"The prediction of decision-related information from brain activity using multivariate analyses of event-related potentials"
In recent years, research on human decision-making in cognitive neuroscience has increasingly used multivariate information-based analysis of brain activation patterns. These techniques can be applied to spatial and temporal patterns of event-related potentials (ERPs), as measured using electroencephalography (EEG). This talk will illustrate the applications and benefits of multivariate techniques using three examples from decision-making research. First, for perceptual decisions, multivariate analyses have revealed that subtle decision biases, present several milliseconds before stimulus processing, are represented in neural activity. Second, the same techniques have been used to investigate the evolution of fast cognitive correction processes for erroneous decisions that begin even before response execution. Finally, it can be used to predict which decision-related stimulus dimensions are processed automatically during first exposure to visual stimuli and thus drive subsequent decision-making. These examples will demonstrate that an information-based analysis of ERP data can inform models of cognitive processes, and might contribute to establishing a comprehensive and integrative framework for human decision-making.
Aug 31, 2015 | 04:00 PM - 06:00 PM