Information technology adoption behavior life cycle: Toward a Technology Continuance Theory (TCT)
Introduction
Over the years, information system (IS) usage has been a prominent topic in IS research. Prior efforts have sought to establish a theoretical base by explicating the determinants and mechanisms of users’ adoption decisions. It is widely believed that the adoption process influences successful use of information systems (Grover et al., 1998; Karahanna, Straub, & Chervany, 1999). Many scholars have investigated the factors that influence the diffusion of IS innovations in organizations (e.g., Gallivan, 2001, Rogers, 2003, Swanson and Ramiller, 2004; Zhu, Kraemer, & Xu, 2006). Others have proposed psychological models for explaining and predicting users’ behavior toward IS adoption at the individual level (e.g., Bhattacherjee, 2001, Bhattacherjee and Premkumar, 2004; Davis, Bagozzi, & Warshaw, 1989; Venkatesh and Davis, 2000, Venkatesh et al., 2003). These two streams of research suggest that the determinants and mechanisms for an individual's adoption decision may vary from stage to stage during the lifecycle of IS usage, i.e., at initial adoption and then subsequent stages of continued usage. Thus using the same or misdirected managerial tactics to facilitate adoption behavior across various stages may result in negative consequences and reduced IS effectiveness. Though different behavioral models (Karahanna, Straub, & Chervany, 1999; Jasperson, Carter, & Zmud, 2005) have been recognized as relevant to user adoption behavior at different stages, what is lacking is a clear comparison of these models in terms of their theoretical underpinning and application practices. Without a clear understanding of the differences in users’ adoption behavior over time, both scholars and practitioners will not be able to effectively manage the issues related to system design, individual cognition, and organizational actions.
The Technology Acceptance Model (TAM) (Davis, 1986, Davis et al., 1989) has dominated IS “use” research and has led to much exploration and widespread discussion over its application and extensions (e.g., Lai and Li, 2005, Shih, 2004b). In more recent years, the expectation confirmation model (ECM) (Bhattacherjee, 2001) was proposed to describe user's behavior in “continue to use” an information system. ECM was adapted from the consumer satisfaction/dissatisfaction model (CS/D) (Churchill and Suprenant, 1982, Oliver, 1981, Oliver and Burke, 1999, Tse and Wilton, 1988) that was originally designed in marketing research to model consumer's repurchase behavior. TAM with its focus on initial acceptance of an IS, theorizes that system use is directly determined by behavioral intention to use, and in turn motivated by the user's attitude toward system use. At the same time, ECM's objective is to evaluate an individual's continuance and loyalty for system use and argues that user satisfaction is the most important requirement determining a user's intention for continued use. While TAM has enjoyed widespread use and related literature has grown tremendously, there has been limited activity in ECM, post-adoption behavior and IS continuance research (Bhattacherjee, 2001, Bhattacherjee and Premkumar, 2004). Many studies have been conducted to verify TAM with diverse empirical data and in various application contexts (Shih, 2004a, Vijayasarathy, 2003, Yu et al., 2005), although the results have not always been consistent.
In comparing the theoretical underpinnings and application practices of TAM and ECM, three major differences can be found. First, while TAM has been applied to examine continuance and post-adoption behavior (Gefen, Karahanna, & Straub, 2003; Karahanna et al., 1999, Shih, 2004b, Taylor and Todd, 1995), its emphasis is on examining variables that lead to initial acceptance. On the contrary, ECM focuses on factors that influence retention and loyalty, as the long-term viability of an IS and success depends on continued use rather than first-time use alone (Bhattacherjee, 2001). Second, TAM proposes that the behavior toward system use can be determined by the user's attitude. However, ECM hypothesizes that IS continuance is primarily affected by user satisfaction. Many theorists believe that conceptually satisfaction and attitude are two distinct constructs (Oliver, 1980, Oliver, 1981, Tse and Wilton, 1988). Third, TAM considers two salient beliefs: perceived usefulness and perceived ease of use as underlying motivators affecting user's attitude and intention toward behavior. These behavioral beliefs are highly related to outcome expectations (Ajzen, 1991, Bandura, 1986, Davis et al., 1989). Thus TAM only adopts user expectations, usually measured in a single time period, to explain and predict behavioral intention. On the other hand, ECM is based on CS/D which posits that user satisfaction has a strong relationship with disconfirmation, which is a function of the difference between user expectations and perceived performance. According to CS/D, a user's expectation must be measured before system use whereas perceived performance is measured after the experience. However, CS/D ignores potential changes in outcome expectations across the accumulation of user's experience and the impact on user's psychological state and cognitive process. ECM replaces pre-consumption expectations with post-consumption expectations and postulates that satisfaction is a function of expectations and confirmation.
Although many theoretical differences exist between TAM and ECM, no empirical study, to our knowledge, has examined the influences of these differences on explaining and predicting users’ psychological states and behavior. In comparing them, many interesting issues are worthy of exploring. For example, which model is more powerful or more suitable? How do the predictive power of the models change from initial acceptance of IS to continued use? Is it appropriate to apply TAM to predict and explain user behavior toward technology continuance? Note that while TAM uses attitude and ECM uses satisfaction, an earlier landmark paper by Oliver (1980) used both attitude and satisfaction as antecedents to intention. In fact, in the Cognitive Model (COG) proposed by Oliver, satisfaction is postulated as an antecedent to post-exposure attitude. Is it possible to develop a hybrid model which combines attitude and satisfaction and relevant parts of TAM and ECM, and has higher explanatory power for describing user behavior toward technology continuance?
