An Empirical Study of Malaysian Consumers’ Channel-switching Intention: Using theory of Planned Behaviour

Introduction

Internet World Stat stated that the number of Internet users in Malaysia was 17,723,000 (60.7 per cent) in 2012. Meanwhile, the increase of Internet users from 2000 to 2012 was 376.8 per cent (http://www.internetworldstat.com). Shopping via Internet is a common occurrence in the Western countries, but for Malaysia, it is very challenging and majority of Malaysians prefer traditional shopping. In Malaysia, online shopping is something new and the transactions are very limited. Syed, Ismail and Ahsan (2008) noted that the growing utilization of Internet by the younger generation in Malaysia offers an emerging opportunity for online retailers. If online retailers know the factors influencing young Malaysian consumers’ shopping behaviour, and the relationship between these factors and the type of online shoppers, they can further develop their marketing strategies to convert potential customers into active ones. Thus, with a large number of Internet users in Malaysia, it is necessary to find out the factors that shape Malaysian online shopping behaviours and to develop more studies in this area. Therefore, Internet as a new retailing channel in Malaysia plays an important role in the channel switching and multichannel environment, and this needs further consideration from retailers. Likewise, marketers and retailers need to find out what is important for customers in selecting/switching a channel/ multichannel in order to settle on a suitable channel strategy and to manage resource allocation. From a marketing point of view, synchronizing multiple channels deeply enhances the difficulty of a marketing strategy and it would have an effect on consumers’ multichannel behaviour.

Channel switching and multichannel retailing strategies offer some advantages to the retailers and companies with two or more channels to send their services or/and products to the consumers. It would also be more interesting and valuable for customers to choose from more than one channel to seek for information and purchase their products. Consumers choose and use channels based on their specific goals at a particular point in time, their prior experiences and their expertise (Wesley & Eisenstein, 2008). A number of multichannel studies suggest that channel choice is not static but changes over time, as consumers migrate from one channel to another (Albesa, 2007; Black, Lockett, Ennew, Winklhofer & McKechnie, 2002; Dholakia, Zhao & Dholakia, 2005; Gupta, Su & Walter, 2004). Besides, consumers may switch from one channel to another in order to find ways to maximize the benefits of shopping and minimize the costs associated with shopping, in terms of money, time and energy, whether in a brick-and-mortar store, or over the Internet (Downs, 1961; Kim & Kang, 1997). In addition, customers would prefer to migrate from online channels to offline channels and vice versa according to their behavioural factors (e.g., attitude, beliefs, as well as perceived behavioural control [PBC]) of a channel. Therefore, it is important for retailers to not only get knowledge about the benefits and costs of the channels, but also they need to know more about the consumers.

Moreover, consumers may search for product information online but prefer to buy it in a traditional store (Albesa, 2007; Balasubramanian, Raghunathan & Mahajan, 2005; Van, Minocha & Laing, 2007). They may find it easy to make a search about product information through the online channel of A-retailer but do their purchase from B-retailer through the offline channel (Kim & Lee, 2008; Shim, Eastlick, Lotz & Warrington, 2001; Verhoef, Neslin & Vroomen, 2005). This would result in a sort of difficulty for the retailer to retain customers due to the switching situation (Jensen, Jakus, English & Menard, 2004; Kumar & Venkatesan, 2005; Stone, Hobbs & Khaleeli, 2002). Thus, the concept of buying and switching intention concerning up-to-date consumers has become a major issue to marketers and retailers (Albesa, 2007; Kumar, Shah & Venkatesan, 2006; Noble, Griffith & Weinberger, 2005; Pookulangara, Hawley & Xiao, 2011). Information technology has made people use the Internet in conducting their transactions rather than the traditional way of trading (Fuller, Serva & Benamati, 2007). Furthermore, it is possible that multiple channels’ retailers meet the desires of customers’ flexibility for how, where, when and what to shop (Kumar & Venkatesan, 2005; Stone et al., 2002). The challenge is to recognize when, how and where consumers utilize the stores and/or Internet and how consumers consider migrating among channels and among retailers (Albesa, 2007; Bickle, Buccine, Makela & Mallette, 2006; Kumar & Venkatesan 2005). The purpose of this study is to examine channel-switching intention between Malaysian consumers in the Internet and brick-and-mortar stores channels by applying the theory of planned behaviour (TPB). Theory of planned behaviour has been used in studies that have examined multichannel consumer behaviour (Grabner-Krauter & Kaluscha, 2003; Kamarulzaman, 2008; Keen, Wetzels, Ruyter & Feinberg, 2004; Wen, Prybutok & XU, 2011) as well as channel-switching behaviour (Kim & Park, 2005; Pookulangara & Natesan, 2010; Pookulangara et al., 2011). Thus, using TPB is an appropriate and acceptable theory for studying consumers’ channel-switching intention.

