Perceptual organization facilitates object-based distractor filtering

: Statistical learning-based filtering mechanisms reduce attentional capture more for frequent than for infrequent distractors, a phenomenon known as the "distractor frequency effect". In the field of object-based attention, this effect has been studied with real-world object distractors, which engage both perceptual and semantic processing of distracting information. Whether similar mechanisms are also engaged by meaningless distractors remains unknown. Within a template-for-rejection framework of filtering mechanisms, we hypothesized that perceptual grouping of distracting visual input into coherent units is the minimal sufficient condition for these mechanisms to operate. In Experiment 1, Gabor patches were presented with superimposed distractors, consisting of either outlined meaningless distracting shapes or scattered-dot displays, which appeared within a block of trials at either high or low frequency. Shapes produced greater interference in Gabor orientation discrimination than scattered dots, despite the two being equivalent in terms of sensory energy impinging on the retina, indicating that additional processing was required to group the elements into shapes. Furthermore, less frequent shape distractors caused greater interference than more frequent ones. This distractor frequency effect was absent for scattered-dot displays but re-emerged in Experiment 2 where the scattered dots were grouped by proximity into multi-regional patterns, indicating that perceptual grouping was critical for this effect to occur. Experiment 3 demonstrated that the frequency effect stems from statistical learning rather than perceptual learning of the compound target-distractor pattern. We propose that Gestalt grouping may facilitate the formation of distractor templates, which are then used to filter task-irrelevant information, thereby reducing object-based distractor interference.