We address the above questions in this study. The study investigates differences in model descriptions, model-fit, and explanatory power of the three intention models: TAM, ECM, and COG. Subsequently, we develop an enhanced theoretical model, called Technology Continuance Theory (TCT) which integrates the three existing models, for representing and explaining user behavior toward technology continuance. Specifically, the objectives of the study are:
- 1.
Compare the three models: TAM, ECM, and COG. Test their hypotheses, and compare their path coefficients and explanatory powers.
- 2.
Compare the three models across various stages of IS continuance. The three stages included are: initial adopters, short-term users, and long-term users.
- 3.
Propose an enhanced Technology Continuance Theory (TCT) based on the characteristics of TAM, ECM and COG.
- 4.
Evaluate the model fit and explanatory power of TCT, and compare it with TAM, ECM, and COG.
Section snippets
Technology Acceptance Model
In the late 1980s, the TAM was developed for the IS discipline (Davis, 1986, Davis et al., 1989). It was based on the theory of reasoned action (Fishbein & Ajzen, 1975), an intention theory that has been widely accepted for several decades. TAM received wide attention from IS researchers for at least three reasons. First, it has a strong foundation in psychological theory (Chau, 1996, Taylor and Todd, 1995). Second, it is parsimonious and can be used as a guideline to develop a successful
Research models
We plan to carefully evaluate the three theoretical models described above: Davis’ TAM (Davis, 1986), Bhattacherjee's ECM (Bhattacherjee, 2001), and Oliver's COG (Oliver, 1980). This evaluation would lead to a comparison of the three models as well as identify relationships across the various constructs. Although these models have been investigated individually in various empirical studies, their comparison has not been reported in the literature. Especially for Oliver's COG model, we could not
Research design
In order to empirically test the three models and their hypotheses, a survey of three distinct user groups with varying levels of experience in using the Cyber University System (CUS) developed by a National University in southern Taiwan was conducted. The instrumentation, sampling methods, and scale validation process are described below.
Evaluation of research models and hypotheses testing
Structural equation modeling was used to evaluate the three structural models and various hypotheses. Visual tools provided by AMOS were used to depict the relationships among the constructs and their items. Since the respondents comprised three groups (initial adopters, short-term users, long-term users), the technique of multiple-group analysis with maximum likelihood estimation (MLE) was chosen to examine the moderating effect of usage experience and estimate the parameters of the model for
Discussion
In general, the three theoretical models TAM, ECM, and COG, and their hypotheses are all supported with relatively few exceptions. We make several important observations from the results.
The effect of attitude on intention in the TAM model is influenced by user experience, as the path coefficient increases across the three groups (initial adopter, short-term users, and long-term users). At the same time, the direct influence of perceived usefulness on intention is much less, and is further
Conclusions and limitations
The goal of this study was the development of an enhanced model for IS continuance suitable for the entire life cycle of adoption. In that pursuit, we analyzed three models: the TAM, the ECM, and the COG. The three models have different assumptions about the underlying constructs that dictate user behavior. Results indicate that the three have different explanatory powers with relative strengths and weaknesses. In general, in explanatory power, the Cognitive Model was superior to the other two,
References (54)
The theory of planned behavior
Organizational Behavior and Human Decision Process
(1991)- et al.
The influence of information technology diffusion and business process change on perceived productivity: The IS executive's perspective
Information & Management
(1998) - et al.
Technology acceptance model for Internet banking
Information & Management
(2005) An empirical study on predicting user acceptance of e-shopping on the Web
Information & Management
(2004)- et al.
The role of moderating factors in user technology acceptance
International Journal of Human-Computer Studies
(2006) - et al.
Extending the TAM for a t-commerce
Information & Management
(2005) - et al.
The antecedents and consequences of customer satisfaction for firms
Marketing Science
(1993) Social foundations of thought and action: A social cognitive theory
(1986)EQS structural equations program manual
(1989)Understanding information system continuance: An expectation-confirmation model
MIS Quarterly
(2001)
Understanding changes in belief and attitude toward information technology usage: A theoretical model and longitudinal test
MIS Quarterly
Structural equations with latent variables
A longitudinal analysis of the impact of service changes on customer attitudes
Journal of Marketing
An empirical assessment of a modified technology acceptance model
Journal of Management Information Systems
An investigation into the determinants of customer satisfaction
Journal of Marketing Research
User acceptance of computer technology: a comparison of two theoretical models
Management Science
A theory of cognitive dissonance
Belief, attitude, intention, and behavior: An introduction to theory and research
Evaluating structural equations with unobservable variables and measurement error
Journal of Marketing Research
Organizational adoption and assimilation of complex technological innovations: Development and application of a new framework
Data Base Advances
Inexperience and experience with online stores: the importance of TAM and trust
IEEE Transactions on Engineering Management
Multivariate data analysis
A structural equation model of end-user satisfaction with a computer-based medical information systems
Information Resources Management Journal
CS/D-overview and future research directions
A comprehensive conceptualization of post-adoption behaviors associated with information technology enabled work systems
MIS Quarterly
Information technology adoption across time: A cross-sectional comparison of pre-adoption and post-adoption beliefs
MIS Quarterly
Cited by (341)
Exploring the adoption patterns of matrimonial apps: An analysis of user gratifications
2024, Journal of Retailing and Consumer ServicesWhy do so many digital transformations fail? A bibliometric analysis and future research agenda
2024, Journal of Business ResearchThe state of financial inclusion research on developing countries
2023, Transnational Corporations ReviewWhat leads end-users to continue using social networking sites: Insights from Cote D'Ivoire
2023, Computers in Human Behavior ReportsThe impact of e-service quality on word of mouth: A higher education context
2023, International Journal of Management EducationConsumer post-adoption of e-wallet: An extended UTAUT2 perspective with trust
2023, Journal of Open Innovation: Technology, Market, and Complexity