ELSC Seminar Series

Over 10 years ago we developed a manipulation named the cue-approach task to influence choices of snack food items relying on non-externally reinforced mechanisms (Schonberg et al., Nature Neuroscience, 2014). In the task, a neutral tone and a button press are repeatedly associated with specific pictures of items in a session lasting less than 1 hour. In a subsequent probe phase participants choose between pairs of items where only one of the items in each pair was previously associated with the cue and button press. Since then replicated results in tens of samples show that this training leads to preference changes favoring the previously cued item and that these changes last up to several months. Functional MRI results show a value change signature in the ventro-medial prefrontal cortex during the choice phase and changes in several regions after training, including high level visual regions and connectivity changes during training with the striatum linked to individual changes. In this talk, I will discuss a new computational modeling meta-analysis (Salomon et al., 2025) of studies with the task alluding to the potential mechanism of the task involving internal reinforcements. Furthermore, I will discuss our suggested novel non-reinforced pathway in the brain (Schonberg and Katz, TiCS, 2020) and how it reconciles previously well-known effects such as the Mere-exposure effect (Zajonc, 1968). Finally, I will show why I suggest that virtual reality is potentially a great research tool to study human value-based decision making and its underlying physiological markers.