The Influence of Emotion on Object-Based Attention Using Reward

Poster Presentation 56.463: Tuesday, May 19, 2026, 2:45 – 6:45 pm, Pavilion
Session: Attention: Reward

Sanweda Mahagabin1, Sarah Shomstein1; 1The George Washington University

Object-based attention (OBA) enhances processing of features within attended objects beyond their spatial position. While reward history is known to bias attention, it remains unclear how the emotional valence of feedback– reward, punishment, or neutral– modulates space- and object-based selection. Using performance-contingent feedback to link each object with a positive or negative emotional value, we examined how these associations modulate OBA in a modified Egly-rectangle paradigm. Participants completed a valenced session where one object was consistently rewarded and the other punished, and a separate group completed a structurally identical no-reward session with neutral feedback. Each trial contained a cue at one end of a rectangle, followed by a target at the cued location (valid), the opposite end of the same object (same-object, SO), or the equidistant location on the uncued object (different-object, DO). In the valenced session, objects carried fixed emotional value: targets on the rewarded object always yielded gains, whereas targets on the punished object always yielded losses. The no-reward session provided non-contingent, neutral feedback. Accuracy was uniformly high (≥97%) with no group differences, ruling out speed-accuracy tradeoffs. Space-based attention was robust: responses were faster for valid than invalid trials in all conditions, with no modulation of this effect by valence. Within the valenced session, reward produced faster responses than punishment for both trial types (ps<.01). OBA was also reliable: invalid SO trials were faster than DO trials across groups (SO-DO: 22.62 ms reward, 37.35 ms punishment, 33.15 ms no-reward). Importantly, a significant Valence × Trial interaction emerged (F(2,76)=3.73, p=.029), indicating slower DO responses under punishment than under reward. In a set of follow-up experiments, we tested whether similar emotional influences arise when affective content, rather than monetary reward, is embedded in background scenes or used as targets themselves.