Mindsets: Intuition vs. Deliberation
The issue of whether intuitive thinking or unconscious thought can potentially enhance decision-making is a topic of intensive debate. We are carrying out experimental research to test the impact of intuitive mind-sets on the quality ecological decisions. For example, we have found that instructions to rely on intuition enhance decision consistency (transitivity) in decisions that involve holistic stimuli. At present, we are developing a computational theory that is based on neural constraints, and which provides a neural mechanism (based on population codes) for fast numerical averaging. This theory accounts for a large set of experimental data, indicating the surprising fact that intuitive numerical averaging improves with the number of values to average, even at very fast presentation rates (100 ms per item).
Illustration of the model:
PopulationVector

Each presented number activates the network with a noisy activation profile (middle panel: blue, red and green lines). When combining these activations together (right panel: blue line) a combined profile is activated in the network. Using simulations, we have shown that the center of mass of the combined network represents the sequence mean. We are currently working on a biologically plausible neural network that could extract the activation center of mass.

NeuronProfileActivationMovie
According to the model, evaluating more numbers results in higher activation (better signal-to-noise-ratio) yielding more precision.