The Trier Social Stress Test (TSST) is a method to induce stress for research, traditionally done in controlled settings but adapted for remote use with employees.
The study utilized noninvasive sensors to gather data on stress responses from twenty participants, highlighting the importance of the State Trait Anxiety Inventory (STAI) score in evaluating effects.
The research employed machine learning techniques, achieving an F score of 0.723 with STAI as a baseline, and improved to 0.847 by using changes in STAI scores to account for subjective variations.