Humans strategically shift decision bias by flexibly adjusting sensory evidence accumulation

Abstract

Decision bias is traditionally conceptualized as an internal reference against which sensory evidence is compared. Instead, we show that individuals implement decision bias by shifting the rate of sensory evidence accumulation towards a decision bound. Participants performed a target detection task while we recorded EEG. We experimentally manipulated participants’ decision criterion for reporting targets using different stimulus-response reward contingencies, inducing either a liberal or a conservative bias. Drift diffusion modeling revealed that a liberal strategy biased sensory evidence accumulation towards target-present choices. Moreover, a liberal bias resulted in stronger midfrontal pre-stimulus 2-6 Hz (theta) power and suppression of pre-stimulus 8—12 Hz (alpha) power in posterior cortex. The alpha suppression in turn mediated the output activity of visual cortex, as expressed in 59—100 Hz (gamma) power. These findings show that observers can intentionally control cortical excitability to strategically bias evidence accumulation towards the decision bound that maximizes their reward.

Boosting Brain Signal Variability Underlies Liberal Shifts in Decision Bias

“…Here, we provide first evidence that flexible adjustment of neural variability in frontal regions (but not signal SD or oscillatory brain dynamics) may underlie such strategic shifts in decision bias. The relationship between shifts in bias and neural variability in frontal regions reported here complements our previous findings in the frequency domain that humans can intentionally control prestimulus 8-12 Hz (alpha) oscillatory power in posterior cortex to strategically bias decision making (10). Notably, we previously observed increased oscillatory 2-6 Hz (theta) power in the liberal condition in the same midfrontal electrodes implicated here in the bias-entropy correlation, but this theta power difference was not correlated with the bias shift.…”

“…We report a novel analysis of a previously published dataset involving a target detection task during two different decision bias manipulations (10).…”

“…Due to this dependence of the r parameter on signal SD, higher SD can yield lower indicated entropy and vice versa, possibly resulting in negative correlations between the two metrics. In addition, electrophysiological (e.g., EEG) data is often quantified in terms of oscillatory spectral power in canonical delta (1-2 Hz), theta (3-7 Hz), alpha (8)(9)(10)(11)(12), beta (13-30 Hz) and gamma (60-100 Hz) bands, which might be related to entropy. Therefore, we tested whether the ∆ bias-∆entropy correlation could be explained by broadband signal SD and oscillatory spectral power.…”

“…We used a single Hann taper for the frequency range of 3-35 Hz (spectral smoothing, 4.5 Hz, bin size, 1 Hz) and the multitaper technique for the 36 -100 Hz frequency range (spectral smoothing, 8 Hz; bin size, 2 Hz; five tapers) (47). See (10) for similar settings. Finally, to investigate spectral power between 1-3 Hz (delta band), we performed an additional time-frequency analysis with a window size of 1 s (i.e.…”

“…We addressed this hypothesis using data from humans performing a challenging, continuous target detection task under two different decision bias manipulations, while non-invasively recording their electroencephalogram (EEG) (10). Sixteen participants (three experimental sessions each, mean of 1733 trials/participant) were asked to detect orientation-defined squares within a continuous stream of line textures of various orientations and report targets via a button press ( Figure 1A).…”