FIGURE 1.

Hypothetical experiment in which the conditioned stimulus (CS; cue) is followed by the unconditioned stimulus (US; outcome) on 0, 50, or 100% of the trials. The 0% cue reliably predicts the absence of the US, the 100% cue reliable predicts the occurrence of the US, the 50% induces uncertainty, and consistently causes error in the prediction of the US. From left to right, the panels depict different perspectives on how associative learning ascribes value to the CS based on these contingency. This learned value in turn might create an attention bias that influences automatic attentional capture by the cues. The Left panel depicts the idea that attention linearly increases with the strength of the association between the CS and the US. In the Mid panel attention is selectively allocated to cues that are good predictors of the outcome. The Right panel depicts the idea that learners allocate attention to uncertain cues. Figure adopted from Gottlieb (2012).