A dissociation between detection and identification of phobic stimuli: Unconscious perception?

Abstract

A psychophysical paradigm for investigating unconscious perception was used to test the hypothesis of dissociation between detection and identification of phobic stimuli. Spider-phobic and non-phobic participants were presented with masked images of spiders and flowers and an equal number of control stimuli in a random sequence. After each masked stimulus was flashed, participants first reported whether or not an object was presented. Then they identified each stimulus as either a spider or a flower, regardless of their prior detection response. Phobic participants identified both detected and undetected spiders better than chance, as assessed by two measures of response bias. They did not exhibit dissociation between detection and identification for flowers. Non-phobic participants did not exhibit detection–identification dissociation for either spiders or flowers. These results are consistent with the interpretation that phobic individuals unconsciously perceive their feared stimulus, and constitute the first direct demonstration of such for emotional stimuli.

Keywords:

• Detection–identification dissociation
• Unconscious perception
• Phobia

View correction statement:

Corrigendum

Notes

¹ The SOA is primarily a function of the refresh rate of the monitor, or how quickly it regenerates stimuli on the screen. By default, SuperLab erases the target stimulus before the masking stimulus is presented on the next refresh cycle. This makes the SOA between target and mask approximately 25 ms (SuperLab, personal communication).

² The response bias measures also yielded qualitative differences associated with hits and misses. On detection misses, false alarm rates for spiders exceeded those for flowers, indicating that missed flowers were more likely to be interpreted as spiders than vice versa. Paired samples t-tests showed that this effect was exhibited by both phobics, t(24) = 2.14, p=.043, and by non-phobics, t(21) = 3.07, p=.006. On detection hits, however, these false alarm rates were comparable in both groups. That is, non-phobic/fear-irrelevant stimuli were misinterpreted as phobic/fear-relevant only when either group did not detect the former stimuli. Furthermore, a paired t-test showed that while non-phobics more likely to interpret “undetected” control stimuli as spiders rather than flowers, they were more likely to interpret “detected” control trials as flowers rather than as spiders, t(21) = 2.79, p=.011, in both cases. This suggests that when the perceptual situation was utterly ambiguous, they may have used different interpretation strategies based on their awareness of the stimuli, as was reported by a few non-phobics during debriefing, e.g., “If I couldn't make it (the stimulus) out, I thought it was probably something interesting. If I could make it out, it was probably something ordinary.”