Visual Search and Automaticity

Psy315 - Class 11, Part 2

(This page last updated 12 October, 2006.)

 Spatial cuing experiments measure the allocation of attention to location and objects.

Visual search tasks are used to determine which aspects of visual processing can be performed preattentively, in parallel across the visual field, and which tasks require attentional selection.

In all the search displays below, search for the red vertical line.

Feature Search by Color
Feature Search by Orientation
Conjuunction Search

Varying the Set Size of the Search Array

For conjunction search, RT increases with set size (serial search). For feature search, RT is fast, regardless of set size (parallel search).

Other methods besides visual search have also been used to investigate the role of visual features in attention.

• texture segregation
• illusory conjunctions
  • Attention is necessary to combine features from different dimensions.
  • If features cannot be correctly "bound" by attention, then they may be combined incorrectly.

Treisman's Feature Integration Theory

• Differences among features of a single type (such as color) can be detected in parallel, without attention.
• Features of different types (such as color and orientation) can only be combined with attention. If a search task requires finding a feature conjunction, search will be serial.

What are the basic features in vision?

• Color
• Orientation
• Curvature
• Size
• Motion
• Shape
• Depth
• …

Automaticity: How Practice Affects Visual Search

Experiment by Shiffrin & Schneider

• Search for target letter among distractors.
• First see target set - remember them.
• Then see test stimulus
  • Array of letters
• If any target letter is present in test stimulus, respond yes.
  • Otherwise, respond no.
• The number of letters in the test stimulus varies from one trial to the next.
  • from 1 to 4 letters
• Response time increases with number of letters
  • serial search!

Shiffrin & Schneider used Two Kinds of Practice

• Consistent Mapping
  • Letters that were targets on one trial were never distractors on other trials, and letters that were distractors on one trial were never targets.
• Varied Mapping
  • Letters that were targets on one trial were distractors on other trials, and letters that were distractors on one trial were targets on others.

The effect of practice depended on the type of mapping.

• With varied mapping practice, search got somewhat faster with practice, but remained serial.
  • Response time still increases with number of letters.
• With consistent mapping practice, search became parallel!
  • Response time is the same whether there is 1 letter or 4 letters.

When subjects can learn which letters will always be targets and which will be distractors, their search improves greatly (from serial to parallel).
When they must search for a given letter on one trial and ignore it on another, their improvement is very limited.

 After extended consistent practice, the task was reversed, so that target letters became distractors and distractors became targets.
Now search was serial and very difficult.

With enough practice, some cognitive processes can be come "automatic". They can be done more easily, without interfering with other tasks.
However, the practice must be consistent.

Stroop Task: Interference in Naming Colors

YELLOW RED BLUE

Reading is automatic; you cannot turn it off. The color name that you read interferes with the color name you are trying to recall and to speak.

next topic: Reading

Psych 315: Cognitive Psychology Kyle Cave Psychology Dept. U. Mass.