Abstract
Attention could be guided to a target object in a search display when either the fixed spatial layouts or the same object sets were presented repeatedly (contextual cueing; Chun & Jiang, 1998; Chun & Jiang, 1999). Previous studies have shown that objects’ locations were selectively learned whereas their identities were not, even when both object locations and identities were available for predicting the target location (Endo & Takeda, 2004). However, our previous study found that only objects’ identities were selectively learned when objects were sequentially presented at multiple fixed locations (spatio-temporal contextual cueing; Higuchi et al., VSS2012). We investigated whether implicit selective learning of location and identity depends on tasks by devising two different tasks using the identical experimental paradigm: the spatio-temporal contextual cuing. Participants were asked to respond to a target among the stream of distractors presented sequentially at different locations. The target was defined as an object belonging different category (Experiment 1: identity task) and an object with spatial offset (Experiment 2: location task). In the learning phase, invariant sequences both in locations and identities associated with a particular target were presented repeatedly. In the following test phase, either locations or identities of the invariant sequences were randomized. In Experiment 1, reaction times increased in the test phase compared with the last block of the learning phase when the object identities were randomized, whereas reaction times did not change when the locations were randomized. In contrast, in Experiment 2, reaction times increased in the test phase when the object locations were randomized. The participants could not recognize the invariant sequences in both experiments. These results indicate that selective learning of location and identity are task-dependent, suggesting that task sets play an important role in deciding which information was learned implicitly.
Meeting abstract presented at VSS 2013