Sequential gains and losses during gambling feedback: Differential effects in time-frequency delta and theta measures

One critical aspect of reward-feedback is the impact of local outcome history—how past experiences with choices and outcomes influences current behavior and neural activity. Yet, prior event-related potential work in this area has been contentious. This study contributes to this field by using time-frequency measures to better isolate constituent processes. Specifically, we identify how theta and delta are differentially sensitive to local outcome history. Participants completed a binary monetary choice task while we collected EEG data. Unbeknownst to them, trial outcomes were manipulated into pre-determined sequences, ranging from one to eight gains or losses in a row. Analyses were arranged by sequence establishment (first 2 trials of a sequence) and continuation (prolonged sequences of 3–8 trials). During the establishment of a sequence, delta activity to gains and losses were virtually identical on the first (change) trial, demonstrating marked divergence only on the second trial. This difference grew throughout the continuation period, as delta activity was sustained with accruing gains but declined with multiple losses. Theta activity, conversely, demonstrated a maximal loss-gain difference on the change trial but was insensitive to the establishment of a new sequence. Differential theta activity between outcomes decreased as sequences continued, with theta activity increasing over accruing gains and remaining stable over losses. Results indicate that delta-gain and theta-loss signals are relatively stable across sequential outcomes. Furthermore, theta is most sensitive to loss-gain differences on the initial change trial, while delta is more sensitive to gain-loss differences with the continuation of a sequence.