Literature Review

There is an opportunity for consumers to choose only one organization to seek for information (Dholakia & Uusitalo, 2000; Lihra & Graf, 2007), buy goods as well as return unwanted products by selecting one of the following channels, for example, Internet and brick-and-mortar stores (Kumar & Venkatesan, 2005). For that reason, it is very significant for multiple channel retailers to understand and perceive how their consumers behave and to know the issues, in particular, the purchase incentive that is under this behaviour. In this way, it is more possible for retailers to classify different groups of consumers and it will diminish the losses related to non-selective coverage (Schroder & Zaharia, 2008).

Multichannel consumers are defined as those who regularly shop through more than a single channel (Kumar & Venkatesan, 2005). Multichannel shoppers are important for retailers’ success because they are more open to new experiences, purchase more frequently and are more loyal to a brand (Pastore, 2001). Several researchers have conducted studies related to consumers’ online shopping behaviours in a multichannel shopping context (Choi & Park, 2006; Kim & Park, 2005; Kim, Kim & Kumar, 2003; Noh, 2008). They have found that multichannel consumers’ behaviours provide a number of chances to increase the sales and profits of multichannel retailers. Therefore, understanding multichannel consumer behaviours is crucial for multichannel retailers’ success (Dholakia, Zhao & Dholakia, 2005; Mathwick, Malhotra & Rigdon, 2002).

A consumer may choose one channel but at the middle stage of his/her decision making process may change the channel and use another one. Therefore, a consumer might migrate to another channel if he/she is not happy with a current channel, which results in channel switching (Albesa, 2007; Balasubramanian et al., 2005). An important issue of interest to both practitioners and academicians is to find out consumers’ channel-switching behaviour (from online to offline and vice versa) and also to recognize the major drivers which influence such behaviour (Gupta et al., 2004). For example, a client may tend to switch to online channels in case his/her intention towards online purchase is better than offline buying intention and vice versa. Necessarily, consumers declare their preferences according to utility maximization when it comes to costs and benefits of the retail structures given to them (Alba et al., 1997). It indicates that the utility gained by the clients through online shopping needs to be more than the utility obtained through the traditional format, which causes the consumer to switch to an online environment. This study recognizes the main behavioural determinants that affect consumers’ switching tendency from shopping offline to the online one and vice versa.

Nowadays, the multichannel strategy is becoming more significant and crucial for both businesses and consumers (Albesa, 2007; Baal & Dach, 2005; Dholakia, Zhao & Dholakia, 2005; Rangaswamy & Bruggen, 2005). Retailers and companies should search for a multiple channels’ design that offer ‘channel advantages’, due to the fact that each channel provides some degree of difference of benefits (e.g., mutual communication between consumers and retailers, customer satisfaction, longer tenure and consumers are more loyal to channel) (Black et al. 2002; Kumar & Venkatesan, 2005), but at the same time, it also has complications and challenges (Van, Minocha & Laing, 2007). One of the challenges is that retailers might lose their consumers in the path of the buying process (Inman, Venkatesh & Rosellina, 2004; Nicholson, Clarke & Blakemore, 2002; Nunes & Cespedes, 2003). This is related to the issue that consumers could easily move through channels. For example, they can seek information through one channel (e.g., Internet), and after that buy the product across another channel (e.g., store) (Pookulangara et al., 2011). For this reason, employing only one channel limits the performance in the marketplace to what that channel is proficient in doing predominantly well (Johnson & Greco, 2003).

