Object Spatial Certainty and Attention: Quantifying Variability in Object Locations in Scenes

Some objects have specific places where you can expect them to be found (e.g., toothbrush), while others vary widely (e.g., cat). Previous studies have pointed to the importance of the spatial associations between objects and scenes in informing search strategies. However, the assumptions about objects having a specific location that they are typically found does not take into account the imprecision inherent in the spatial associations of objects. In the current study, we investigated the effects of the variability of objects’ spatial associations on visual search. First, we quantified this variation and developed the Object Spatial Certainty index by having participants rate where 150 objects were expected to be found in scenes; the index provides a relative measure that ranks these objects from the most spatially predictable (almost always found in one region of the scene, e.g., boots) to the least spatially predictable (equally likely to be in every region of the scene, e.g., plant). In two experiments, we examined how these variations affected search by manipulating targets are either being High Certainty or Low Certainty. Our findings demonstrate that the variability of spatial association of objects significantly affected how effectively scene context influences search performance.