Visual Search Task

Version: v1 (current)

A classic attention paradigm measuring the efficiency of finding targets among distractors based on feature similarity.

Overview

Visual Search tasks present participants with arrays of items and ask them to locate a target among distractors. Search efficiency depends on whether the target "pops out" due to a unique feature (feature search: fast, parallel) or requires scrutiny of individual items (conjunction search: slow, serial). The relationship between response time and set size (number of items) reveals the underlying search strategy.

Feature searches (e.g., red circle among green circles) produce flat RT slopes (~5ms/item), indicating parallel processing. Conjunction searches (e.g., red circle among red squares and green circles) produce steep slopes (~20-40ms/item), indicating serial, attention-demanding search. This dissociation has been central to attention theories for decades.

Researchers use visual search to study attention, perception, working memory, and aging. Clinical applications include measuring attentional capacity in ADHD, neglect, and visual processing disorders.

Scientific Background

Classic Findings:

• Feature vs. Conjunction: Feature searches are efficient (parallel); conjunction searches are inefficient (serial)
• Set Size Effect: RT increases linearly with number of distractors in conjunction search
• Pop-Out: Targets with unique features are detected in <5ms/item regardless of set size
• Target-Present vs. Target-Absent: Absent trials show steeper slopes (~2x present slopes in serial search)
• Guidance: Partial information can guide attention to reduce effective set size

Search Types:

• Feature Search: Single dimension difference (color, shape, orientation)
• Conjunction Search: Target defined by combination of features
• Spatial Configuration: Targets defined by unique spatial arrangement

Seminal Papers:

• Treisman & Gelade (1980): Feature Integration Theory
• Wolfe (1994): Guided Search model
• Duncan & Humphreys (1989): Visual search and stimulus similarity

Why Researchers Use This Task

1. Attention Research: Distinguish parallel vs. serial processing mechanisms
2. Visual Processing: Study preattentive vs. attentive vision
3. Clinical Assessment: Measure attentional capacity and visual processing speed
4. Aging Studies: Track age-related slowing in visual search
5. Interface Design: Test visibility of targets in complex displays (e.g., medical imaging)

Current Implementation Status

Fully Implemented:

• ✅ Feature search (single-dimension differences)
• ✅ Conjunction search (two-dimension combinations)
• ✅ Configurable set sizes (4, 8, 12, 16, etc.)
• ✅ Target-present and target-absent trials
• ✅ RT slope calculation by set size
• ✅ Practice trials with feedback

Partially Implemented:

• ⚠️ Limited to color and shape features (no orientation, size, motion)
• ⚠️ Fixed spatial layouts (random positions)

Not Yet Implemented:

• ❌ Contextual cueing (repeated configurations)
• ❌ Eye-tracking integration for scan paths
• ❌ Custom feature dimensions

Configuration Parameters

Display Options

Parameter	Type	Default	Description
Item Font Size (px)	number	48	Font size of each search item (range 8-200)
Grid Columns	number	6	Number of columns in the search grid (range 2-20)
Grid Rows	number	6	Number of rows in the search grid (range 2-20)
Time-based mode	boolean	false	When enabled, trials auto-advance on the per-trial stimulus timeout; otherwise the trial is self-paced
Practice enabled	boolean	true	Whether a practice phase runs before the main trials

Trial Spreadsheet

Trials (and practice trials) are defined row-by-row in a spreadsheet rather than as global parameters. Each row has the following columns:

Column	Type	Default	Description
Set Size	number	4	Number of items shown (clamped to 1-50, and to grid capacity)
Target Present	checkbox	false	Whether the target is present on this trial
Target Feature	text	T	Character used for the target
Distractor Feature	text	L	Character used for the distractors
Search Type	dropdown	feature	feature or conjunction
Block	text	(empty)	Optional block label
Fixation (ms)	number	500	Pre-display fixation cross duration
Stimulus (ms)	number	0	Search display duration (0 = until response)

The default main set ships with 12 rows (set sizes 8 and 12, feature and conjunction, target present/absent); the default practice set ships with 4 rows (set sizes 4 and 6).

Keyboard Shortcuts

Researchers can customize the keyboard bindings used during the task:

Parameter	Type	Default	Description
Show keyboard hint	boolean	True	Display an on-screen hint showing the configured keys
Present key	key	P	Key for "target present" response
Present action label	text	"Target present"	Label shown in the keyboard hint
Absent key	key	A	Key for "target absent" response
Absent action label	text	"Target absent"	Label shown in the keyboard hint

Data Output

Markers and Responses

The task records high-resolution timestamps in two separate collections:

Markers (stimulus_shown):

{
  "type": "stimulus_shown",
  "ts": "2024-01-01T00:00:01.000Z",
  "hr": 1234.56,
  "data": {
    "trial_index": 1,
    "stimulus_id": "visual_search_0_1",
    "set_size": 8,
    "target_present": true,
    "search_type": "feature",
    "expected_response": "present",
    "block": "main",
    "is_practice": false
  }
}

Response Data:

{
  "trial_index": 1,
  "stimulus_id": "visual_search_0_1",
  "source": "keyboard",
  "raw_key": "p",
  "set_size": 8,
  "target_present": true,
  "search_type": "feature",
  "expected_response": "present",
  "response_value": "present",
  "response_correct": true,
  "latency_ms": 485,
  "block": "main",
  "is_practice": false
}