In spite of the growing attention which has been paid to multichannel-oriented topics, study on multichannel retailing and channel-switching behaviour is still considered to be at its early stages. As Neslin et al. (2006) and Slack, Rowley and Coles (2008) stated, previous research works on multichannel-based topics chiefly concentrated on attaining knowledge regarding the elements of customers’ choice of channel. Very few studies have investigated customer channel migration in terms of multi-channel retailing and the factors that affect consumer’s channel behaviour among different channels in a multichannel environment (Ansari, Mela & Neslin, 2008). A study done by Choi and Park (2006) has shown that there is a lack of knowledge concerning important predictors in terms of consumers’ beliefs, attitudes and intentions for online as well as traditional stores shopping on the basis of multiple channels and channel switching. According to Pookulangara et al. (2011), the study on consumer channel-switching intention is still not sufficient and needs more research. Hence, the present study investigates potential elements related to customer channel migration behaviour.

In addition, to convince and attract Malaysian consumers to use the Internet, as their retailing channel instead of traditional channels, is still a challenging task for web retailers in Malaysia (Haque & Khatibi, 2005; Salehi et al., 2011). The development of Internet technology in Malaysia has massive opportunities such as its increasing benefits, decreasing costs of product and service delivery and expanding geographical boundaries in bringing buyers and sellers together (Haque & Khatibi, 2005; Syed et al., 2008). Salehi et al. (2011) reported that the acceptance of the Internet channel among Malaysian consumers is not as advanced compared to their counterparts in other countries. The question is why? In addition, what are the factors that influence their acceptance? The other challenge is to understand how and when Malaysian consumers use the Internet, and what drives their propensity to switch between retailers and between channels. There are some barriers, which have contributed to the unwillingness of the Malaysian people to shop online, like being afraid of their personal information being stolen by others (Salehi, 2012). Despite the potential among Malaysian consumers, there is still a lack of understanding towards online shopping. Meanwhile, Haque and Khatibi (2005) and Mumtaz, Islam, Ariffan and Karim (2011) stated that studies regarding consumer behaviour by using of behavioural models (e.g., theory of reasoned action [TRA] and TPB) towards online shopping in the Malaysian environment are still limited and also they claimed that Internet in Malaysia is still considered as a new medium between the retailers and the consumers.

Theory of planned behaviour is a well-known theory to evaluate and examine consumers’ behaviour and channel-switching behaviour in online and/or offline channels. The effect of components of TPB on channel choice/switching is different (Pookulangara et al., 2011). Some factors in some channels may significantly affect consumers’ channel-switching behaviour, but in contrast, some other channels may have no influence on consumers’ channel-switching behaviour. Meanwhile, the purpose of this article is to study on Malaysian consumers’ channel-switching intention. Thus, Haque and Khatibi (2005), Harn, Khatibi and Ismail (2006) and Mumtaz et al. (2011) stated that studies regarding consumer behaviour by using behavioural models (e.g., TRA and TPB) for online shopping in the Malaysian environment are still limited and also they claimed that Internet in Malaysia is still considered as a new medium between the retailers and the consumers. Therefore, the objective of this study is to identify factors that are significant in explaining Malaysian consumers’ channel-switching behaviour. This study is based on the TPB, a theory well grounded in consumer behaviour research, and we used Malaysia as the sampling frame.

Decomposing Attitude (relative advantage, compatibility and complexity)

Rogers (1995) argued that the attitudinal belief has three innovation characteristics that influence behavioural intentions, which consist of relative advantage, complexity and compatibility. Relative advantage is defined as an innovation factor that significantly affects attitude (Taylor & Todd, 1995). This component presents some benefits to the model, such as image, satisfaction, economic benefits improvement and convenience (Rogers, 1983). Relative advantages should be positively connected to attitude (Taylor & Todd, 1995). The link between perceived relative advantage and attitude has been supported by numerous studies related to IT usage. Morris and Dillon’s (1997) study on Netscape usage among students revealed that the attitude towards using the browser was positively influenced by perceived relative advantage (perceived usefulness). George and Kumar (2013) indicated that perceived usefulness positively impact on customer satisfaction in Internet banking. A survey by Chau and Hu (2001) investigating physicians’ acceptance of telemedicine technology also indicated the significant influence of perceived relative advantage on attitude. In an online survey exploring Internet users’ intention to shop online, Chen, Gillenson and Sherrell (2002) found that higher perceived relative advantage (perceived usefulness) resulted in a more favourable attitude towards online shopping.

OPEN IN VIEWER

Figure 1.

Source: Adapted from Taylor and Todd (1995), Pookulangara et al. (2011) and Elliott and Ainsworth (2012).

