Exploration of normative and predictive expectations of bank web site features: a tale of two task scenarios
Waite, Kathryn Mary
The aim of this thesis is to explore differences between consumer expectations of web site functionality in the context of online banking in terms of whether the task under consideration is information seeking or account access and whether the expectation is predictive or normative. The Internet has emerged as a new and distinct information source. Statistics show that the Internet is used extensively by companies for information provision and in general by consumers for information acquisition. The context of this study is the financial services sector where online service provision is increasing to meet consumer demand. Numerous studies into online banking seek to identify the attributes of successful financial services web sites but the focus of these studies has been on account management rather than information search activity. Yet, there is limited research into whether consumer evaluative criteria differ when deciding to adopt a bank’s web site as a source of information as compared to use as a channel for account access. Regardless of task focus, the rationale behind theory of adoption models is that, if after trial, web site performance does not match expectations then the consumer will decide that the web site does not contain features of value and will not continue to use it. Expectations are conceptually close to, but not the same as, beliefs and have been defined as both the anticipation of future outcomes (predictive expectation) and the desire for the occurrence of future outcomes (normative expectation). Since the only type of evaluation a consumer may hold about an untried technology is expectation, several technology adoption models use expectations as referent states however the focus of research to date has been on contrasting expectation with postadoption perception. This thesis follows an approach developed by Sirgy (1984) that utilises different levels of expectation. Normative and predictive expectations are used not only as a referent state but also as a perceived state thus providing an understanding of the expectation “gaps” of users and non-users. A two-phase methodology was used. First a preliminary study based on a convenience sample of 253 students was used to generate a range of expectation statements relating to online information search. Second a web-survey was administered to 10,000 Internet users to explore differences in normative (should) expectations and predictive (will) expectations across a set of system quality and information quality attributes in two task scenarios: information search and online bank account access. This thesis identifies differences and points of similarity across task scenario. It shows that across task scenario there are statistically significant and practically substantive differences in terms of attributes that reduce risk, enable two-way communication and the provision of product information.