Visual Search Remains Efficient When Visual Working Memory


Published on

  • Be the first to comment

  • Be the first to like this

No Downloads
Total views
On SlideShare
From Embeds
Number of Embeds
Embeds 0
No embeds

No notes for slide

Visual Search Remains Efficient When Visual Working Memory

  1. 1. VISUAL SEARCH REMAINS EFFICIENT WHEN VISUALWORKING MEMORY IS FULL<br />Geoffrey F. Woodman, Edward K. Vogel, and Steven J. Luck<br />PSYCHOLOGICAL SCIENCE (2001)<br />
  2. 2. This article examines interactions between visual attention and visual working memory in the context of visual search.<br />
  3. 3. visual attention<br />the RTs increase linearly as the number of items in the arrays is increased<br />
  4. 4. visual working memory<br />visual working memory also plays a significant role in visual search<br />a template of the search target may be stored in VWM (Desimone & Duncan, 1995) <br />once a target is detected,it must be transferred into WM so that it can beused to control overt behavior (Duncan, 1980)<br />when attention is focused on an item, this item is automatically transferred into VWM<br />
  5. 5. attention & VWM<br />the searcheditems are stored in WM<br />serial models<br />parallel models<br />
  6. 6. attention & VWM<br />the searcheditems are NOT stored in WM<br />visual objectidentification can occur more rapidly than the encoding of informationinto visual working memory (Jolicoeur & Dell’ Acqua, 1998; Potter,1976; Thorpe, Fize, & Marlot, 1996; Vogel, Luck, & Shapiro, 1998)<br />observers do not even remember the locations of the items they have searched, so attention may revisit a given nontarget item multiple times.(Horowitz and Wolfe, 1998) <br />
  7. 7. goal<br />whether visual search does indeed involve thetransfer of information into visual working memory<br />
  8. 8. EXPERIMENT 1<br />SS = 4, 8, or 12<br />
  9. 9. Results<br />3.2 items<br />2.7 items<br />Equation (Pashler, 1988)<br />
  10. 10. Discussion<br />change in intercept; NO change in slope<br />memory task led to a slowing of a process that eitherpreceded or followed the actual search<br />there was NO evidence that the search process itself was impaired by the addition of a memory load<br />
  11. 11. Discussion<br />impairment in memory accuracy<br />nonspecific disruption of the WM representation by the mere appearance of the search array<br />->Experiment 3<br />
  12. 12. Discussion<br />performance: single task = dual task<br />different strategies<br />memory set size: 2, 4 (within blocks)<br />-> = EXP 1<br />the memory task used in Experiment 1 did not actually fill working memory to capacity<br />condition 1: 4 colored squares<br />condition 2: 4 colored squares + square with the gap<br />-> accuracy: condition 1 &gt; condition 2<br />
  13. 13. EXPERIMENT 2<br />
  14. 14. Results<br />
  15. 15. EXPERIMENT 3<br />Experiment 3 investigated thecause of the small impairment in memory performance that was observed in the dual-task conditions of Experiments 1 and 2.<br />ignore the search array<br />
  16. 16. Results and Discussion<br />the deficit in memory performanceobserved in the dual-task conditions of the previous experimentswas not due to visual search per se, but instead reflects somesort of nonspecific masking or interruption<br />
  17. 17. GENERAL DISCUSSION<br />visual search causes minimal displacementof information already in VWM<br />theefficiency of the search process is not impaired when VWM is filled to capacity<br />intercept ↑: a delay in the onset of the search process or a delay in postsearch processes such as response selection<br />
  18. 18. GENERAL DISCUSSION<br />it is possible that a different working memory task would interfere with the search process<br />objects canbe attended at a perceptual level without automatically being enteredinto WM<br />