Facial Expression Recognition Task

Version: v1 (current)

An emotion processing task measuring the ability to identify basic emotions from facial expressions.

Overview

The Facial Expression Recognition task presents photographs or drawings of faces displaying emotions and asks participants to identify which emotion is shown. This fundamental social-cognitive ability is essential for social interaction and emotional communication.

Recognition of basic emotions develops early in childhood and is relatively preserved in typical aging, but deficits are seen in autism spectrum disorder, schizophrenia, depression, and alexithymia. The task is used to assess emotional processing across development, aging, and clinical conditions.

Scientific Background

Classic Findings:

• Universal Emotions: Six basic emotions (happiness, sadness, anger, fear, surprise, disgust) recognized across cultures (Ekman & Friesen, 1976)
• Happiness Advantage: Happy faces recognized fastest and most accurately
• Fear-Anger Confusion: Fear and anger sometimes confused, especially at low intensity
• Clinical Deficits: Reduced accuracy in autism (especially for complex emotions), schizophrenia, and depression
• Intensity Effects: Recognition accuracy improves with expression intensity

Seminal Papers:

• Ekman, P., & Friesen, W.V. (1976). Pictures of Facial Affect. Consulting Psychologists Press.
• Adolphs, R., Tranel, D., & Damasio, A.R. (2003). Dissociable neural systems for recognizing emotions. Brain and Cognition, 52(1), 61-69.

Why Researchers Use This Task

1. Emotion Research: Study basic emotion processing and recognition mechanisms
2. Clinical Assessment: Evaluate emotion recognition deficits in psychiatric and neurological conditions
3. Developmental Studies: Track emotion recognition development in children and changes in aging
4. Autism Research: Core social-cognitive deficit area in autism spectrum disorder
5. Treatment Evaluation: Assess improvements in social cognition interventions

Configuration Options

Visual Parameters

Parameter	Type	Default	Description
Image width	number	400px	Width of facial expression images (200-800px)
Font size	number	18px	Font size for emotion label buttons (8-72px)
Show image number	boolean	False	Display trial counter (e.g., "Image 3 of 24"); hidden by default, enable it only if your protocol calls for the participant to see their position in the sequence
Randomize option order	boolean	False	Randomize emotion button order on each trial

Timing Parameters

Parameter	Type	Default	Description
Response timeout	number	0ms	Maximum time to respond (0 = no timeout, self-paced)

Practice Trials

Parameter	Type	Default	Description
Practice enabled	boolean	True	Enable practice trials with visual feedback
Show feedback	boolean	True	Display checkmark (correct) or X (incorrect) during practice

Trial Configuration

Each trial is defined in the trials spreadsheet with the following columns:

Column	Required	Description	Example
image_url	Yes	URL to the facial expression image	https://example.com/faces/happy_01.jpg
correct_emotion	Yes	The correct emotion label	Happy, Sad, Angry, Fear
emotions	Yes	Comma-separated list of emotion options to display	happy,sad,angry,fear,surprise,disgust
intensity	No	Expression intensity level (for analysis grouping)	Low, Medium, High
block	No	Block identifier for grouping trials	Practice, Main, Block 1
response_timeout_ms	No	Response timeout for this trial in ms (empty = use general setting)	5000, 0

Example Trials:

| image_url              | correct_emotion | emotions                              | intensity | block |
|------------------------|-----------------|---------------------------------------|-----------|-------|
| faces/happy_high_01.jpg| happy           | happy,sad,angry,fear,surprise,disgust | high      | main  |
| faces/fear_low_03.jpg  | fear            | happy,sad,angry,fear,surprise,disgust | low       | main  |
| faces/sad_med_02.jpg   | sad             | happy,sad,angry,fear,surprise,disgust | medium    | main  |

Notes:

• Each trial can have a different set of emotion options
• Options are displayed as clickable buttons below the image
• If Randomize option order is enabled, button order changes on each trial

Practice Trials

If Practice enabled is checked:

1. Practice trials are presented first with visual feedback
2. After each response, a green checkmark (correct) or red X (incorrect) appears briefly
3. Participant clicks "Next" to continue (or moderator advances in moderated sessions)
4. After all practice trials, trials instructions are shown
5. Main trials begin without feedback

Practice is recommended to ensure participants understand the task and response options.

Instructions

The task supports four types of customizable instructions:

Instruction Type	When Shown	Purpose
Main instructions	Before the entire task	Explains the task: identify the emotion from facial expressions
Practice instructions	Before practice trials (if enabled)	Explains practice with feedback
Trials instructions	Before main trials (if practice enabled)	Signals transition from practice to scored trials
Hint instructions	Available via "?" button	Quick reminder accessible during the task

Data Output

Markers and Responses

Markers:

• stimulus_shown: Logged when a facial expression image is shown (both practice and main trials use this marker name)
  • Includes: trial index (trial_index in main, practice_trial_index in practice), is_practice, stimulus_id, image_url, correct_emotion, emotions, intensity, block
• response_recorded: Logged when the participant responds, carrying the phase-appropriate trial index and is_practice

Response Data:

