Understanding consumer online shopping behaviour from the perspective of transaction costs

2.2.3 The Theory of Planned Behaviour (TPB) 
The development of TPB originated from the TRA (Ajzen and Fishbein 1980, Ajzen 1991, 
Ajzen 2011) and is designed to 
predict and explain human behaviour across various 
information technologies (Wu and Chen 2005, Wang and Ritchie 2012). According to TPB, a 
person’s actual behaviour in performing certain actions is directly influenced by his or her 
behavioural intention and, in turn, is jointly determined by his or her attitude, subjective 
norms (SN) and perceived behavioural controls (PBC) toward performing the behaviour. In 
essence, TPB differs from TRA in its addition of the component of PBC (Taylor and Todd 
1995c, Bagozzi and Kimmel 2011). PBC refers to the individual’s perception of ease or 
difficulty in performing the behaviour of interest (Ajzen 1991). It is believed that behaviour 
is strongly influenced by an individual’s confidence in his/her ability to perform a behaviour 
(Ajzen 1991). The more an individual believes that the resources and opportunities exist to 
perform the behaviour, the greater their PBC over the behaviour should be. SN refers to “the 

---

24
perceived social pressure to perform or not to perform the behaviour”. In other words, SN is 
related to the normative beliefs about expectation from other people (Wu and Chen 2005).
TPB has received good empirical support in a variety of application areas (Armitage and 
Conner 2001, Ajzen and Fishbein 2005, Sutton 2006). It has been applied to a variety of 
human behaviours, including adoption behaviour of internet banking (Lee 2009a, Yousafzai
 
et al.
2010, Nasri and Charfeddine 2012), online tax (Wu and Chen 2005), e-service (Chen 
and Li 2010, Lee 2010), e-learning (Lee 2010), e-procurement (Aboelmaged 2010), users’ 
acceptance of instant messaging (Lu
 et al.
2009), health-related services (e.g., diet, drinking, 
drug usage, smoking, weight loss, etc.) (Godin and Kok 1996, Hoie
 et al.
2012), tourists’ 
behavioural intention and actual behaviour of visiting the destination
(Quintal
 et al.
2010, 
Filho
 et al.
2012, Hsu and Huang 2012), environmental behaviour (Chao 2012), 
business 
start-up intentions and subsequent behavior (Kautonen
 et al.
2013), crisis planning intention 
(Wang and Ritchie 2012), 
intention to exercise (Spink
 et al.
2012), 
and so on.
Although current studies demonstrate that the TPB has great power in predicting and 
understanding consumers’ adoption behaviour across a variety of service contexts, it does not 
mean that TPB has no limitations. The main argument focuses on whether perceived 
behavioural control can be regarded as a good representative of actual behavioural control 
(Armitage and Conner 1999, Armitage and Conner 2001). In the literature, support for the 
PBC as an accurate proxy for actual control remains equivocal (Armitage and Conner 2001). 
In addition, the TPB is based on a specific behaviour, thus, each behaviour requires its own 
distinctive and specific belief set. Each behaviour in the TPB is explained by a salient belief 
set, so the application of TPB across a variety of situations may not be consistent (Hoie
 et al.
2012). Another limitation of TPB derives from the fact that this theory treats a set of beliefs 

---

25
(those influencing attitude, SN, or PBC) as a one-dimensional construct (Hoie
 et al.
2012), 
which makes it difficult for understanding the specific beliefs that affect user behaviour in the 
different technology adoption contexts (Taylor and Todd 1995c, Riemenschneider
 et al.
2003, 
Lin 2008).
In an attempt to address the potential limitations of TPB, scholars argue that extending TPB 
by incorporating the additional key constructs which are deemed important to the specific 
usage context can 
increase the variance of explanation of usage behaviour (Hsu and Huang 
2012). 
The major constructs, such as the achievement of personal goals (Perugini and 
Bagozzi 2001), self-identity processes (Shaw
 et al.
2000, Booth
 et al.
2014), descriptive 
norms (Hoie
 et al.
2012, Leyland
 et al.
2014), moral norms (Hoie
 et al.
2012, Newton
 et al.
2013), anticipated emotions (Ajzen and Sheikh 2013, Kim
 et al.
2013e), perceived risk and 
benefit (Lee 2009a), uncertainty (Quintal
 et al.
2010), past behaviours (Lam and Hsu 2006), 
user’ satisfaction (Baker and Crompton 2000, Liao
 et al.
2007), and technology readiness 
(Chen and Li 2010) were added to enhance the TPB’s predictive power. The extended TPB 
provided a more complete understanding of behaviour and behavioural intention.
In particular in the e-commerce setting, extant studies have applied TPB to online consumer 
behaviour (Bhattacherjee 2002, Choi and Geistfeld 2004, Hsu and Lu 2004, Ramus and 
Nielsen 2005, Wu 2006, Hansen 2008, Lee 2009a, Su and Huang 2010, Burns and Roberts 
2013). Researchers usually draw upon TPB to build a new research model by integrating 
other theories or constructs into it. For example, Hsu et al. (2006) extended TPB by 
incorporating constructs drawn from the expectation disconfirmation theory (EDT) and 
examined the antecedents of users’ intention to continue using online shopping. The results 
indicated that disconfirmation from EDT and satisfaction with prior online shopping exerted 

---

26
dominant influence on the continuance intention compared to the impacts of attitude, social 
norms, and PBC in the online shopping process. Limayem et al. (2000) introduced perceived 
innovativeness and perceived consequences, both as antecedents to attitude and intention into 
TPB. The results of their longitudinal study showed the positive effects of personal 
innovativeness and perceived consequences on attitude and intentions to online shopping. 
Behjati and Pandya (2012) extended TPB by including the effects of perceived reliability, 
trust and faithfulness on online purchasing intention. The findings showed that trust and 
faithfulness have significant relationships on online purchasing behaviour while perceived 
reliability has insignificantly relationship on online purchasing intention. 
Generally speaking, the extended TPB has improved the understanding of online consumer 
behaviour. Despite the growing body of knowledge of TPB in online environments, several 
issues exist in the literature in relation to the application of TPB in explaining online 
purchasing behaviour amd deserve attention from researchers. One obstacle in using TPB has 
been found in applying it to the research of continued online shopping behaviour. Recently, 
some researchers pointed out that a weakness of TPB is its lack of explanatory power of 
continued online shopping behaviour (Hsu
 et al.
2006). This is because TPB constructs do 
not fully reflect the context of user continuance decisions. Karahanna et al. (1999) also 
indicated that the beliefs users hold for continuance intention may not be the same set of 
beliefs which lead to initial adoption.
Additionally, among the existing studies, there is a relative paucity of knowledge about the 
roles that cost-related constructs reflected in time and effort expended have on the consumers’ 
decision-making (Mukherjee
 et al.
2012, Kim
 et al.
2013b, Wu
 et al.
2014). Darley et al. 
(2010) and Kim et al. (2014) further assert that the research on online consumer behaviour 

---

27
needs a more comprehensive model, describing not only the effect of personal motivation 
beliefs, but also the impacts of TCs incurred during the online transaction process. Indeed, 
there is growing recognition that cost reduction appears central to the business models 
pursued by firms (Williamson and Ghani 2012). However, little is still known about the 
specific costs associated with online shopping and how they determine consumers’ online 
shopping behaviour (Wu
 et al.
2014). As such, understanding how the costs involved in 
online transaction-related activities together with the key constructs in TPB affect consumers’ 
online decision making is an important research topic that needs further theoretical and 
empirical attention.