Taking advantage of unique longitudinal task data from the Berea Panel Study, we provide a new examination of the gender wage gap, paying particular attention to gender differences in types of work experience. Access to longitudinal, individual-level job-task information, along with unique time-allocation information, allows us to produce quantitative measures of current and past tasks. We provide the first empirical evidence on gender differences in time spent on tasks and show that gender differences in task-specific experience, in particular high-skilled information experience, are important for predicting the widening of the gender wage gap over the career.
Frontier AI systems are increasingly capable and deployed in high-stakes multi-agent environments. However, existing AI safety benchmarks largely evaluate single agents, leaving multi-agent risks such as coordination failure and conflict poorly understood. We introduce GT-HarmBench, a benchmark of 2,009 high-stakes scenarios spanning game-theoretic structures such as the Prisoner's Dilemma, Stag Hunt and Chicken. Scenarios are drawn from realistic AI risk contexts in the MIT AI Risk Repository. Across 15 frontier models, agents choose socially beneficial actions in only 62
IntroductionContributing to global public health within the context of interprofessional education is an imperative of healthcare programs, with expectations for academic and experiential activities in health promotion, disease prevention, and rehabilitation. This study evaluated the effectiveness of a Collaborative Online International Learning (COIL) pedagogy implemented across three countries (Japan, South Korea, and the United States) to teach baccalaureate undergraduates about remote island healthcare. The program sought to facilitate cross-cultural understanding and global competency related to remote island healthcare.MethodsThis study used a descriptive design with a sample of 144 allied health undergraduate students. Over 4 weeks, participants completed structured online assignments, prepared collaborative presentations, and engaged in recorded group discussions focusing on healthcare challenges faced by remote island populations. The Beliefs, Events, and Values Inventory (BEVI) was used for pretest and posttest data on participants’ cross-cultural understanding and global competency.ResultsJapanese participants demonstrated meaningful positive changes in cross-cultural understanding and global competency, whereas South Korean and US participants showed declines in these outcomes.DiscussionThese findings suggest that COIL programs may have differential effects on cross-cultural understanding and global competency depending on participants’ contexts. These results should be interpreted with caution due to differences in baseline scores across groups.
This study examines an evaluation that used Schwandt's post-normal evaluation design to create a structure that would capture the uncertainties and complexities encountered in implementing a novel professional development (PD) program at four university teacher preparation programs. Three distinct levels of evaluation were addressed in the study: (1) the evaluation of the PD implementation (RQ 1-2), the evaluation within the project conducted by the PD team internally (RQ 3-4), and an evaluation of the evaluation design itself (RQ 5-6). The PrimeD framework guided the PD implementation and all levels of the evaluation. PrimeD supported continuous collaboration among and within evaluator and implementation teams, which strengthened the focus and consistency of the project vision and goals across all four institutions. Monthly Networked Improvement Community (NIC) meetings provided a forum for participants to share ideas and develop plans for improving their professional practice using Plan-Do-Study-Act (PDSA) cycles. Conclusions, best practices, and contextual challenges were shared across institutions by the participants (teacher candidates, program alumni, classroom mentor teachers, and field experience supervisors). Observations of NIC meetings and focus groups of participants provided the data used to adjust approaches to the implementation and evaluation. Throughout this 5-year PD project, evaluation and research grew to become an integral part of everyday lesson development conversation. The PD implementation team also shared findings across institutions, resulting in modifications of the teacher preparation and early career programs across all four sites that led to better alignment across programs while retaining necessary contextual differences.
Organizations invest heavily in formal Knowledge Management (KM) systems to capture organizational wisdom, yet knowledge hoarding and inefficiency persist. While existing quantitative literature has mapped the antecedents of sharing, it often fails to explain the micro-foundational resilience of hoarding under operational pressure. This exploratory, interpretivist multiple-cases study investigates knowledge behaviors within two high-technology organizations in the medical device sector. Data were collected through thirty-six in-depth semi-structured interviews, nonparticipant observation, and document analysis. This data set establishes a foundational baseline for “performative compliance” that persists in modern digital environments. Analysis reveals a profound institutional decoupling between formal KM policy and actual practice. Employees engaged in performative compliance, gaming formal training systems to satisfy audit trails without genuine learning. Knowledge hoarding was identified not as malicious behavior, but as a rational coping mechanism—a calculation of adaptive cost in response to time famine. Consequently, employees relied on exclusionary informal networks, increasing strategic fragility. The study concludes that the “click–click” culture of modern compliance is a structural inevitability when governance systems ignore the temporal cost of sharing. To transition to genuine sharing, organizations must treat time as a critical knowledge resource.
