An AI agent can complete the Attention Network Test with human-like behavioral signatures: Implications for the bot-or-not debate

2026-03-27

Can AI agents produce behavioral data that passes as human? This question carries direct consequences for any field that relies on online reaction time (RT) experiments. Recent proposals for bot detection emphasize distributional shape, mean-variance scaling, and trial-wise autocorrelation of RTs, though the sufficiency of these markers has been challenged. We report the iterative development and empirical evaluation of an autonomous AI agent that completes the Attention Network Test (ANT) on a live Pavlovia experiment, producing behavioral data in real time. Across seven code revisions, each informed by analysis of the agent's output, the bot achieved attention network scores within published human norms (alerting = 65.1ms, orienting = 52.1ms, executive = 72.6ms), 95.8% accuracy, and an RT distribution exhibiting positive skew and trial-to-trial autocorrelation. We evaluated the agent against 796 human participants who completed the same ANT implementation across three university sites. The bot fell within the human range on QQ normality (z = -0.09), skewness (z = -0.77), and all three network scores, but showed elevated autocorrelation and a bimodal RT distribution from intermittent detection failures. Building this agent was technically feasible but required substantial iterative effort, experiment-specific reverse engineering, and repeated access to behavioral output. These constraints make widespread deployment unlikely for complex RT tasks in the near term, though the barrier will lower as agentic AI tools mature.

A collective review on some potential negative impacts of smartphone and social media use on adolescent mental health: Results from a Delphi process

2025-05-15 · 6 citations

The literature on how smartphone and social media use affects adolescent mental health is highly fragmented. To synthesize the evidence, we convened over 120 researchers with diverse perspectives to evaluate 26 commonly cited claims using a Delphi process. A large majority agree that: Adolescent mental health has declined in several Western countries; heavy smartphone and social media use can cause sleep problems; such use correlates with attention problems and behavioural addiction; among girls, social media use may be associated with body dissatisfaction, perfectionism, exposure to mental disorders, harassment and predation. Most other claims were judged to have insufficient evidence due to limited, inconsistent, or non-causal data. Researchers also raised broader concerns, including challenges in measuring mental health and establishing causality, geographic bias in existing evidence, and the need for policies that account for diverse risks and avoid unintended harms. This collective review offers a foundation for future research and policy.

A collective review on some potential negative impacts of smartphone and social media use on adolescent mental health: Results from a Delphi process

2025-12-12 · 1 citations

The literature on how smartphone and social media use affects adolescent mental health is highly fragmented. To synthesize the evidence, we convened over 120 researchers with diverse perspectives to evaluate 26 commonly cited claims using a Delphi process. A large majority agree that: Adolescent mental health has declined in several Western countries; heavy smartphone and social media use can cause sleep problems; such use correlates with attention problems and behavioural addiction; among girls, social media use may be associated with body dissatisfaction, perfectionism, exposure to mental disorders, harassment and predation. Most other claims were judged to have insufficient evidence due to limited, inconsistent, or non-causal data. Researchers also raised broader concerns, including challenges in measuring mental health and establishing causality, geographic bias in existing evidence, and the need for policies that account for diverse risks and avoid unintended harms. This collective review offers a foundation for future research and policy.

A Consensus Statement on Potential Negative Impacts of Smartphone and Social Media Use on Adolescent Mental Health

SSRN Electronic Journal · 2025-01-01 · 4 citations

Is It Time to Abandon the Media Multitasking Index?

Journal of Media Psychology Theories Methods and Applications · 2025-05-20

Abstract: For over a decade, the media multitasking index (MMI) has been the de facto standard measure for media multitasking. Thousands of studies in communication and adjacent disciplines have used the measure to investigate the relationship between media multitasking and cognitive functioning. In this paper we pose one central question: Is it time to abandon the MMI as a valid measure of individual differences in media multitasking? To answer this question, we highlight a selection of practical, methodological, and theoretical concerns regarding the validity and usefulness of the MMI for indexing media multitasking. Thereafter, we outline a research agenda to evaluate the veracity of these concerns and to advance the development of new measures of media multitasking and the adoption of new research designs. In doing so, we aim to integrate disjointed perspectives and stimulate new, improved theoretical and methodological practices in media multitasking research.

Limited capacity model

Edward Elgar Publishing eBooks · 2025-12-16

What Can Cognitive Load and Processing Fluency Tell Us About Difficult Processing?

Media Psychology · 2025-10-30

This experiment investigated the impacts of difficult processing in the context of a health narrative by integrating two distinct literatures: cognitive load and processing fluency. Difficult processing was manipulated via an auditory cue induction of reactive cognitive control (a manipulation of cognitive load) and language difficulty (a manipulation of processing fluency). We examined whether cognitive load can impact persuasion-relevant outcomes, and similarly whether processing fluency can impact memory. Results revealed that the processing fluency manipulation impacted memory, suggesting subjective difficulty was indicative of resource availability, as well as narrative persuasion. Additionally, the processing fluency and cognitive load manipulations interacted such that when the language was difficult, the presence of auditory cues enhanced memory. Our results show that cognitive load and processing fluency interact to shape message processing and effects.

Computational Methods in Media Psychology

Oxford University Press eBooks · 2025-07-22

Abstract As media environments become increasingly complex and data-rich, computational approaches have become more mainstream for studying media selection, processing, and effects. These tools afford media psychology researchers new forms of data, new ways to pre-process data to clarify and add structure to unstructured information, and richer ways to build and test theories. This chapter provides an overview of computational methods and their applications in media psychology research. First, the author highlights a number of tools for collecting data, including web scraping, digital trace data analysis, and custom experimental platforms. Next, he describes how pre-processing tools like natural language processing, computer vision, and network analysis can be used to clean, organize, and structure large datasets. Finally, he discusses the growing role of computational modeling in building and testing psychological and social theories, with a focus on cognitive and agent-based models. By providing means of integrating data-driven and theory-driven approaches, computational methods can help advance a more comprehensive understanding of the complex dynamics at play within individuals, messages, and social interactions.

Chapter 44 The Biology of Entertainment

2024-10-07

Corporate Purpose: Theoretical and Empirical Foundations/Confusions

2024-10-03

Abstract Does it matter if corporate leaders pursue a broader, social corporate purpose rather than a narrow, shareholder-centric one, and can legal and governance levers influence their choice? Theoretically—and limited by substitution, regulation, and legitimacy—socially-minded corporate decision-making can benefit society ex post, while commitment to either purpose may be required to motivate various constituencies’ contributions ex ante. Empirically, however, even structural measures like employee co-determination have barely detectable effects, let alone mere exhortations such as those in (unenforceable) nuances of (misunderstood) fiduciary duties. Many arguments for or against (particular) corporate purpose(s) are fallacies, red herrings, or, for empirics, cherry-picking.