Fig. 1.

Experimental design and trial-by-trial shock expectancy ratings. (A) The Black-Shocks-First (BSF) and White-Shocks-First (WSF) groups both saw the same Black and White male exemplar. For the BSF group, the Black male (CS1) was originally paired with an electric shock unconditioned stimulus (US; depicted as the lightning bolt), and the White male (CS2) was unpaired. For the WSF group, the White male (CS1) was originally paired with shock, and the Black male (CS2) was unpaired. In both groups, the CS–US association reversed midway. (B) On each trial, subjects rated expectancy using a three-point scale corresponding to ‘no shock’ (=1), ‘maybe shock’ (=2), and ‘shock’ (=3). These declarative expectancy ratings tracked the CS–US contingencies accurately in both groups. Dashed line indicates the start of reversal.