Changes in implicit associations do not occur simultaneously to Pavlovian conditioning of physiological anxiety responses.

Previous research shows that an 'angry face' conditioned stimulus will elicit conditioned skin conductance responses (SCR) after a small number of trials pairing it with a mild electric shock. Such conditioning occurs even with masked presentations of the facial stimulus. Furthermore, the Implicit Association Test (IAT) has been demonstrated to be sensitive to the presence of information processing biases seen in specific phobias. We examined the ability of the IAT to detect changes in implicit associations that arise from pairing one facial image (CS+) with a mild electro-tactile stimulation whereas a second facial image (CS-) was presented alone. Participants (N=117) were assigned to one of four groups: a group that received no shocks with unmasked presentation of the facial images; a group that received no shocks with masked presentation; a group with shocks delivered simultaneously to the masked presentation of the CS+; and a group that received shocks simultaneous to unmasked CS+ presentation. Learning was assessed through both SCR and the IAT. Results showed that while it was possible to condition a SCR to the CS+, this learning was not detectable using the IAT. Implications of the discrepancies in the outcome measures are discussed.