14234 research outputs found
Sort by
PREDICTING FOOTBALL MATCH OUTCOMES USING LARGE LANGUAGE MODELS: A COMPARATIVE STUDY WITH TRADITIONAL MACHINE LEARNING METHODS
No full textAccurately predicting football match outcomes is valuable for stakeholders such as fans, analysts, sports betting companies, and team strategists. In this study, we explore the potential of large language models (LLMs) for predicting football match results by transforming numerical features into contextual inputs. Key features include historical match results, player ratings, coach ratings, and other relevant conditions, which are processed by the LLM to predict the match winner. We compare the performance of LLM-based predictions with traditional machine learning (ML) models, including random forest and XGBoost. Our findings demonstrate that LLMs achieve comparable accuracy to these conventional ML techniques. Additionally, the LLM offers a significant advantage in that it requires no model training, simplifying implementation and reducing computational costs. This makes LLMs a promising, resource-efficient alternative for football match prediction, presenting new opportunities for AI-driven sports analytics
Strategic Interactions in Science Communication: A Complex Adaptive Systems Framework
No full textScience communication plays a crucial role in maintaining public trust in science amid complex societal challenges. This study addresses a gap in understanding science communication dynamics by conceptualizing it as a complex adaptive system of actors. It introduces the CASSCO (Complex Adaptive System of Science Communication) model, which integrates complex adaptive systems theory with game theory to analyze strategic interactions in science communication. The model encompasses decision-making processes, impact evaluation, and learning mechanisms among actors, distinguishing between institutional and organizational roles across three communica-tion modes: dissemination, dialogue, and participation. By applying the CASSCO model to two scenarios - citizen science and generative AI – the study demonstrates its potential for predicting non-linear dynamics and emergent outcomes in science communication. This approach yields insights into the impact of communication strategies on public trust and contestations of science. The CASSCO model serves as a strategic thinking template, enabling actors to select strategies while considering the behaviors of others. The study concludes with theoretical and practical implications, model limitations, and future research directions
Set-for-Variability Predicts Responsiveness to Tier 2 Reading Interventions
No full textThis is the OSF project page for the manuscript "Set-for-Variability Predicts Responsiveness to Tier 2 Reading Interventions"
Teaching in the Garden of Forking Paths: Implementing Many-Analysts Designs in the Classroom
No full textThe rise of many-analysts studies has underscored substantial variability in research outcomes when different teams independently analyze identical data sets and hypotheses. This paper demonstrates how the many-analysts framework can be effectively integrated into research methods education. We propose a pedagogical approach in which multiple students independently analyze the same research question using identical datasets in their term papers, followed by a meta-analysis of their collective results. This instructional method highlights critical methodological concepts, such as researcher degrees of freedom and epistemic humility. Moreover, this study applies the many-analysts approach specifically to causal inference using observational data with difference-in-differences analyses. The analyses conducted by students revealed a broad range of effect sizes and led to divergent conclusions, thus emphasizing the educational and methodological value of robustness checks involving multiple independent analysts. These findings illustrate how integrating a many-analysts framework in teaching can enhance students' methodological rigor and appreciation of empirical uncertainty
Examining the Relations Between Self-Construal, Culture, and Cognitive Dissonance Using the Induced Compliance Paradigm
No full textCognitive dissonance, a fundamental psychological process involving inconsistent cognitions causing discomfort, may vary across cultures. These variations could be attributed to differences in the way people define themselves, known as “self-construal”. Previous cross-cultural studies on the role of self-construal in cognitive dissonance have mainly employed the free-choice paradigm. However, many concerns have been raised about the validity of this procedure and these studies, more generally. To address this issue, we will conduct secondary analyses to explore unexamined associations in a large existing dataset (Vaidis et al., 2024). Specifically, the current study will investigate the moderating role of individual (self-construal) and cultural (individualism) variables on dissonance effects following an induced-compliance paradigm across 18 countries (N = 3822). Based on the literature, we hypothesised that induced-compliance effects (i.e., adjusting attitude to match behaviour, particularly when it has been adopted freely) will be stronger for participants with higher individual self-construal scores (H1), in more individualistic countries (H2), and in countries with higher aggregated levels of independent self-construal (H3). The analyses [supported / did not support] H1, [supported / did not support] H2 and [supported / did not support] H3.
Keywords: Self-Construal, Culture, Cognitive Dissonance, Induced Compliance, Cross-cultural stud
Disrupted Physiological Coregulation in Youth at Clinical High Risk for Psychosis: Insights from a Dyadic Interaction Study
No full textInterpersonal difficulties have long been implicated in psychopathology. However, we know quite little about how social (dis-) connection unfolds at the physiological level in real time in clinical populations, including among youth at clinical high risk for psychosis (CHR). The present laboratory-based dyadic interaction study examined physiological coregulation in 70 dyads across 137 participants (32 CHR youth-caregiver dyads; 38 healthy youth-caregiver dyads) and linked coregulation with clinical symptoms – concurrently and prospectively. This study was based in a mid-sized Midwestern city and recruited from the broader community and from mental health clinics. Data were collected from 2018 to 2022. Dyads engaged in 10-minute neutral, conflict, and pleasant conversations while their autonomic physiology was continuously monitored. In conflict and neutral conversations, CHR youth exhibited contrarian physiological coregulation (i.e., slowing heart rate in response to caregiver’s escalating heart rate and vice versa). Contrarian coregulation was associated with elevated risk for psychosis, was linked to greater baseline psychosis symptomatology, and prospectively predicted increases in symptoms one year later. These findings document alterations in physiological coregulation between CHR individuals and their caregivers, highlight their relevance for clinical symptomatology, suggest novel avenues for relationship-focused treatments, and contribute to a biologically-grounded science of social connection
Sexual Minority Persons’ Life Course Experiences with Arrest and Incarceration
No full textSecondary data analysis conducted by undergraduate students in The University of Alabama's Flexa Lab