Summary Artifact

A JSON file (visual_search_summary_<taskIndex>.json) with aggregated statistics including RT x set size slopes:

{
  "task_kind": "visual_search",
  "task_index": 0,
  "total_trials": 40,
  "overall": {
    "total": 40,
    "valid_responses": 38,
    "correct": 35,
    "accuracy": 0.921,
    "mean_rt_ms": 1245,
    "mean_correct_rt_ms": 1198,
    "timeouts": 2
  },
  "breakdowns": {
    "target_presence": {
      "present": { "total": 20, "accuracy": 0.95, "mean_correct_rt_ms": 1050 },
      "absent": { "total": 20, "accuracy": 0.89, "mean_correct_rt_ms": 1350 }
    },
    "search_type": {
      "feature": { "total": 20, "accuracy": 0.98, "mean_correct_rt_ms": 850 },
      "conjunction": { "total": 20, "accuracy": 0.86, "mean_correct_rt_ms": 1550 }
    },
    "set_size": {
      "4": { "present": 900, "absent": 1100 },
      "8": { "present": 1100, "absent": 1500 },
      "16": { "present": 1400, "absent": 2200 }
    }
  },
  "trials": [ /* per-trial data */ ]
}

When practice is enabled, an additional practice object (with its own total_trials, overall, breakdowns.target_presence, and trials) is appended.

Key metrics:

• breakdowns.target_presence: Performance split by target-present vs. target-absent trials
• breakdowns.search_type: Performance split by feature vs. conjunction search
• breakdowns.set_size: Mean RT by set size for present and absent correct trials (for slope analysis)
• response_correct: Whether participant correctly identified target presence/absence

Example Research Configurations

Feature Search (Efficient)

Target: Red circle
Distractors: Green circles
Set Sizes: 4, 8, 12, 16
Expected: Flat slopes (<10ms/item), high accuracy

Conjunction Search (Inefficient)

Target: Red circle
Distractors: Red squares AND green circles
Set Sizes: 4, 8, 12, 16
Expected: Steep slopes (20-40ms/item present, 40-80ms/item absent)

Aging Comparison

Search: Both feature and conjunction
Set Sizes: 4, 8, 12 (smaller range for older adults)
Analysis: Compare slopes between age groups

Participant Experience

1. Instructions: "A grid of letters will appear. Decide whether the target letter (shown at the top) is present or absent. Press P for present, A for absent."
2. Practice: Try searches with feedback (default 4 trials, mixed set sizes)
3. Main Trials: For each trial:
   • Fixation point (500ms)
   • "Target: X" header appears above the grid showing the current trial's target letter (driven by the target_feature column, defaults to T)
   • Search display appears with multiple items
   • Scan for target as quickly as possible
   • Respond present/absent
4. Completion: See overall accuracy and average RT

All keyboard bindings are configurable by the researcher in the study configuration. The keys listed above are the defaults.

Design Recommendations

General Guidelines

• Set Size Range: Use 4-16 items for adults; 4-8 for children/clinical
• Target Presence: Balance present/absent (50/50) to prevent bias
• Trials per Condition: 10-15 trials per set size for stable slope estimates
• Feature Selection: Use highly discriminable features (avoid subtle color differences)

Designing Efficient (Feature) Search

• Single feature difference: color, shape, orientation, or size
• Large perceptual difference (red vs. green, not red vs. orange)
• Homogeneous distractors (all same)

Designing Inefficient (Conjunction) Search

• Target defined by 2+ features (red circle = red + circle)
• Each feature also present in distractors separately
• Requires checking feature combinations

Population-Specific Adaptations

Children (8+ years):

• Smaller set sizes (4, 6, 8)
• Larger items (60px)
• Simpler features (shape > color for young children)
• Fewer trials

Older Adults (65+):

• Larger items and spacing
• Fewer set sizes (4, 8, 12)
• Generous time limits
• Expect overall slowing but similar slope patterns

Clinical Populations:

• Neglect: Watch for left-right asymmetry in search
• ADHD: May show increased RT variability
• Visual processing disorders: Consider feature discriminability

Common Issues and Solutions

Issue	Solution
No set size effect in conjunction search	Ensure target truly requires conjunction (check distractor types)
Set size effect in feature search	Target may not be sufficiently distinct; increase feature salience
High error rate on absent trials	Give more practice; emphasize accuracy over speed
Response bias (favoring present)	Balance target presence; provide accuracy feedback
Items overlap	Increase minimum spacing or reduce item size

References

• Treisman, A. M., & Gelade, G. (1980). A feature-integration theory of attention. Cognitive Psychology, 12(1), 97-136.
• Wolfe, J. M. (1994). Guided Search 2.0: A revised model of visual search. Psychonomic Bulletin & Review, 1(2), 202-238.
• Duncan, J., & Humphreys, G. W. (1989). Visual search and stimulus similarity. Psychological Review, 96(3), 433-458.
• Wolfe, J. M., & Horowitz, T. S. (2017). Five factors that guide attention in visual search. Nature Human Behaviour, 1(3), 0058.

See Also

• Attentional Blink - Temporal attention limits
• Posner Cueing - Spatial attention orienting