• Each response row records: trial index, is_practice, stimulus_id, source, image_url, correct_emotion, emotions, response_value (selected emotion), response_correct, responded, intensity, block, and latency_ms (reaction time)
• Practice trials use the same response structure, marked with is_practice: true

Summary Artifact

A JSON file named fer_summary_<taskIndex>.json with comprehensive emotion recognition statistics:

{
  "task_kind": "facial_expression_recognition",
  "task_index": 0,
  "total_trials": 24,
  "overall": {
    "total": 24,
    "valid_responses": 24,
    "correct": 21,
    "score": 21,
    "total_possible": 24,
    "proportion_correct": 0.88,
    "mean_rt_ms": 1650,
    "timeouts": 0
  },
  "breakdowns": {
    "emotion": {
      "happy": {
        "total": 4,
        "correct": 4,
        "accuracy": 1.00,
        "mean_rt_ms": 1120,
        "confusion": { "happy": 4 }
      },
      "fear": {
        "total": 4,
        "correct": 3,
        "accuracy": 0.75,
        "mean_rt_ms": 2180,
        "confusion": { "fear": 3, "anger": 1 }
      }
    },
    "intensity": {
      "high": {"total": 8, "correct": 7, "accuracy": 0.95},
      "medium": {"total": 8, "correct": 7, "accuracy": 0.85},
      "low": {"total": 8, "correct": 6, "accuracy": 0.75}
    }
  },
  "trials": [...],
  "practice": {
    "total_trials": 6,
    "overall": {
      "total": 6,
      "valid_responses": 6,
      "correct": 5,
      "score": 5,
      "proportion_correct": 0.83,
      "mean_rt_ms": 1400
    },
    "breakdowns": {},
    "trials": [...]
  }
}

The intensity breakdown is only present when at least one trial supplies an intensity value. The practice block is only present when practice is enabled and practice responses were recorded.

Key Metrics:

• proportion_correct: Proportion of correctly identified emotions (over answered trials)
• breakdowns.emotion: Accuracy, reaction time, and a confusion response tally for each emotion type
• breakdowns.intensity: Performance grouped by expression intensity (if the intensity field is used)

Design Recommendations

Standard Configuration

Trials:

• 4-6 images per emotion × 6 emotions = 24-36 total trials
• Balanced design (equal number of each emotion)
• Mix of expression intensities

Options:

• Present all 6 basic emotions as choices on every trial
• Or use a reduced set (e.g., only positive/negative emotions)

Practice:

• 6-12 practice trials with feedback
• Include all emotion types in practice
• Use clear, high-intensity expressions

Population-Specific Adaptations

Children (6+ years):

• Larger images (500-600px width)
• Larger button text (20-24px)
• Simpler emotion labels (happy/sad/mad instead of happy/sad/angry)
• More practice trials (10-12)
• High-intensity expressions only

Older Adults (65+):

• Larger images and text
• Self-paced (no timeout)
• Fewer trials (4 per emotion = 24 total)
• Practice trials mandatory

Clinical Populations:

• Simplified emotion set (e.g., only 4 emotions: happy, sad, angry, fear)
• Extended practice with feedback
• Self-paced responses
• Consider intensity manipulation (start with high intensity)

Advanced Configurations

Intensity Study:

• Include intensity levels (low, medium, high) in the trial sheet
• Analyze accuracy as a function of expression intensity
• Expect linear improvement with intensity

Emotion-Specific Assessment:

• Focus on specific emotions of interest (e.g., fear and anger only)
• Useful for targeted clinical assessments

Dynamic Emotion Options:

• Vary the emotion options per trial (not always all 6)
• Increases difficulty and reduces guessing

Common Issues and Solutions

Issue	Solution
Very high accuracy (>95%)	Use lower intensity expressions; increase number of emotion options; use more subtle expressions
Very low accuracy (<60%)	Use higher intensity expressions; reduce number of options (4 instead of 6); add more practice
Participant not understanding task	Ensure practice trials enabled; clarify instructions; verify emotion labels are clear
Long reaction times (>3000ms)	Reduce image size; simplify emotion labels; check for reading difficulties
Technical: Images not loading	Verify image URLs are accessible; check internet connection; ensure correct file paths

Participant Experience

1. Main Instructions: Overview of the task (identify emotions from faces)
2. (Optional) Practice Instructions: Explanation that feedback will be shown
3. (Optional) Practice Trials: View face → click emotion → see feedback (✓ or ✗) → Next
4. (Optional) Trials Instructions: Transition message (feedback no longer shown)
5. Main Trials:
   • Facial expression image appears
   • Emotion buttons displayed below
   • Click the emotion button that matches the expression
   • Next trial automatically appears
6. Completion: Summary feedback (optional)

References

• Ekman, P., & Friesen, W.V. (1976). Pictures of Facial Affect. Consulting Psychologists Press.
• Adolphs, R., Tranel, D., & Damasio, A.R. (2003). Dissociable neural systems for recognizing emotions. Brain and Cognition, 52(1), 61-69.
• Baron-Cohen, S., Wheelwright, S., & Jolliffe, T. (1997). Is there a "language of the eyes"? Evidence from normal adults, and adults with autism. Visual Cognition, 4(3), 311-331.

See Also

• RMET Revised - Theory of mind assessment