299645 research outputs found
Sort by
On Bekker's many instrument asymptotic framework
No full textA specification test is developed to examine Bekker's many instrument asymptotic framework (Bekker 1994), where the concentration parameter and the number of instruments grow at the same rate as the sample size. The test relies on the fact that the difference between the two-stage least squares (2SLS) estimator and the ordinary least squares (OLS) estimator asymptotically converges to a non-zero limit under Bekker's specification, but otherwise disappears. The limiting distribution of this difference is established within the two specifications, and a delete-d Jackknife procedure is introduced to estimate the asymptotic variance of the difference. Monte Carlo experiments demonstrate the good performance of the test procedure for both single and multiple endogenous variables. Additionally, an empirical illustration to returns to education data indicates the reliability of the test.</p
Evaluating the use of BERT and Llama to analyse classroom dialogue for teachers' learning of dialogic pedagogy
Get PDFClassroom dialogue is crucial for effective teaching and learning, prompting many professional development (PD) programs to focus on dialogic pedagogy. Traditionally, these programs rely on manual analysis of classroom practices, which limits timely feedback to teachers. To address this, artificial intelligence (AI) has been employed for rapid dialogue analysis. However, practical applications of AI models remain limited, often prioritising state-of-the-art performance over educational impact. This study explores whether higher accuracy in AI models correlates with better educational outcomes. We evaluated the performance of two language models—BERT and Llama3—in dialogic analysis and assessed the impact of their performance differences on teachers' learning within a PD program. By fine-tuning BERT and engineering prompts for Llama3, we found that BERT exhibited substantially higher accuracy in analysing dialogic moves. Sixty preservice teachers were randomly assigned to either the BERT or Llama3 group, both participating in a workshop on the academically productive talk (APT) framework. The BERT group utilized the fine-tuned BERT model to facilitate their learning, while the Llama3 group employed the Llama3 model. Statistical analysis showed significant improvements in both groups' knowledge and motivation to learn the APT framework, with high levels of satisfaction reported. Notably, no significant differences were found between the two groups in posttest knowledge, motivation, and satisfaction. Interviews further elucidated how both models facilitated teachers' learning of the APT framework. This study validates the use of AI in teacher training and is among the first to investigate the relationship between AI accuracy and educational outcomes. Practitioner notes What is already known about this topic Given the significance of classroom dialogue, many teacher professional development programmes have been implemented focusing on dialogic pedagogy. To provide timely feedback to teachers, artificial intelligence (AI) techniques are increasingly utilised to investigate classroom dialogue. However, a small proportion of studies have investigated the impacts of AI models in practice, with a predominant focus on pursuing state-of-the-art performance. It is unclear whether more accurate AI models necessarily lead to more positive educational outcomes. What this paper adds This study evaluated the performance of two AI-powered language models, BERT and Llama3, in dialogic move analysis through fine-tuning and prompt engineering. BERT exhibited significantly higher accuracy than Llama3. Through an experimental study, this paper revealed that teachers using either the more accurate BERT model or the less accurate Llama3 model showed substantial improvements in their knowledge and motivation to learn the APT framework and reported high levels of satisfaction. The performance difference between BERT and Llama3 did not cause significant differences in teachers' knowledge, learning motivation, and satisfaction during the learning of the APT framework. Implications for Practice and/or Policy Deep learning models and large language models can be integrated into professional development programs to effectively facilitate teachers' learning of dialogic pedagogy. AI models with moderate performance can also produce impressive outcomes and provide a satisfactory experience. In some scenarios, the manner in which teachers collaborate with AI may be more pivotal than the AI's accuracy.published_or_final_versio
Feature Preserving Shrinkage on Bayesian Neural Networks via the R2D2 Prior
