Automatic Mechanisms for Measuring Subjective Unit of Discomfort: Erratum, datafile and analysis scripts
datasetposted on 15.10.2018 by Dwi Hartanto, N. (Ni) Kang, Willem-Paul Brinkman, I.L. (Isabel) Kampmann, N. (Nexhmedin) Morina, P.M.G. (Paul) Emmelkamp, M.A. (Mark) Neerincx
Datasets usually provide raw data for analysis. This raw data often comes in spreadsheet form, but can be any collection of data, on which analysis can be performed.
Current practice in Virtual Reality Exposure Therapy (VRET) is that therapists ask patients about their anxiety level by means of the Subjective Unit of Discomfort (SUD) scale. With an aim of developing a home-based VRET system, this measurement ideally should be done using speech technology. In a VRET system for social phobia with scripted avatar-patient dialogues, the timing of asking patients to give their SUD score becomes relevant. This study examined three timing mechanisms: (1) dialogue dependent (i.e. naturally in the flow of the dialogue); (2) speech dependent (i.e. when both patient and avatar are silent); and (3) context independent (i.e. randomly). Results of an experiment with non-patients (n = 24) showed a significant effect for the timing mechanisms on the perceived dialogue flow, user preference, reported presence and user dialog replies. Overall, dialogue dependent timing mechanism seems superior followed by the speech dependent and context independent timing mechanism. This dataset contains erratum, data file and analysis script to the publication Automatic Mechanism for Measuring Subjective Unit of Discomfort