Humans can learn to use repeated spatial arrangements of irrelevant, non-target items to direct the focus of attention towards behaviorally relevant – target – items, a phenomenon known as contextual cueing (CC). However, whether CC is itself dependent on attentional resources is a controversial issue. Here, we used visual search to test how CC is affected when attention varies through two types of manipulations: perceptual load (as induced by target-distractor similarity) and postural load (sitting vs. standing). For easy searches (low target-distractor similarity), we observed reliable facilitation of search in repeated-context displays, which was independent of participants’ body posture. For difficult searches (high target-distractor similarity), contextual facilitation was evident only with standing posture. Posture-related benefits remained significant even after controlling for heart rate variability (HRV), body mass index, and physical activity. Decomposing aggregated reaction times by drift-diffusion modeling revealed that CC in difficult searches decreased the amount of evidence required for target-response decisions. Our results suggest that statistical learning is effectively supplemented during standing posture when visual search is challenging, possibly because posture manipulation and contextual manipulation affect common response-selection stages of processing.
