Leonidas Doumas

Leonidas Doumas is a Senior Lecturer in Psychology at the University of Edinburgh’s School of Philosophy, Psychology and Language Sciences. He holds a PhD in Cognitive and Developmental Psychology from UCLA and has held academic positions at Indiana University and the University of Hawaii before joining Edinburgh in 2013. His research focuses on relational reasoning, analogy, cognitive development, and computational modelling, particularly exploring how distributed systems like neural networks represent and generalize relational concepts. He teaches courses in developmental science, cognitive science, and statistics, and actively supervises PhD students in these areas.

Dr. Doumas’s work combines empirical studies with computational models to investigate topics such as cross-domain generalization, relational learning in children and adults, and the role of structured representations in cognition. His research also employs neuroscientific techniques like EEG and TMS to understand cognitive processes. He is a core member of the Edinburgh Cognitive and Neuroethology Lab (CNE) and leads initiatives in immersive virtual reality (VR) for neuropsychological assessment, such as the Virtual Reality Everyday Assessment Lab (VR-EAL).

His educational background includes a PhD under John Hummel, Keith Holyoak, and Cathy Sandhofer at UCLA, followed by postdoctoral training in Linda Smith’s lab at Indiana University. His current projects emphasize the intersection of symbolic and connectionist approaches to cognition, with a focus on how structured representations emerge through experience.

• Research Interests: Relational learning, analogy, cognitive development, computational modelling, neural networks, and VR applications in cognitive science
• Teaching: Year 1/2 Psychology courses, Cognitive Science, and advanced options on neural networks and induction
• Advising: Open to PhD/MSc students interested in relational cognition, computational models, or VR applications

Publications span topics from neural network limitations in language processing to mechanisms of hierarchical linguistic structure, with a focus on bridging computational and empirical cognitive science.