Abstract

A frequently used paradigm to investigate unconscious processing is the masked priming paradigm. Traditionally, a direct task is compared to an indirect task. In the direct task, detection of a masked prime stimulus is measured, typically being close to chance. In the indirect task, effects of the prime on reaction times to a subsequent target stimulus are analyzed (priming effects). Here, we focus on the direct task, where participants usually provide a binary response. In a previous experiment we have shown that confidence covaried with the accuracy of prime detection in number discrimination tasks, suggesting that participants have access to a richer, continuous representation of the prime in the direct task and therefore should not be restricted to a binary decision. To assess whether this remains true for different sets of stimuli, N=11 participants performed the masked priming paradigm with simple line stimuli that were horizontally or vertically oriented. Participants classified the orientation of the prime (direct task) or of the later target stimulus (indirect task). Additionally, in the direct task, participants judged their confidence in their response on a continuous scale. We found that confidence covaried with the accuracy of prime detection (F(1,10)=14.0, p=.004, η²=.58), with higher accuracies of prime detection for trials with higher confidences. We discuss to which extent continuous direct measures open up new analysis strategies and why they might be favorable for the investigation of unconscious priming.

[This project is supported by the Deutsche Forschungsgemeinschaft (DFG, German Research Foundation) through the CRC1233 “Robust Vision”, project number 276693517]