Abstract
Elevations in exclusion error rates (i.e., responding with the target stimulus despite instructions to the contrary) in experiments with masked, briefly presented stimuli have been attributed to unconscious perception. The present studies tested the validity of exclusion methods for studying unconscious perception. Experiment 1 replicated Merikle, Joordens, and Stolz (1995; Experiment 1) by showing more exclusion errors (exclusion failure) for masked word stimuli in a stem completion task. However, this experiment did not replicate the finding of fewer exclusion errors (exclusion success) at long stimulus durations. Experiments 2 and 3 showed that exclusion errors are accompanied by significant discrimination sensitivity to the target stimulus, which suggests conscious perception of the target stimulus. Experiments 3 and 4 demonstrated that the exclusion errors obtained from stem completion tasks depend in part on the exclusion method. Altogether, elevated exclusion failure does not provide an unambiguous demonstration of unconscious perception.
