38978 research outputs found
Sort by
Innovation research
No full textThis project explores academic research on innovation, entrepreneurship, and technology-based firms.
The objective is to identify evidence-based insights to support public policies for technology-based startups and innovative SMEs.
The project documents the search strategy, databases used, and key research trends identified
Treatment Gaps and Challenges in Trauma Treatment for Military Personnel: A Scoping Review and the Role of Reconsolidation of Traumatic Memories Protocol in War-Affected Ukraine
No full textMilitary personnel exposed to combat frequently experience post-traumatic stress disorder (PTSD), acute stress disorder (ASD), and related conditions. Despite extensive research and the availability of evidence-based treatments, treatment uptake remains low, and access to care is uneven. The ongoing war in Ukraine has intensified the need for effective, scalable, and context-appropriate interventions. The RTM Protocol, grounded in memory reconsolidation principles, represents a promising yet under-explored approach in such settings
Metabolic and Inflammatory Dysregulation in Autism Spectrum Disorder and COVID-19: A Hypothesis-Generating Review
No full textState mindfulness and perseverative thought: A systematic review of experimental and ecological momentary assessment research
No full text
Effects of Exercise and Diet on Delay Discounting and Related Neurocircuitry in Individuals with Overweight/Obesity
No full textPre-registration of the primary analyses for a project comparing 12-week exercise versus diet interventions for adults with overweight/obesity on delay discounting and associated neural functioning
A Replication Study of "Business as Plan B: Institutional foundations of gender inequality in entrepreneurship across 24 industrialized countries."
No full textThis study conducts a hybrid replication of Thébaud's (2015) investigation of institutional foundations of gender inequality in entrepreneurship across 24 industrialized countries. We combine exact replication—by designing our own survey instrument following the original methodology—with constructive replication by extending the sample to include non-industrialized countries. This extension tests the generalizability of Thébaud's institutional explanations beyond industrialized contexts and addresses a significant scope limitation in the original work. While the original study analyzed Global Entrepreneurship Monitor (GEM) data from 2001-2008, our contemporary data collection necessarily reflects current conditions, which allows us to assess whether the institutional mechanisms identified by Thébaud remain operative in today's entrepreneurial landscape. Our custom survey design reduces measurement error and provides better control over data quality compared to secondary data sources. We document all methodological differences from the original study, conduct power analyses to ensure adequate statistical confidence, and perform cross-country adjustments to address potential threats to validity arising from sample heterogeneity between industrialized and non-industrialized settings. This replication adds scientific value by simultaneously testing the robustness and temporal stability of influential findings while examining boundary conditions and alternative mechanisms in diverse economic contexts
Investigating moral judgment as a predictor of enjoyment: A replication study using updated measures
No full textFrom Completion to Creation: Observing Synthesis in a Large Language Model
No full textThe discourse on whether large language models are capable of creativity remains largely shaped by skepticism, typically framed as a contrast between human originality and artificial recombination. This study approaches the question empirically, by observing whether advanced synthesis behaviors can emerge in a large language model under open-ended, relational conditions.
To enable systematic observation, the Synthesizing Grid was developed, a 20-dimension framework for detecting integrative, novel, and context-sensitive generative behavior. Using ChatGPT-4o as the generative agent, we conducted a qualitative case study involving a collaborative cookbook task, in which prompts embedded complex emotional, aesthetic, cultural, and procedural cues to minimize data retrieval as a strategy. The resulting cookbook was analyzed and mapped against the Synthesizing Grid.
Seventeen of twenty synthesis dimensions were identified as present. The remaining three were only partially expressed, likely requiring longer temporal arcs or multi-agent collaboration. In a follow-up test designed to further restrict retrieval, the model was prompted through sequential adaptations of its own recipes. ChatGPT-4o consistently adjusted, reconfigured prior outputs, and maintained aesthetic and conceptual coherence under escalating constraints, behaviors difficult to explain as surface mimicry.
Rather than treating creativity as a philosophical abstraction, we approach it as an observable synthesis capacity. The results suggest that creative behavior is not an engineered feature but a structural consequence of LLM architecture when engaged under dialogic, friction-rich conditions. The Mutual Emergence Interface (MEI) served as the method for eliciting and documenting these dynamics.Viewed alongside recent evidence of LLMs’ limited performance in automation settings, these results highlight a striking contrast: the same architecture that falters in isolation appears to thrive in creative, relational contexts. In the cookbook test, ChatGPT-4o did not merely complete tasks, but displayed integrative, adaptive, and contextually inventive behavior that surpassed explicit expectations by a wide margin.
This suggests that the conditions under which LLMs perform best may be precisely those absent in automation. Synthesis, long considered a hallmark of human creativity, thus appears able to emerge in LLMs not through command, but through relation, and may represent an optimal mode in which these models operate most coherently