No full textBayesian neural networks (BNNs) treat neural network weights as random variables, which aim to provide posterior uncertainty estimates and avoid overfitting by performing inference on the posterior weights. However, selection of appropriate prior distributions remains a challenging task, and BNNs may suffer from catastrophic inflated variance or poor predictive performance when poor choices are made for the priors. Existing BNN designs apply different priors to weights, while the behaviours of these priors make it difficult to sufficiently shrink noisy signals or they are prone to overshrinking important signals in the weights. To alleviate this problem, we propose a novel R2D2-Net, which imposes the R2-induced Dirichlet Decomposition (R2D2) prior to the BNN weights. The R2D2-Net can effectively shrink irrelevant coefficients towards zero, while preventing key features from over-shrinkage. To approximate the posterior distribution of weights more accurately, we further propose a variational Gibbs inference algorithm that combines the Gibbs updating procedure and gradient-based optimization. This strategy enhances stability and consistency in estimation when the variational objective involving the shrinkage parameters is non-convex. We also analyze the evidence lower bound (ELBO) and the posterior concentration rates from a theoretical perspective. Experiments on both natural and medical image classification and uncertainty estimation tasks demonstrate satisfactory performances of our method
Using cognitive diagnosis to guide workshop design: A preliminary study of examining novice mathematics teachers’ knowledge of dialogic teaching
No full textTeacher knowledge assessment has received little attention in professional development programs. This study addresses this gap by utilizing cognitive diagnosis to guide the workshop design and examine the differences in novice teachers' knowledge of dialogic teaching. The study involved 45 Chinese secondary mathematics teachers who participated in a knowledge-based workshop. Pre- and post-tests demonstrated a significant improvement in teachers' knowledge. Cognitive diagnosis analysis revealed that around half of the participating teachers sufficiently comprehended the concepts of “elaborating,” “reasoning,” and “thinking with others.” The study underscores the potential of cognitive diagnosis as a design strategy to support teachers’ professional growth. </p
The role of artificial intelligence in surgeon-performed ultrasonographic evaluation of cytologically indeterminate thyroid nodules
No full textIntroduction: Evaluating indeterminate thyroid nodules(ITN) is challenging, especially without molecular tests. This study examines whether artificial intelligence (AI) assistance can improve ITN diagnostic accuracy and bridge expertise gaps in surgeon-performed ultrasound. Methods: 134 ultrasound clips from 67 patients with ITN were reviewed by doctors of four levels: endocrine-surgery specialist, senior residents, junior residents, and medical student. After a 2-week wash-out, they re-evaluated the clips using AI-SONIC, an AI platform analyzing ultrasound real-time to predict cancer risk. Performance was validated against final histopathology. Results: Without AI, medical students, junior residents and senior residents performed significantly worse than specialists(AUROC 0.530–0.560 vs 0.771, p Conclusion: AI enhances ultrasound evaluation of ITN by junior surgeons and medical students, elevating their accuracy to expert levels, supporting clinical assessment and medical education.</p
Green partnership across borders: the location-specific institutional and stakeholder pressures on firm participation
No full textFirms operating production and sales across different locations face environmental challenges that regulatory regimes struggle to address. Therefore, regulators have developed voluntary cross-border environmental programs that encourage firm participation. However, the location-specific effects of institutional and stakeholder pressures remain underexplored. By drawing on voluntary club, stakeholder, and institutional isomorphism theories, this study examines how pressure from production sites and headquarters influences firms' participation and the timing of involvement. Using data from cross-border programs between Hong Kong and mainland China's Guangdong province, we find that regulatory pressures from both headquarters (Hong Kong) and production sites (Guangdong province) and peer pressures from competition at production sites significantly increase firm participation. Among all the stakeholder influences in both regions, only regulatory pressures increased Hong Kong-based firms' early involvement in these programs. These findings suggest that policymakers should tailor cross-border programs to address institutional dynamics and leverage local competition and regulatory enforcement to enhance firm participation and improve environmental governance.</p
RIS Assisted Near-Field NOMA Communications: A Security-Fairness Trade-off