Compatibility is the degree to which the innovation fits with the positional adapter’s existing values, previous experiences and current needs (Rogers, 1983). Compatibility is likely to be positively connected to adoption. Tan and Teo (2000), in their study on adoption of Internet banking in Singapore, suggested that users who perceive Internet banking as compatible to their values developed positive attitude about the technology and were more likely to adopt it. In an exploratory study comparing adopter and non-adopter beliefs about Internet banking, Gerrard and Cunningham (2003) suggested the need for banks to highlight positive characteristics of the technology such as compatibility to encourage its usage. They found that the adopters of Internet banking felt the technology was more compatible to their lifestyle. Finally, in a study involving 82 brokerage firms, Lau (2002) found that compatibility had a significant effect on attitude of using online trading.

Complexity signifies the degree to which an innovation is perceived to be complicated to understand, learn or operate (Rogers, 1983). In general, the simpler an innovation is to realize and utilize, the more likely that it will be adopted. Complexity (and its corollary, ease of use) has been found to be a significant factor in a technology adoption decision (Davis, 1989; Moore & Benbasat, 1991; Taylor & Todd, 1995). It should expect channel that is easy to use would encourage individuals to utilize the channel by developing a positive attitude about it. Previous studies have indicated that if technology is complicated and complex to use, the impact of complexity on attitude will be negative (Beiginia, Besheli, Ahmadi & Soluklu, 2011; Moore & Benbasat, 1991). In addition, the effect of perceived ease of use on attitude has been theorized and validated by numerous studies. In a cross-sectional study on potential adopters and users of Microsoft’s Windows 3.1, Karahanna, Straub and Chervany (1999) found that perceived ease of use had a significant influence on attitude towards adopting the software among the potential adopters. Lau’s (2002) study on online trading revealed the significant influence of perceived ease of use on attitude. A similar result was also revealed in the study by other researchers (e.g., Bhattacherjee, 2000; Taylor & Todd, 1995).

Moreover, Haque and Khatibi (2005) and Mumtaz et al. (2011) demonstrated that online shopping was still not compatible with Malaysian consumers. In fact, a majority of Malaysians use the Internet for non-shopping purposes. In line with the argument and findings discussed, brick-and-mortar store is the channel which is compatible with a majority of Malaysian consumers. In addition, it would be expected that Malaysians who perceive relative advantage and compatibility of the Internet channel would more likely have a positive attitude towards using the technology. Taylor and Todd (1995) illustrated that according to prior empirical study on the correlations among these perceived characteristics (Rogers, 1983; Tornatzky & Klein, 1982), it would be likely that relative advantage and compatibility are positively associated to attitude. On the other hand, Beiginia et al. (2011) found out that relative advantage and complexity have a positive effect on attitude (Figure 1).

Method

The survey method in the form of Likert scale questionnaire has been widely used in marketing research (Kim et al., 2003; Sadeghi & Hanzaee, 2010; Sierra & McQuitty, 2005; Yoon, Lee & Lee, 2010; Yousafzai, Pallister, & Foxall, 2005; Yuksel, Yuksel & Bilim, 2010). This study adopted/adapted the questions from Pookulangara et al. (2011), Pookulangara and Natesan (2010), Ajzen and Fishbein (1980), George (2004), Beiginia et al. (2011), Taylor and Todd (1995), Yang, Park and Park (2007), Verhoef et al. (2005), Sproles and Kendall (1986), Mokhlis and Salleh (2009), Noble, Griffith and Weinberger (2005), Smith et al. (2008), Elliott and Ainsworth (2012) and Ajzen (2002). According to the aim of the current study, questionnaire was distributed among Malaysian consumers. To assess and estimate the intention of consumers, components’ scales have been followed by the recommendations of Ajzen and Madden (1986).

To assess relative advantage, compatibility and complexity, items were adapted from Taylor and Todd (1995) and Beiginia et al. (2011). Ten items are developed to examine the impact of relative advantage, compatibility and complexity on attitude. Six items to assess NB (friends, family and co-workers) are adapted from Pookulangara et al. (2011), Ajzen and Fishbein (1985), Pookulangara and Natesan (2010) and George (2004). To assess the effect of self-efficacy and facilitating condition (i.e., price and information) on PBC, ten items are adapted from Chiu et al. (2011), Pookulangara et al. (2011), Eastlick and Feinberg (1999), Noble, Griffith & Weinberger (2005) and Dickerson and Gentry (1983).

