Validity Assessment of Wrist-Worn Electrodermal Activity Measurement During Relaxation and Acute Stressor Tasks: Comparing the EmbracePlus and Shimmer3 GSR+

This is the authors' abstract. We don't add key points for this paper.

Wearable devices, such as the Empatica E4, are increasingly used to assess stress-related electrodermal activity (EDA). The EmbracePlus has replaced the widely used Empatica E4. This study evaluated the validity of the EmbracePlus by comparing wrist-worn EDA measurements with the Shimmer3 GSR+ across signal-, parameter-, and event-level analyses. Healthy volunteers (n=53) completed breathing exercises, a Stroop stress task, and a virtual reality (VR) height stressor while wearing both devices simultaneously on the non-dominant wrist. Signal-level agreement was assessed using cross-correlation, parameter-level agreement using Bland–Altman analyses of mean skin conductance level (SCL), total phasic activity, skin conductance response (SCR) amplitude, and SCR count, and event-level agreement using within-device baseline-to-task comparisons and inter-device difference plots with a priori acceptability boundaries. Signal-level agreement was weak-to-moderate. Shimmer3 yielded systematically higher values across all metrics, indicating limited agreement in absolute EDA values. Shimmer3 detected significant increases across all metrics during both stress tasks. EmbracePlus detected the stronger VR stressor across all metrics, whereas during the milder Stroop task only total phasic activity and mean SCL increased. These findings indicate that EmbracePlus and Shimmer3 are not interchangeable for absolute EDA assessment and that EmbracePlus validity depends on stressor intensity and metric selection.