Experiment 2

Is multiple target search always inefficient (target switching) or is efficiency constrained by the similarity of multiple targets (sub-set selection)?

Possible outcomes:

1. Simultaneous search not possible: target switching cost will always render multiple target search inefficient.
   
2. Efficiency constrained by colour difference: search efficiency for multiple targets will vary as a function of their separability in feature space.

Method

• Stimuli: 16 colour patches as in Experiment 1.
  
• Participants: N=26.
  
• Task: search for either of two targets.
  
• Target pairs: Distance between targets in colour-space varied using three separations of 1, 4 and 8 steps apart

Figure 6: Target A and B pairs with different separations in colour space

• Display: as in Figure 3, but with the search array displayed until a response was given.
  
• Set sizes: 4, 8, 12 or 16 items.
  
• Mean RT was compared for present and absent trials as a function of separation and set size.

Results

Figure 7: Mean RT as a function of set size for target separations of 1, 4 and 8 steps in colour space.

• Relatively efficient search for two targets that are adjacent in colour space.
  
• Inefficient search for two targets 4 or 8 steps apart in colour space.

Conclusions

Figure 8: Possible subsets which may be formed for each target pair.

• Separation of 1: subset can be formed containing targets and no distractors, giving efficient search.
  
• Larger separations: subset containing targets also contains distractors, reducing search efficiency.
  
• Supports attentional mechanisms that select a subset of items within a contiguous region of colour space.
  
• Implies top-down activation constrained to a single attentional template - multiple templates cannot be constructed.
  
• Suggests that search for multiple targets with between-dimension differences (Quinlan and Humphreys 1987, Treisman 1988) is much easier than for within-dimension differences.

On to Implications and Further Work.

Back to Experiment 1.