Selected work

Representative Papers

Observing Teachers' Instrumental Pedagogical Orchestration in Synchronous Online Learning: A Multimodal Grid Based on Videoconferencing Traces

Jun 12, 2026

This study addresses the lack of systematic multimodal analytical approaches in synchronous online teaching. Integrating instructional orchestration theory, a multimodal interaction framework, and the data affordances of video conferencing platforms, it proposes the first reproducible, prioritized multimodal observation schema specifically designed for synchronous online classrooms. The schema encompasses observable instructor behaviors such as gestures, posture, gaze, and digital tool usage. The resulting structured observation grid not only fills a critical methodological gap in the field but also provides a practical and actionable analytical framework for future empirical research on online teaching practices.

0 citationsRead paper

AI Governance under Political Turnover: The Alignment Surface of Compliance Design

Apr 22, 2026

This study addresses the challenge of designing compliance mechanisms for probabilistic AI systems in public administration under conditions of political alternation, ensuring their auditability, reproducibility, and legal legitimacy while guarding against strategic misuse by subsequent political actors. It introduces, for the first time, the concept of an “alignment surface” within an AI governance framework that explicitly accounts for government turnover. Drawing on game-theoretic and institutional modeling approaches, the paper formally characterizes the dynamic interactions among government agencies, AI systems, and political successors. The analysis reveals that AI-driven compliance reforms initially intended to enhance regulatory efficacy may, through automation expansion and rule entrenchment, inadvertently heighten long-term risks of political manipulation, thereby offering a theoretical warning about potentially irreversible governance trajectories.

0 citationsRead paper

Distracting from the Epstein files? Media attention and short-run shifts in Trump's Truth Social posts

Nov 14, 2025

This study investigates whether political figures strategically manipulate social media discourse during negative media coverage (e.g., scandals) to divert public attention and thereby undermine democratic accountability. Method: Focusing on Donald Trump’s Truth Social posting behavior during the Epstein scandal, we employ computational text analysis and a time-series design, augmented by synthetic control-based counterfactual estimation and placebo-period simulations for causal inference. Contribution/Results: A one-standard-deviation increase in scandal-related news coverage corresponds to a 0.28-standard-deviation deviation of Trump’s posts from baseline thematic content—an effect persisting approximately four days—providing robust evidence for short-term “distraction-driven communication.” This strategy is contingent upon a supportive media ecosystem, underscoring structural weakening of accountability mechanisms under political polarization. To our knowledge, this is the first study to offer rigorous empirical evidence of deliberate, real-time attention diversion by elected officials, revealing a novel challenge to democratic accountability in the digital age.

0 citationsRead paper

Training for Obsolescence? The AI-Driven Education Trap

Aug 27, 2025

This study identifies a critical skill-mismatch risk arising from AI adoption in education: policymakers overemphasize AI’s short-term pedagogical efficiency gains while neglecting its long-term wage-suppressing effects on AI-complementary skills, leading to growing misalignment between curricular offerings and labor market demands—intensifying with higher AI penetration. Method: We innovatively incorporate the crowding-out effect of non-cognitive skills (e.g., perseverance) into an integrated analytical framework, combining theoretical modeling with empirical data from pilot educational interventions to characterize the dynamic interplay among AI penetration, skill returns, and non-cognitive development. Contribution/Results: Findings reveal that unguided AI deployment in education not only exacerbates skill mismatch unidirectionally but also systematically erodes foundational non-cognitive competencies, thereby inducing long-term human capital depreciation. The model provides a rigorous foundation for evidence-based educational policy under AI-driven labor market transformation.

0 citationsRead paper

Recent publications

Latest Papers

Observing Teachers' Instrumental Pedagogical Orchestration in Synchronous Online Learning: A Multimodal Grid Based on Videoconferencing Traces

Jun 12, 2026

This study addresses the lack of systematic multimodal analytical approaches in synchronous online teaching. Integrating instructional orchestration theory, a multimodal interaction framework, and the data affordances of video conferencing platforms, it proposes the first reproducible, prioritized multimodal observation schema specifically designed for synchronous online classrooms. The schema encompasses observable instructor behaviors such as gestures, posture, gaze, and digital tool usage. The resulting structured observation grid not only fills a critical methodological gap in the field but also provides a practical and actionable analytical framework for future empirical research on online teaching practices.

0 citationsRead paper

AI Governance under Political Turnover: The Alignment Surface of Compliance Design

Apr 22, 2026

This study addresses the challenge of designing compliance mechanisms for probabilistic AI systems in public administration under conditions of political alternation, ensuring their auditability, reproducibility, and legal legitimacy while guarding against strategic misuse by subsequent political actors. It introduces, for the first time, the concept of an “alignment surface” within an AI governance framework that explicitly accounts for government turnover. Drawing on game-theoretic and institutional modeling approaches, the paper formally characterizes the dynamic interactions among government agencies, AI systems, and political successors. The analysis reveals that AI-driven compliance reforms initially intended to enhance regulatory efficacy may, through automation expansion and rule entrenchment, inadvertently heighten long-term risks of political manipulation, thereby offering a theoretical warning about potentially irreversible governance trajectories.

0 citationsRead paper

Distracting from the Epstein files? Media attention and short-run shifts in Trump's Truth Social posts

Nov 14, 2025

This study investigates whether political figures strategically manipulate social media discourse during negative media coverage (e.g., scandals) to divert public attention and thereby undermine democratic accountability. Method: Focusing on Donald Trump’s Truth Social posting behavior during the Epstein scandal, we employ computational text analysis and a time-series design, augmented by synthetic control-based counterfactual estimation and placebo-period simulations for causal inference. Contribution/Results: A one-standard-deviation increase in scandal-related news coverage corresponds to a 0.28-standard-deviation deviation of Trump’s posts from baseline thematic content—an effect persisting approximately four days—providing robust evidence for short-term “distraction-driven communication.” This strategy is contingent upon a supportive media ecosystem, underscoring structural weakening of accountability mechanisms under political polarization. To our knowledge, this is the first study to offer rigorous empirical evidence of deliberate, real-time attention diversion by elected officials, revealing a novel challenge to democratic accountability in the digital age.

0 citationsRead paper

Training for Obsolescence? The AI-Driven Education Trap

Aug 27, 2025

This study identifies a critical skill-mismatch risk arising from AI adoption in education: policymakers overemphasize AI’s short-term pedagogical efficiency gains while neglecting its long-term wage-suppressing effects on AI-complementary skills, leading to growing misalignment between curricular offerings and labor market demands—intensifying with higher AI penetration. Method: We innovatively incorporate the crowding-out effect of non-cognitive skills (e.g., perseverance) into an integrated analytical framework, combining theoretical modeling with empirical data from pilot educational interventions to characterize the dynamic interplay among AI penetration, skill returns, and non-cognitive development. Contribution/Results: Findings reveal that unguided AI deployment in education not only exacerbates skill mismatch unidirectionally but also systematically erodes foundational non-cognitive competencies, thereby inducing long-term human capital depreciation. The model provides a rigorous foundation for evidence-based educational policy under AI-driven labor market transformation.

0 citationsRead paper