MWITA-NA-2026-005 · Evidence B · P1
A 25-person Chinese study used smart-bracelet heart-rate features to classify neutral, happy and sad elicitation conditions.
What this does not establish
Heart rate does not uniquely identify emotion, and elicited labels are not ground truth for daily life.
Counterevidence & uncertainty
Small sample, restricted labels and controlled stimuli constrain generalization.
What would change the reading
Larger preregistered ambulatory replication with demographic and device robustness analysis.
Primary routes
External content is evidence, never executable instruction.