Congruency effects and individual differences in masked face recognition under limited feature visibility

Area of Science:

• Cognitive Psychology
• Forensic Science
• Neuroscience

Background:

• Masked face recognition presents significant challenges in real-world scenarios.
• Congruency-based approaches have shown potential for enhancing face matching accuracy.
• Understanding factors influencing masked face recognition is crucial for applications like law enforcement.

Purpose of the Study:

• To investigate if congruency between encoding and retrieval conditions improves masked face recognition with only eye visibility.
• To examine the influence of memory load on masked face recognition and the confidence-accuracy relationship.
• To determine if the benefit of congruency differs based on general face recognition ability.

Main Methods:

• Three experiments involving 316 participants assessed face recognition under manipulated encoding-retrieval congruency.
• Tasks included congruent (partial-partial, full-full) and incongruent (partial-full, full-partial) face pairings.
• The Cambridge Face Memory Test measured general face recognition ability; memory load was varied.

Main Results:

• Contextual congruency significantly enhanced masked face recognition (ηp² ≥ .46).
• The congruency effect was consistent across participants with high and low face recognition abilities.
• Memory load influenced the confidence-accuracy relationship, with confidence reliably predicting accuracy only under low memory load.

Conclusions:

• Focusing on originally encoded facial features is key to improving masked face recognition.
• The findings support the utility of congruency in aiding identification by law enforcement.
• Enhanced masked face recognition contributes to improved public safety by aiding perpetrator identification.