Unequally masked: Indexing differences in the perceptual salience of "unseen" facial expressions

Two experiments investigated the degree to which different facial expressions, when backwardly masked by neutral faces, are blocked from different levels of perceptual access as indexed by explicit (self-reported) awareness, forced-identification performance, and stimulus sensitivity in a signal detection paradigm. Results indicate that a 16.67 ms target-mask stimulus onset asynchrony (SOA) is insufficient to reliably block either the simple detection or the forced-identification of backwardly masked expressions. This finding holds even for participants who deny possessing any explicit awareness of the masked images whatsoever. Furthermore, "unseen" happy faces are less effectively masked by neutral expressions than are unseen angry or neutral faces, an observation that might also extend to other affective expressions (e.g., fear). These results expand upon previous findings (Esteves & Öhman, 1993) and provide new information about the multitude of factors that may be operating in research that involves backwardly masked facial expressions.