterns and nature of Latin and Old Welsh glosses as they appear in the manuscript, it is argued that, as expected, the monks would learn Classical Latin with the aid of Vulgar Latin and vernacular glosses and that they would tackle texts of a gradually higher complexity, conversely reducing their reliance on glosses. They would then proceed to learn Greek, using biblical excerpts (in Greek and in Latin script) as reference material; by analysing these texts, it is argued here that these British monks in the 9th century worked with the help of a Greek-speaking teacher.

We propose an evolutionary competition model to investigate the green transition of firms, highlighting the role of adjustment costs, state-dependent transition risk, and positive externalities in green technology adoption. Firms base their decisions to adopt either green or brown technologies on relative performance. To incorporate the costs of switching to another technology into their decision-making process, we adopt a novel, ad hoc crafted, replicator dynamics. Our global analysis reveals that increasing transition risk, e.g., by threatening to impose stricter environmental regulations, effectively incentivizes the green transition. Economic policy recommendations derived from our model further suggest maintaining high transition risk regardless of the industry’s level of greenness. Subsidizing the costs of adopting green technologies can reduce the risk of a failed green transition. While positive externalities in green technology adoption can amplify the effects of green policies, they do not completely eliminate the possibility of a failed green transition. Finally, evolutionary pressure reduces the extent of green economic policies required to ensure a successful green transition.

The Council of Europe has very recently adopted the Framework Convention on Artificial Intelligence and Human Rights, Democracy and the Rule of Law. This article provides an initial analysis of the CoE AI Convention. It emphasises the necessity of understanding the CoE AI Convention within the context of its adoption as an international treaty negotiated within the Council of Europe. This context has affected its scope in terms of how the treaty includes the regulation of the usage of AI systems by both public authorities and private actors. The detailed review of the available negotiation documents reveals that the concrete level of protection offered by the Convention has been lowered. This includes the risk-based approach, which shapes the obligations undertaken by States under the treaty. This approach is explained and contrasted with the approach under the EU AI Act. The argument that emerges is that the absence of categorisation of risk levels in the treaty is related to its higher level of abstraction, which does not necessarily imply less robust obligations. The content of these obligations is also clarified in light of the requirement imposed by the treaty of consistency with human rights law. An argument is advanced that the principles formulated in the treaty – human dignity and individual autonomy, transparency and oversight, accountability, non-discrimination, data protection, reliability, risk-management – can offer interpretative guidance for the development of human rights standards.

Blockchain technology has attracted attention from public sector agencies, mainly for its perceived potential to improve transparency, data integrity, and administrative processes. However, its concrete value and applicability within government settings remain contested, and real-world adoption has been limited and uneven. This raises questions regarding the conditions that promote or impede adoption at the institutional level. Fuzzy-set qualitative comparative analysis is employed in this research to explore how the combined effects of national-level regulatory clarity, financial provision, digital readiness, and ecosystem engagement shape patterns of blockchain adoption in the European public sector. Rather than identifying any single factor as decisive, our findings reveal a plurality of institutional paths leading to high adoption intensity, with regulatory certainty and European Union funding appearing most frequently on high-consistency paths. In contrast, digital readiness indicators and national research and development budgets are substitutable, challenging resource-based perceptions of technology adoption and supporting a configurational understanding that accounts for institutional interdependence and contextuality. We argue that policy strategies cannot look for overall readiness but should place key institutional strengths relative to local conditions and public value objectives.

This study examines public-private partnerships (PPPs) in home- and community-based services (HCBS) for older adults in Guangzhou, China, amid growing efforts to increase non-state actor participation and market mechanisms in welfare provision. Based on semi-structured interviews with non-state actors, this research examines both contractual and relational dimensions of PPPs and why achieving their full potential remains challenging. Findings reveal that state actors maintain a leading role in PPPs, with non-state actors primarily positioned as implementers of predefined welfare objectives. The power imbalance embedded in China’s sociopolitical context blurs the boundaries among the Party, state, and society, shaping coordination challenges in PPPs. These structural constraints hinder integrated service delivery and limit the extent to which PPPs can meaningfully leverage the expertise of various actors. This study enhances understanding of PPPs in China’s authoritarian, state-led market context and offers insights into the evolving landscape of welfare socialisation reforms.

Divine simplicity is plausibly seen as a biblical doctrine, given a standard account of the way doctrine is derived from Scripture. The polemic of Jeremiah 10 against ancient Near Eastern mis pî or ‘mouth opening’ rituals involves a commitment to a radical account of divine aseity. In dialogue with Thomas Aquinas and a number of contemporary figures, I suggest this view of divine aseity might plausibly be thought to lead to the inference to divine simplicity.

Acetyl-CoA carboxylase (ACCase)-inhibiting herbicides are primarily applied for controlling grass weeds in broadleaf crops. These herbicides are foliar-active, providing minimal residual weed control. This review aims to summarize 1) the history and use of ACCase-inhibiting herbicides in the United States; 2) ACCase-inhibitor-resistant weeds, their mechanisms of resistance, and management strategies; and 3) the future of ACCase-inhibiting herbicides. Herbicides that inhibit ACCase belong to three chemical families: aryloxyphenoxypropionates, cyclohexanediones, and phenylpyrazoles. They function by inhibiting the enzyme ACCase activity, thereby blocking the first step in de novo fatty acid biosynthesis and thus preventing the production of phospholipids and essential secondary metabolites in susceptible plants. Diclofop-methyl was the first ACCase inhibitor discovered in 1975, and commercialized in 1982 in the United States. Pinoxaden was the last herbicide to be commercialized in 2005. As of 2025, a total of 51 grass weed species have been documented as being resistant to ACCase-inhibiting herbicides worldwide, including 16 in the United States. The resistance in these weeds is attributed to both target-site and non–target site mechanisms. Mixing ACCase-inhibiting herbicides with auxinic herbicides can reduce grass weed control due to antagonistic interactions. Therefore, selecting an appropriate tank-mix partner with an ACCase inhibitor is crucial for achieving broad-spectrum weed control, or a dual-tank precision sprayer could be used. Clethodim is the most widely used ACCase-inhibiting herbicide, with 920,339 kg applied to approximately 16% of soybean crops planted in the United States in 2023, at an average application rate of 179 g ha‒1. A recent discovery, metproxybicyclone, will be the first carbocyclic aryl-dione herbicide from a new ACCase inhibitor family. This novel herbicide will be applied postemergence to control sensitive and ACCase inhibitor-resistant grass weeds in broadleaf crops. Continued research efforts are focused on discovering new ACCase-inhibiting herbicides capable of controlling ACCase inhibitor-resistant grass weeds.