A neural marker of confidence predicts information seeking – new paper in Journal of Neuroscience

Awesome news: our paper on the neural markers of confidence and information seeking, with Peter Murphy, Annika Boldt, Tom Verguts and Nick Yeung, has now been in accepted in The Journal of Neuroscience. You can find the paper here! Data collection for this study was done back in 2015 (ages ago, it feels) when I spent three months in Nick Yeung’s lab at Oxford University. Actually, I collected these data in the very last week of my stay there – so the ratio data analysis/writing versus data collection for this project must be about 205:1.

In this work, we followed up on our behavioral finding that decision confidence predicts information seeking, even across conditions matched for difficulty. Here, we wanted to see whether known neural markers of confidence are predictive of such information seeking. Given previous work on this by Annika and Nick, we expected this to be the case for the Pe. We used a paradigm in which participants had to make a choice about the average color of eight elements. Before making their final choice, however, participants could decide whether or not they wanted to pay a small cost to see the stimulus again in an easier version.

Using multivariate decoding, we then trained a decoder at each point in time to predict based on EEG data whether a trial was judged with high or low confidence. Critically, it was then tested whether this decoder was able to classify whether participants wanted to sample more information or not. This was indeed possible in a time window following the initial speeded response. This time window corresponded to that of the Pe, a previously established neural marker of confidence (Boldt & Yeung, 2015).

In sum, our study shows that neural indices of confidence are functionally involved in information-seeking decisions.