Attitude has been adapted based on the instrument applied and validated by Pookulangara et al. (2011) and Yang et al. (2007). According to the report of Pookulangara et al. (2011), behavioural attitude for brick-and-mortar stores and the Internet were conceptualized and was calculated by a four-item scale. Items to evaluate SN were adapted from the instrument used by Verhoef et al. (2005) and Pookulangara et al. (2011). Subjective norm was conceptualized brick-and-mortar stores and Internet and was measured via a five-item scale (Pookulangara et al., 2011). Three items were adapted to assess PBC from the instrument used and validated by Pookulangara et al. (2011).

Data Analysis

The purpose of this study is to investigate the influence of behavioural attitude and SN and PBC on channel-switching intention. Theory of planned behaviour is used in the current research. The partial least square (PLS) based on structural equation modelling (SEM) technique was used to test research hypotheses as well as research model as suggested by other researchers who have studied based on the behavioural models such as TPB and TRA models (Blue, Wilbur & Marston-Scott, 2001; Bock, Lee, Zmud & Kim, 2005; Chang, 1998; Lin & Lee, 2004; Millar & Shevlin, 2003; Ryu, Ho & Han, 2003). The SEM is a technique that seeks to represent the observed data in terms of a number of structural parameters defined by a hypothesized underlying model (Hair, Ringle & Sarstedt, 2011). Structural equation modelling is a theory-based approach that has the ability to bring data and theory together. Structural equation modelling is used in business and marketing studies to empirically test the complex models. This study also used the SEM approach to test the research hypotheses and data analysis. Unlike other statistical methods, SEM tests the model paths and model fit. Structural equation modelling also allows the assessment of complex interrelated dependence relationships and incorporates the effects of measurement error on the structural coefficients (Hair, Anderson, Tatham & Black, 1998). In SEM, two types of variables are used: exogenous variables and endogenous variables. Therefore, since in this study there are some exogenous variables (i.e., relative advantage, compatibility, complexity, NB, facilitating conditions, self-efficacy, attitude, SN and PBC) and their effect on the endogenous variable (i.e., channel-switching intention) has to be examined, the use of SEM is necessary. Pilot testing was conducted to clarify the terms used in the questionnaire before operating the main survey.

The survey instrument was pre-tested for content validity and adjustments were made prior to main data collection. The survey instrument was pre-tested with consumers (N = 30). It was assumed that these consumers had used at least one channel (i.e., brick-and-mortar store and/or the Internet) in the last 6 months. These consumers were comprised of professors and senior lecturers at the University of Malaysia (UM) and Universiti Putra Malaysia (UPM). Based on feedback from the pilot study group, minor adjustments were made to the instrument scale. This feedback was implemented into the instrument and content validity claim was established accordingly. Items were revised to ensure readability and for logical flow of questions. The results of the data analyses were organized into the following sections: (i) respondents profile, (ii) measurement model and (iii) structural model.

Respondents Profile

With reference to Table 1, a remarkable percentage of the respondents (57.9 per cent) are less than 34 years old. In addition, majority of the respondents are females. There are 283 female respondents (56.9 per cent) and 214 male respondents, which contributes to 43.1 per cent of the total respondents participated in this study. Of these respondents, 51.1 per cent of the respondents are Malay, followed by 28.6 per cent Chinese and 20.3 per cent Indian. Table 1 further indicates that 57.7 per cent of the practitioner possess bachelor’s degree, 33.8 per cent possess master’s degree and 8.5 per cent possess doctoral degree.

OPEN IN VIEWER

Table 1.

Respondents Profile

Variable	Frequency (N = 497)	Percentage
Age
18–24	172	34.6
25–34	116	23.3
35–44	74	14.9
45–54	68	13.7
55–64	48	9.7
65+	19	3.8
Gender
Male	214	43.1
Female	283	56.9
Race
Malay	254	51.1
Chinese	142	28.6
Indian	101	20.3
Education Level
Bachelor’s Degree	287	57.7
Master’s Degree	168	33.8
Doctoral Degree	42	8.5

Source: Researcher’s own construction.