No full textA reconfigurable intelligent surface (RIS) assisted near-field secure non-orthogonal multiple access (NOMA) communication system is investigated. In particular, a challenging secure NOMA communication scenario is considered, where the user closing to the RIS is untrusted. Exploiting the near-field beamfocusing capability, a far-to-near successive interference cancellation (SIC) operation is employed to facilitate the secure NOMA communications. Based on this, the trade-off between security and fairness is characterized by maximizing the weighted sum of security capacity and minimum capacity. An alternating optimization based algorithm is developed to solve this highly coupled problem, where RIS beamforming and power allocation are optimized using semidefinite relaxation and successive convex approximation methods, respectively. Numerical results demonstrate: (1) The secure communication for the far user can be achieved in the near field but is impossible in the far field; (2) As the distance between two users increases, the security capacity initially increases and then decreases, while the minimum capacity continuously declines.</p
Enhancing Preservice Teachers’ Use of Dialogic Teaching and Dynamic Visualizations in Mathematics Classes: Bridging the Knowing–Doing Gap
Get PDFTalking productively with students and sufficiently integrating technology into mathematics classrooms have long been regarded as two hurdles for mathematics teachers. To enhance preservice mathematics teachers’ dialogic teaching skills and integration of GeoGebra-scaffolded dynamic visualizations, this study proposed and examined the effectiveness of a video-based professional development (PD) approach supported by a digital platform called Classroom Discourse Analyzer. Adopting the QUAL-quan method, one preservice teacher was selected as a representative case. The results showed that the PD approach effectively improved the preservice teacher’s declarative knowledge and teaching practice of using lower-order talk moves. The preservice teacher’s self-awareness and self-reflection on dialogic teaching informed her future practices. Furthermore, the preservice teacher was able to integrate GeoGebra-scaffolded dynamic visualizations into the instructions with different pedagogical decisions, reflecting how she reacted to student errors and the affordances and constraints of dynamic visualizations. This study suggests that the theoretically robust PD approach can serve as pioneering work in simultaneously promoting dialogic teaching and GeoGebra-scaffolded dynamic visualizations among preservice mathematics teachers. It also demonstrates the potential of integrating digital technologies to design hybrid PD programs to enhance preservice teachers’ self-reflection and facilitate improvement in their future teaching practices.published_or_final_versio
Contract-Inspired Contest Theory for Controllable Image Generation in Mobile Edge Metaverse
No full textThe rapid advancement of immersive technologies has propelled the development of the Metaverse, where the convergence of virtual and physical realities necessitates the generation of high-quality, photorealistic images to enhance user experience. However, generating these images, especially through Generative Diffusion Models (GDMs), in mobile edge computing environments presents significant challenges due to the limited computing resources of edge devices and the dynamic nature of wireless networks. This paper proposes a novel framework that integrates contract-inspired contest theory, Deep Reinforcement Learning (DRL), and GDMs to optimize image generation in these resource-constrained environments. The framework addresses the critical challenges of resource allocation and semantic data transmission quality by incentivizing edge devices to efficiently transmit high-quality semantic data, which is essential for creating realistic and immersive images. The use of contest and contract theory ensures that edge devices are motivated to allocate resources effectively, while DRL dynamically adjusts to network conditions, optimizing the overall image generation process. Experimental results demonstrate that the proposed approach not only improves the quality of generated images but also achieves superior convergence speed and stability compared to traditional methods. This makes the framework particularly effective for optimizing complex resource allocation tasks in mobile edge Metaverse applications, offering enhanced performance and efficiency in creating immersive virtual environments
Influence of New Internet Usage on Depressive Symptoms Among Older Adults: Does the Effect Vary in People with Different Economic Status?
Get PDFThis study investigated the longitudinal effect of new internet usage on depressive symptoms and whether economic status modified this association. Data were from the China Health and Retirement Longitudinal Study, involving 5,259 participants who were 60 +, did not use the internet in 2015, and were followed up in 2018. Linear regression with standard errors clustered at the city level was employed. We found that new internet usage was associated with less depressive symptoms, and the association was more profound among the poor participants. It implies that bridging the digital divide requires special attention to those with disadvantaged economic status.published_or_final_versio