Modelling the relationship between design/performance factors and perceptual features contributing to Quality of Experience for mobile Web browsing

Abstract

Browsing different types of Web sites on mobile devices has become a part of many people's everyday activities. In the context of Quality of Experience (QoE) research, there is an interest in understanding the impact of various web site design and performance factors, i.e., QoE influence factors (IFs) on different perceivable characteristics of a user’s experience which contribute to its quality, i.e., QoE features. In this paper, we refer to our recently published multidimensional QoE modelling study (Baraković and Skorin-Kapov, 2015), whereby 77 participants took part in a field study and rated sixteen different versions of an information portal and a thematic portal, and eight versions of an e-mail portal by using a smartphone or a tablet. We extend the previously reported data analysis with new findings explicitly exposing the relationships between QoE IFs and QoE features. In this paper, we quantify mutual relations between identified QoE IFs (Web site loading time, Web site aesthetics, number of taps to reach desired Web content, quality of Web site information) and selected QoE features (perceived Web site loading time, perceived Web site usability, perceived Web site aesthetics, perceived quality of Web site information). Results of the analysis show the existence of : (i) negative impact of number of taps to reach the desired Web content on perceived Web site loading time; (ii) negative impact of Web site loading time (in case of smartphones), and positive impacts of Web site aesthetics and quality of Web site information (in case of thematic portal) on perceived Web site usability; (iii) negative impact of number of taps to reach the desired Web content and positive impact of quality of Web site information (in case of thematic portal) on perceived Web site aesthetics; and (iv) positive impact of Web site aesthetics on perceived quality of Web site information. Moreover, obtained multiple linear regression models, which describe the mutual relations between considered QoE IFs and QoE features, extend previously reported multidimensional mobile Web QoE models and thereby enable the identification of the importance of distinct factors in terms of perceived QoE features.