Unifying scene-object congruency and incongruency benefits in object perception.

Area of Science:

• Cognitive Psychology
• Visual Perception
• Computational Semantics

Background:

• The impact of scene-object semantic congruency on object perception is known, but its directional influence is debated.
• Existing research lacks a nuanced understanding of how congruency affects perception across different tasks and mechanisms.

Purpose of the Study:

• To investigate the relationship between scene-object congruency and object perception using a novel vector-space semantic model.
• To explore the dual mechanisms of attention and processing efficiency underlying this relationship.
• To reconcile conflicting findings in previous studies on congruency effects.

Main Methods:

• Quantified scene-object congruency as a continuous variable using a vector-space semantic model.
• Conducted four experiments (N=543) using exemplar identification, change detection, and identification tasks.
• Manipulated congruency levels, stimulus sets, scene presentation times, and attentional cues.

Main Results:

• A positive U-shaped relationship was observed between congruency and identification accuracy.
• An inverted U-shaped relationship between congruency and reaction time indicated enhanced attention in both congruent and incongruent contexts.
• Congruency uniquely improved processing efficiency, while incongruency did not, suggesting distinct effects.

Conclusions:

• Scene-object congruency influences object perception via both attentional and processing efficiency mechanisms.
• Both congruent and incongruent contexts can enhance attention, but only congruency directly improves processing efficiency.
• This study reconciles contradictory findings by demonstrating the coexistence of congruency and incongruency benefits, offering a unified framework.