Electrodermal signatures of stress during tactical law enforcement training

Law enforcement tactical training places sustained cognitive and emotional demands on recruits, making stress regulation a critical determinant of learning and performance. This study examined EDA as an objective index of stress regulation during standardized law enforcement scenarios. Objectives were to 1) predict individual change in tactical performance from electrodermal dynamics, 2) test whether a breathing-based intervention produces different pre- to post-physiological change than standard training across the full scenario periods, and 3) quantify peri-threat EDA dynamics during threat recognition. Fifty-three recruits completed repeated pre- and post-training evaluations. Continuous EDA was recorded via wearable sensors across four scenarios. Performance change was defined as post-training minus baseline performance scores and modeled using machine learning with repeated 5-fold cross-validation. Random forest models explained 30% of the variance in performance change, and improvers versus non-improvers were classified with 77.1% accuracy, driven primarily by sympathetic energy and phasic amplitude. Whole-scenario analyses showed robust Time effects but no reliable Group or Time × Group effects. In the peri-threat window, onset latency showed a Time × Group interaction (p = .031, d = 0.86), with shorter post-training latency in Controls and stable latency in Experimental. Findings indicate that officer recruit tactical performance can be predicted from electrodermal dynamics measured during scenario-based training, and that performance improvement can be predicted from stress regulation dynamics. Finally, a brief biofeedback-based breathing intervention significantly slowed down the threat-triggered sympathetic arousal. The results support the use of biofeedback-based stress interventions in recruit training to shape autonomic responses and improve tactical performance.