Purchase this article with an account.
Jennifer McBride, Ute Leonards, Iain D Gilchrist; Flexible target representations underlie repetition priming in visual search. Journal of Vision 2007;7(9):688. doi: 10.1167/7.9.688.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
In visual search, repeating target features across consecutive visual search trials can influence response time (RT). It remains unclear, however, what representation of the target is stored across trials to produce this repetition priming. Two experiments were conducted to address this question and to investigate how priming is affected by feature type and the roles that features plays in defining the target. In Experiment 1, the target was defined as a bar differing from distractor bars in its orientation. Target colour was shared with some of the distractors and so was irrelevant for target definition and the search. Repeating the task irrelevant target colour across consecutive trials produced significantly faster RTs, but repeating the target-defining orientation significantly slowed RTs. In Experiment 2, the roles of colour and orientation were reversed: the target now differed from the distractors in colour, while target orientation was shared with some of the distractors and was irrelevant to the search. For Experiment 2, repetition of (target-defining) target colour produced significantly faster RTs, but no reliable cost or benefit was found for repeating task-irrelevant target orientation. Taken together, the results from Experiments 1 and 2 show that different priming effects are shown for different feature types (colour and orientation) and that swapping the roles played by these different target feature types also induces different priming effects.
The effects of repeating multiple target features interacted in a manner that is not consistent with views that priming operates either on feature-based or episodic (whole-object) representations of search targets. Instead, the representation of the target that is stored across trials and leads to priming seems flexible: it critically depends on the role of target features in the task (e.g. orientation once induced negative priming and once no reliable priming), and on the target feature types themselves.
This PDF is available to Subscribers Only