Investigating the roles of features and priming in visual search
MetadataShow full item record
Identifying and locating specific objects amidst irrelevant, distracting items can be difficult when one is unsure of where, or even what, to look for. Priming the perceptual/cognitive system for specific features or objects is one way of helping observers to locate and identify target items (e.g., Grice&Gwynne, 1985; Laarni and Hakkinen, 1994). Past research has demonstrated that priming single features does indeed affect search performance (e.g., Hailston&Davis, 2006; Huang&Pashler, 2005). But, what happens when more than one feature is primed? Does priming two features result in better performance than priming only one? What about three features? How does feature priming compare to simply priming the entire object itself? The current research addressed these questions with a series of three visual search experiments. In the first experiment performance in simple feature search was compared against triple-conjunction search performance. Three prominent models of visual search were compared to see which best predicted actual performance. In the second and third experiments the effects of multiple feature priming on search accuracy were examined in a triple-conjunction search (Experiment 2) and a whole-object search (Experiment 3). Moreover, in Experiment 3 the effectiveness of whole-object primes were compared to multiple-features primes. Results show that none of the three models can accurately predict performance in all cases, suggesting some modification of each is necessary. Furthermore, valid primes resulted in performance benefits, and these benefits increased with the number of primed features. Finally, no performance costs of invalid priming were observed in the current experiments.