Foundations and determinants of consumer attitudes toward commercial websites : an experimental study

• Main Author: Argyriou, Evmorfia P.
• Published: University of Warwick 2007
• Subjects: 381
• Online Access: http://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.490684

The thesis investigates a process-model of attitude formation in the context of information search in commercial websites. Contemporary research in social and consumer psychology is moving away from traditional functional and multi-attribute models of attitude, towards constructive theories and the notion of perceptual and heuristic information processing. The latter approach to attitude formation stresses the limitations of human working memory and perceptual sensitivity to stimuli, and therefore suggests that consumers are less likely to form and hold attitudes in memory toward an infinite number ofmundane objects. What this view proposes, is that attitudes are temporary evaluations, constructed 'on the spot' of a question from perceptually salient information in the context and are contingent on situational goals. Research in commercial website usage and re-visitation, complements this view of humans as perceptual and perhaps inaccurate information processors. A number of inter-related disciplines, namely usability research, human-computer interaction, information systems, and e-commerce research, are concerned with determining the optimal interface design for improving user experiences, in terms of task completion rates, user satisfaction with technology, technology adoption, and purchase behaviour. However, few research studies to date examine - in a systematic manner - the process by which consumers construct their attitudes towards commercial interfaces combining different levels of web design attributes. Due to the vastness of tools that marketers can choose from when designing their online interface, selecting stimuli that create an optimal context for information search on websites is crucial to website evaluation. In light of the above considerations, we synthesise a conceptual framework of consumer attitudes toward commercial websites, adopting a constructive viewpoint. In order to test our process-model of attitude formation in a natural, realistic setting, we conduct two webexperiments. Following theory and web-designers' insights, we manipulate three core design attributes, namely novelty in informative features, information layout on web pages, and interactivity in website navigation, and monitor the changes in consumers' perceptions of effort, attitudes, and intentions to use websites. In addition, in order to test for moderating effects of motivation, we manipulate search goals experimentally (undirected vs. directed information search). Finally, we assess the nuisance effects of relatively stable individual tendencies. Our first experiment (193 subjects) involves extreme manipulations of design, and results in large effects on consumer responses. Our second, large-scale experiment (676 subjects) involves refined manipulations of the same design attributes, and results in smaller-sized effects but significant interactions. Interpreted as a whole, our results support the contextual and temporary nature of attitudes towards commercial online interfaces, and are consistent with constructive perspectives on attitude formation. Regarding design attributes, absence of novelty in features and low interactivity in navigation have negative direct and combined effects on consumer responses. However, when the two attributes are manipulated unobtrusively, information layout is the sole influential design factor. Moreover, in both studies, individual design factors interact with participants' search goal, although participants involved in general search report consistently more favourable responses than those engaged in directed search. One of our major research contributions lies in the finding that attitudinal responses reflect a temporary evaluation of the context, which is shaped by the interactions of situational search goals and perceptually salient design aspects. Furthermore, responses to commercial websites are not explained by stable individual factors, although a number of these factors may have a direct impact on participants' responses. In accordance with these findings, we put forward a number of implications for marketers, web-designers, and academic researchers.