Cue discriminability predicts instrumental conditioning.

Charting mental acts that succeed or fail under unconscious instances of cognition informs debates on the nature and potential functions of consciousness. A prominent method to exclude conscious contributions to cognition is to render visual stimuli unconscious by short and pattern-masked presentations. Here, we explore a combination of visual masking and pixel noise added to visual stimuli as a method to adapt discriminability in a fine-grained fashion to subject- and stimulus-specific estimates of perceptual thresholds. Estimates of the amount of pixel noise corresponding to perceptual thresholds are achieved by psychometric adaptive algorithms in an identification task. Afterwards, the feasibility of instrumental conditioning is tested at four levels of cue discriminability relative to previously acquired estimates of perceptual thresholds. In contrast to previous reports (Pessiglione et al., 2008), no evidence for the feasibility of instrumental condition was gathered when contributions of conscious cognition were excluded.