WS8: Repeated Measures Designs: Visual Search

Question 0

Treisman’s research

Answer 0

• RTs increased linearly w no of distractors when Ss search for a conjunction of features. Require attention to be focused serially on each object in turn
• for single feature targets, decision RTs are unaffected by the number of distractors. Reject distractors simultaneously.
• pop out - when unique feature calls attention to its location

Question 1

Visuals search

Answer 1

• early processing involves analysis of simple features, such as orientation, colour, spatial frequency
• in VS, Ss is asked to inspect an array and to decide whether a specified target is present or not. Size of disp and rship of target to distractors can be manipulated. Decision times and accuracy are measured for diff types of targets and displays

Question 2

Treisman’s Feature Integration Theory

Answer 2

1) visual features (colour, size, lines) are processed in parallel pre-attentively
2) feature combination to form objects occurs as a serial process (e.g. Large blue square)
3) the serial process is slower than the parallel process
4) focused attention to the location of the object can be used to combine features. It provides the ‘glue’ that constructs unitary objects from the available features
5) feature combination can also be influenced by stored knowledge about the objects class
6) in the absence of focused attention or stored knowledge, features can be combined from different objects in a random fashion causing ‘illusory conjunctions’

Question 3

Search asymmetry

Answer 3

• the distinction b/w pre-attentive and attentive features is complicated by observations that search times can vary when the role of target and distractors are reversed
• the time to identify a circle with an intersecting line (Q) among regular circles (O) is independent of the number of distractors (regular circles), the time to identify a regular circle among circles with lines increases linearly with the number of distractors

Question 4

Order effects can arise with repeated measures designs. What is the best way to control for order?

Answer 4

• counterbalancing: present in diff order to diff Ss.
• randomize order of tasks or items
• order of words within a list can also be randomized