Other External Variables

Table 2 shows descriptive statistics of the Internet usage and brick-and-mortar stores for searching for information as well as purchasing in the last 1 year. Findings reveal that 65.8 per cent of the respondents had searched for information online and 89.3 per cent preferred to search for information in brick-and-mortar stores in the last 1 year. Additionally, 39 per cent of respondents had purchased online and 98.2 per cent had purchased from brick-and-mortar stores in the last 1 year.

OPEN IN VIEWER

Table 2.

Descriptive Statistics of the Respondents

Variable	Channel	Frequency	Percentage
Searched for Information in the Last 1 Year	Internet	327	65.8
	Store	444	89.3
Purchased Products/Service in the Last 1 Year	Internet	194	39
	Store	488	98.2

Source: Researcher’s own construction.

Internet

All measurement variables with loadings under 0.70 were removed. This included the removal of the first item of complexity as well as two items of NB. After excluding these items, factors were computed again. In addition, Cronbach’s alphas were well above the acceptable level ranging from 0.71 to 0.94 for relative advantage, compatibility and complexity, NB with 0.95, information with 0.95, price with 0.95, self-efficacy with 0.84, attitude with 0.93, SN with 0.96, PBC with 0.92 and channel-switching intention with 0.94. In addition, CR was 0.97, 0.98, 0.87, 0.96, 0.96, 0.97, 0.91, 0.96, 0.97, 0.95 and 0.96 for relative advantage, compatibility, complexity, NB, information, price, self-efficacy, attitude, SN, PBC and intention, respectively (Table 3). Therefore, based on Cronbach’s alpha and CR, all these latent variables regarding Internet channel had reliability higher than 0.7. In addition, AVE was 0.88, 0.94, 0.78, 0.86, 0.87, 0.92, 0.76, 0.87, 0.87, 0.86 and 0.89 for relative advantage, compatibility, complexity, NB, information, price, self-efficacy, attitude, SN, PBC and channel-switching intention, respectively. These measurements were well above the 0.50 recommended levels (Fornell & Larcker, 1981). These results indicated that the constructs associated with outer measurement models exhibited satisfactory convergent validity.

OPEN IN VIEWER

Table 3.

Factor Analysis and Reliability

Measurement Variables	Internet				Brick-and-mortar Store
	Factor Loading	α	Composite Reliability	Average Variance Extracted	Factor Loading	α	Composite Reliability	Average Variance Extracted
Relative Advantage (RA)		0.95	0.97	0.88		0.94	0.96	0.86
• RA1	0.944				0.878
• RA2	0.946				0.939
• RA3	0.929				0.948
• RA4	0.933				0.934
Compatibility (Compt)		0.97	0.98	0.94		0.97	0.98	0.94
• Compt1	0.969				0.966
• Compt2	0.973				0.980
• Compt3	0.966				0.963
Complexity (Complx)		0.71	0.87	0.78		0.79	0.87	0.77
• Complx1	0.438*				0.993
• Complx2	0.893				0.331*
• Complx3	0.868				0.742
Normative Belief (NB)		0.95	0.96	0.86		0.97	0.98	0.89
• NB1	0.907				0.928
• NB2	0.952				0.943
• NB3	0.953				0.959
• NB4	0.905				0.955
• NB5	0.509*				0.948
• NB6	0.485*				0.918
Information (Info)		0.95	0.96	0.87		0.90	0.95	0.91
• Info1	0.911				0.647*
• Info2	0.938				0.948
• Info3	0.949				0.962
• Info4	0.933				0.640*
Price		0.95	0.97	0.92		0.94	0.96	0.89
• Price1	0.959				0.912
• Price2	0.954				0.958
• Price3	0.958				0.952
Self-efficacy (S.Eff)		0.84	0.91	0.76		0.92	0.95	0.87
• S.Eff1	0.841				0.913
• S.Eff2	0.912				0.950
• S.Eff3	0.866				0.933
Attitude		0.93	0.96	0.88		0.97	0.98	0.94
• Attitude 1	0.944				0.966
• Attitude2	0.962				0.981
• Attitude3	0.914				0.967
Subjective Norms (SN)		0.96	0.97	0.87		0.96	0.97	0.93
• SN1	0.916				0.962
• SN2	0.946				0.977
• SN3	0.951				0.578*
• SN4	0.935				0.947
• SN5	0.918				0.521*
Perceived Behavioural Control (PBC)		0.92	0.95	0.86		0.89	0.93	0.82
• PBC1	0.922				0.912
• PBC2	0.941				0.958
• PBC3	0.915				0.952
Channel-switching Intention (CSI)		0.94	0.96	0.89		0.90	0.93	0.83
• CSI1	0.942				0.923
• CSI2	0.957				0.915
• CSI3	0.934				0.890

Source: Researcher’s own construction.

