earticle

영어

Most IS studies take the position that the effect of critical success factors (CSFs) of IT services on user satisfaction/usage is one-dimensional or linear: the higher the influence of a CSF, the higher the level of measured performance (e.g., service satisfaction, adoption, and/or usage). Most theories and research models adopted by IS researchers, such as the theory of reasoned action and the technology acceptance model, are grounded on this paradigm in which unidimensionality is presumed implicitly or explicitly. However, sporadic arguments have been made that adequate prediction of user perceptions and behaviors is difficult when research is grounded on such a one-dimensional premise (Ahuja and Thatcher, 2005; Cenfetelli, 2004; Mittal et al., 1998). The motivators and de-motivators to use a service/product could represent two divergent concepts rather than flip sides of a coin. For instance, motivators to use a service improve user satisfaction if they are fulfilled, but their absence may not necessarily result in a proportional degree of user dissatisfaction. Likewise, de- motivators to use a service may increase user dissatisfaction, but their alleviation may not necessarily enhance satisfaction with the service proportionally. However, empirical studies that examine the dichotomy or directional discrepancy of influencing forces on the success of an IT service are scarce, making it difficult to judge the theoretical integrity of what may be called the ‘dichotomy’ view. In this article, I would like to introduce a recent research result, co-produced with Bongsik Shin from San Diego State University, to test the dual nature of CSFs. In adopting the duality perspective, we have taken the position that the influence of a CSF as an antecedent variable on the consequence variable(s) of an IT service can be asymmetric (i.e., non- linear). From the symmetric viewpoint, the association between a given CSF and outcomes remains proportionately uniform: a higher-level CSF results in a higher level of performance, and vice versa. However, if the influence of a CSF is asymmetric, its effect on the consequence variable(s) must be stronger in one direction than in the other direction. To test the validity of our theoretical position, data were gathered on post-adoption usage of digital data services offered by carriers of mobile devices (mobile data service or MDS), currently the most prevalent form of IT-enabled service. The inundation of web-based advanced data services for smart phones (e.g., iPhone), tablets (e.g., iPad), laptops, and other personal digital devices is well known. With the launch of high-speed mobile networks running 4G technologies, there has been a surge of various mobile data services such as social networking, video conferencing, streaming of movies and user-created content, and online gaming. This explosive growth trend should continue with the meteoric rise of smart phones. Two sets of empirical data (a purely self-report-based, cross-sectional dataset and an actual usage-based, longitudinal dataset) were gathered to verify the duality of the CSFs. The different types of datasets complemented each other’s weaknesses, allowing for a more generalized view of the empirical findings.