Statistical Learning Facilitates Access to Awareness

Abstract

Statistical learning (SL) allows us to quickly extract regularities from sensory inputs. Although many studies have established that SL serves a wide range of cognitive functions, it remains unknown whether SL impacts conscious access. We addressed this question, seeking converging evidence from multiple paradigms across four experiments (total N = 153): Two reaction-time based b-CFS experiments showed that objects at probable locations and with probable features are released from suppression faster than improbable objects. In a visual masking experiment, weobserved higher sensitivity to probable (versus improbable) objects, independent of conscious access to the stimulus dimension carrying the regularities. Finally, a pre-registered accuracy-based b-CFS experiment showed higher localization accuracy for interocularly suppressed probable (versus improbable) objects given identical presentation durations, thereby excluding processing differences emerging after conscious access (e.g., criterion shifts). Together, these findings demonstrate that SL prioritizes conscious access of probable over improbable visual input.