Third International Conference on Spoken Language Processing (ICSLP 94)

Yokohama, Japan
September 18-22, 1994

Identifying Salient Usability Attributes for Automated Telephone Services

Stephen Love (1), R. T. Dutton (1), J. C. Foster (1), Mervyn A. Jack (1), F. W. M. Stentiford (2)

(1) Department of Electrical Engineering, The University of Edinburgh, Edinburgh, UK
(2) BT Laboratories, Martlesham Heath, Ipswich, UK

This paper presents experimental results relating to the perceived usability of automated telephone services which incorporate a spoken language dialogue interface. Using a new Wizard of Oz experimental technique, users' attitudes towards automated telephone services are measured by a Likert attitude questionnaire. Factor Analysis performed on subjects' responses to the questionnaire produced a set of five underlying group factors which accounted for 74% of the variance in the questionnaire scores. On the basis of the factor analysis, a multiple regression model was constructed using the highest loading attributes on each of the five factors. The predictive power of this model was assessed.

Full Paper

Bibliographic reference.  Love, Stephen / Dutton, R. T. / Foster, J. C. / Jack, Mervyn A. / Stentiford, F. W. M. (1994): "Identifying salient usability attributes for automated telephone services", In ICSLP-1994, 1307-1310.