Abstract-Wide-area early warning systems are dependent on synchrophasor data-based applications for providing timely information to operators so that preventive actions can be taken. This article proposes the use of voltage sensitivities computed from syncrhophasor data for voltage stability monitoring, and a visualization approach that can be implemented in wide-area early warning systems. In order to provide reliable information, synchrophasor data must be pre-processed to extract only the useful features embedded in measurements and correct for errors. Hence, this article also addresses the issue of data filtering and correction, and proposes a filtering methodology for robust voltage sensitivity computation. The methodology is developed considering both positive-sequence simulations for methodology development purposes, and real phasor measurement data from a real-time (RT) hardware-in-the-loop (HIL) laboratory for testing the robustness of the developed approach for use in the control room. The results from both approaches are contrasted against each other, the limitations of the positive-sequence simulation approach for developing PMU-data applications are highlighted and the challenges of working with the RT HIL lab are recognized.