Saskia Baltrusch

208 Chapter 8 explanation of the exoskeleton (Figure 1, more details on the exoskeleton see [14]), thus, they rated their self-efficacy based on their expectations of the exoskeleton (EXPECTATION condition), and 3) after wearing the exoskeleton during a set of daily activities, such as walking, lifting and forward bending with and without the exoskeleton (TRY-OUT condition) (for more information on test battery see [30]). The performed tasks were not the same tasks as depicted in the questionnaire, but provided the participants with sufficient experience of the device to evaluate its potential benefits. Figure 1: The exoskeleton (a) unloads the back by applying forces at the torso, pelvis, and the thighs, generating a torque by two serially connected passive actuators: an elastic spinal module (a) and a hip actuator (b) . The implemented clutch allows disengagement of the passive hip actuators, by moving a manual switch (c) . Data analysis The total score of the M-SFS questionnaire was calculated for the three different assessments and averaged over participants. To find differences between the self-efficacy scores of BASE, EXPECTATION and TRY-OUT measurement we applied the non-parametric Friedman test. In case of a significant effect of condition, Wilcoxon post-hoc tests were conducted to determine differences between conditions. Alpha of 0.05 was used as the critical level of significance. Statistical analysis was performed using Matlab (R2015b). To get more insight into the effect of the exoskeleton on individual tasks, we categorized the twenty tasks of the M-SFS into 5 categories: 1) Lifting (Tasks 1, 4, 5, 10, 11, 14, 18), Repetitive Bending (Tasks 17, 19), 3) Standing and Walking (Tasks 3, 13, 16), 4) Static Forward Bending (Tasks 6, 15), 5) Sitting (Tasks 2, 20) and 6) Others (Tasks 7, 8, 9, 12). For each category, we calculated the percentage of