Research

ATLAS focuses on turning constructs into instruments that work in the real world. We develop and validate tools that do more than describe outcomes: they measure the mechanisms and processes that drive behaviour. Using behavioural experiments, digital assessments, computational modelling, and AI-enabled methods, we study learning, decision-making, motivation, apathy, and reinforcement learning in ways that are scalable, interpretable, and clinically or educationally useful.

Programmes

1. Metacognition & Adult Learning
   We develop technology-enhanced interventions to strengthen metacognition in adults and evaluate them using converging measures, including self-report, behavioural tasks, cognitive testing, and neuroimaging. Our goal is to understand not only whether people learn, but how they monitor, regulate, and improve their own learning.
2. Developmental Digital Screening (ASD)
   We are developing caregiver-friendly, home-based digital tools for earlier identification of autism risk in children. These tools are designed to be accessible, scalable, and sensitive to meaningful developmental differences, and we validate them across autistic and typically developing cohorts.
3. AI-driven Metacognition Scoring Engine
   We design and evaluate AI-based methods, including large language models, to extract structured psychological indicators from text and behaviour. This work extends our measurement framework by creating scalable scoring tools for constructs such as metacognition and decision processes while maintaining interpretability and validity.
4. Decision-Making, Motivation & Reinforcement Learning
   We use behavioural tasks and computational approaches to study the mechanisms underlying decision-making, apathy, motivation, and reinforcement learning. By developing and applying process-sensitive measures, we aim to identify how people evaluate effort, reward, uncertainty, and feedback, and how these mechanisms vary across individuals and clinical groups.