Randomized trial validates morphed emotional face stimuli for recognizing emotions in varying genders, indicating improved reliability and effectiveness.
Key Points
This research aims to develop and validate a standardized set of morphed emotional face stimuli, addressing biases in emotion recognition.
Developed STEMorph using neutral-anchored morphing and neural-network-generated masks to enhance quality.
Participants rated faces on a 9-point scale for emotional intensity, followed by a two-week reliability assessment.
Employed linear mixed-effects modeling to analyze subjective ratings against intended morph levels.
STEMorph demonstrated a strong association between subjective ratings and intended morph level (β = 0.735, p < 0.001).
A significant variation in ratings associated with participant gender was found (β = 0.325, p < 0.001).
Reliability confirmed with high agreement between initial and follow-up ratings (ICC = 0.858).