Towards Identification of Potential Antecedents and Determinants of Continuance Intention towards M-pesa

A germinating ground within the IS literature is post-adoption, and it is concerned with individuals use of technology ensuing initial adoption (De Guinea & Markus 2009). IS post- adoption has been a focus area for theoretical development, and is now referenced by several labels, which include: IT usage (Straub, Limayem, & Karahanna-Evaristo, 1995; Agarwal &

Karahanna, 2000; Bhattacherjee & Premkumar, 2004; Kim & Malhotra, 2005; Burton-Jones and Gallivan, 2007), IS continuance (Bhattacherjee, 2001a; Cheung & Limayem, 2005; Kim et al, 2007; Limayem et al. 2007; Wu & Kuo, 2008; Larsen et al. 2009; Zhou, 2013), post-adoptive IT

28 Management Information Systems Quarterly is the leading scholarly journal of the IS discipline.

usage (Jasperson et al. 2005), IS Reuse (Kettinger, Park, & Smith, 2009), IS Extended Usage and Exploratory Usage (Saeed & Helm, 2008), and Customer Retention (Vatanasombut, Igbaria, Stylianou, & Rodgers, 2008), amongst others. That said, the expedition for antecedents and determinants of IS use remains a salient venture in theory and practice. On this note, Bhattacherjee (2001a) writes that IT continuance is considered a chain of decisions influencing continued use of a technology. To characterize the decision making process asserted by Bhattacherjee (2001a), De Guinea and Markus (2009) explain that the process is twofold: First is logically based on views of usefulness and ease of use, expectations based on familiarity, and other beliefs (Bhattacherjee, 2001a; Venkatesh, Brown, Maruping, & Bala, 2008). Second involves emotional responses to use of technology, for instance, satisfaction or mental assimilation (Agarwal & Karahanna, 2000; Sun & Zhang, 2006). Simultaneously, advances in the IS literature has taken a commendable leap by drawing on psychology literature, to put forward that continuance with IT could be habitual. The argument here is that in the presence of habitual use, intention loses its guiding force on use. Habitual use of IT is posited as repetitive behavioural series involuntarily prompted by environmental cues, rather than resolute behaviour (De Guinea

& Markus, 2009; Bhattacherjee & Lin, 2014). On habit, its influence has been received with mixed feelings, and its nomological path is still under debate. Some studies have reported that habit directly influences continued IT use (Kim & Malhotra, 2005), is mediated by continuance intention (Gefen, 2003), moderates the influence of continuance intention on continued use (Kim

& Malthotra, 2005; Limayem et al. 2007). While habit is likely to be a salient construct in continuance of M-pesa, for this study, the researcher excludes it selection, until more resolute reports on its causal path are documented. However, on continuance studies as a body, a categorization scheme by Larsen et al. (2009) characterizes extant studies into three somewhat overlapping categories. First category refers to studies that have applied IS adoption as a criterion variable for predicting IS continuance (e.g., Lin, Wu, & Tsai, 2005; Roca, Chiu, & Martinez, 2006). Second category refers to studies of procedures explicating the development of time- variant continued use (e.g., Kim & Malhotra, 2005; Cheung & Limayem, 2005; Chiu, Chiu, &

Chang, 2007). The third category refers to studies that combine the original IS continuance model with perceived harmonizing theoretical frameworks (e.g., Hsu & Chiu, 2004; Liao, Chen, & Yen, 2007; Chiu & Wang, 2008; Larsen, et al. 2009).

Mindful of this, the present study aligns with the third category because its objective is to model and test determinants of user continuance intention. As such, it seeks to adopt and test

complimentary factors from suitable models and theories in predicting user continuance towards M-pesa. Whilst this approach is commonly accepted in the IS community, leading authority- Anol Bhattacherjee and co-author Barfar (2011) have cautioned on the integration of inappropriate models (e.g., utilizing acceptance models- TAM & UTAUT in investigating continuance phenomenon), as identified in some extant studies (e.g., Chiu & Wang, 2008; Hong & Tam, 2006). An ill informed approach to the selection of factors for investigating continuance phenomenon stands to stir spurious correlations amongst unrelated factors and they propel a trajectory of misinformed views. In accordance, Bhattacherjee and Barfar (2011) opine that it is inapt to merge acceptance and continuance theories to predict continuance phenomenon, as these theories were posited to predict two divergent and time-variant separate behaviours (Bhattacherjee

& Barfar, 2011).

Further, in their paper on current state and future direction of continuance research, Bhattacherjee and Barfar (2011) observe the variety of theories utilized in the investigation of IT continuance.

These include TAM (Hong & Tam, 2006), UTAUT (Chiu & Wang, 2008), Theory of Habit (Limayem & Cheung, 2008), Fairness Theory (Chiu et al. 2007), Commitment-trust Theory (Vatanasombut et al. 2008), Innovation Diffusion Theory (Karahanna et al. 1999), amongst others. Following an outline of the aforementioned theories, Bhattacherjee and Barfar (2011) again stresses a distinction between IS adoption and continuance and argue that it is intuitive that factors that predict initial adoption may differ from continuance. To support this stance, the authors cite Karahanna et al’s. (1999) findings that perceived ease of use, a factor from the TAM, and a significant motivator of IT acceptance, has an attenuating influence over time, and thus an insignificant effect in IT continuance. Also, Bhattacherjee (2001a) demonstrated that unique factors emerge in the post acceptance phase of IT usage, such as a user’s confirmation and disconfirmation of expectations, and consequent satisfaction with the IT, which collectively influence IT continuance intention but are void on acceptance intention. Thus, theories and models designed to investigate IT acceptance (e.g TAM and UTAUT), may be inappropriate for explaining continuance.

That said, the next section presents the inception of Bhattacherjee’s (2001a) model, and subsequently identifies extant studies on continuance intention that have been published in reputable IS journals, and studies that combine Bhattacherjee’s (2001a) model with perceived complementary theoretical frameworks that are appropriate in a continuance era. The identified

studies will serve as a guide towards identifying potential antecedents of user continuance intention towards M-pesa.