Force and position perturbations are widely applied to identify muscular and reflexive contributions to posture maintenance of the arm. Both task instruction (force vs. position) and the inherently linked perturbation type (i.e., force perturbations-position task and position perturbations-force tasks) affect these contributions and their mutual balance. The goal of this study is to explore the modulation of muscular and reflexive contributions in shoulder muscles using EMG biofeedback. The EMG biofeedback provides a harmonized task instruction to facilitate the investigation of perturbation type effects irrespective of task instruction. External continuous force and position perturbations with a bandwidth of 0.5-20 Hz were applied at the hand while subjects maintained prescribed constant levels of muscular co-activation using visual feedback of an EMG biofeedback signal. Joint admittance and reflexive impedance were identified in the frequency domain, and parametric identification separated intrinsic muscular and reflexive feedback properties. In tests with EMG biofeedback, perturbation type (position and force) had no effect on joint admittance and reflexive impedance, indicating task as the dominant factor. A reduction in muscular and reflexive stiffness was observed when performing the EMG biofeedback task relative to the position task. Reflexive position feedback was effectively suppressed during the equivalent EMG biofeedback task, while velocity and acceleration feedback were both decreased by approximately 37%. This indicates that force perturbations with position tasks are a more effective paradigm to investigate complete dynamic motor control of the arm, while EMG tasks tend to reduce the reflexive contribution.