Note: * Factor loadings lower than 0.7 were extracted.

Structural Model

PLS can evaluate theoretical hypotheses as well as indicate the existence of relationships for further testing (Chin, Marcolin & Newsted, 2003). Partial least squares can be used in estimating latent structural models that are indirectly observed by multiple indicators for theory testing and development as well as offering predictive applications (Anderson & Gerbing, 1998). The focus of the assessments of structural paths in PLS is on the inner model and the significance of the paths can be measured by bootstrapping critical ratios.

In the structured model of this study, all constructs had reflective items, as depicted in Figures 2 and 3. The significance of the reflective outer-measurement model via bootstrapped t-values of item loadings was assessed. The bootstrapping method of sampling was used to estimate the precision of the reflective outer-measurement models. Bootstrap t-values were computed on the basis of 500 bootstrapping runs. The model parameters as depicted in Figures 2 and 3 were estimated using the SmartPLS with the focus here on the inner results as they relate directly to the hypotheses. Thus, an examination for each exogenous and endogenous constructs of the model was undertaken via path weight coefficients, standard error, R² and bootstrap critical ratios (t-values).

The primary evaluation criteria for the structural model are the R² measures and the level and significance of the path coefficients (Lohmöller, 1989). Because the goal of the prediction-oriented PLS-SEM approach is to explain the endogenous latent variables’ variance, the key target constructs’ level of R² should be high (Lohmöller, 1989). The judgement of what R² level is high depends, however, on the specific research discipline (Hair et al., 2011), whereas R² results of 0.20 are considered high in disciplines such as consumer behaviour (Hair et al., 2011).

With the collected data from the survey, consumer channel-switching behaviour is predicted in regards to the Internet and brick-and-mortar stores channels. All the dimensions are included in the final data analysis, except for the three dimensions for Internet channel (first item of complexity and items five and six of NB). Also five dimensions for brick-and-mortar stores (second item of complexity, first and last items of information and third and fifth items of SN) were removed because of factor loadings less than 0.7. After these low items were extracted, factors were analyzed again and sufficient supports of reliability and validity of the measurement scales were achieved. As the measurement assessment supported the validity and reliability of measured items, a series of hypothesis tests proposed in the model are followed using SmartPLS techniques.

OPEN IN VIEWER

Figure 2.

Source: Researcher’s own construction.

OPEN IN VIEWER

Figure 3.

Source: Researcher’s own construction.

Hypothesis Testing: Internet

Hypotheses are tested in the following discussion for consumer channel-switching intention from the Internet to a brick-and-mortar store. Relative advantage, compatibility and complexity were the exogenous (independent) latent constructs that were utilized to predict attitude as endogenous (dependent) latent construct towards channel-switching intention in the Internet channel. The results in Table 4 and Figure 2 indicate that compatibility and complexity have a positive and strong relationship with attitude (β = 0.37 and 0.36 and t-values = 4.90 and 7.77, respectively) (p < 0.01) and support H2a and H3a. Somewhat unexpectedly, relative advantage does not influence attitude because path coefficient and t-value are not significant; therefore, the findings do not support H1a (relative advantage on attitude), but H2a and H3a (compatibility and complexity) will significantly predict attitude towards switching channel from the Internet to brick-and-mortar stores. The results show that NB with t-value = 16.12 and β = 0.59 (p < 0.01) strongly and significantly influence SN. Thus, the results support H4a, which declared that NB positively influenced SN towards channel switching from the Internet to brick-and-mortar stores.

Information, price and self-efficacy were the exogenous constructs that were used to predict PBC. Findings indicated that self-efficacy with t-value of 7.79 and β = 0.36 (p < 0.01), information with t-value of 2.51 and β = 0.19 (p < 0.05) and price with t-value of 1.71 and β = 0.11 (p < 0.1) positively and significantly affected PBC. Self-efficacy influenced PBC more significantly. Thus, the results revealed that H5a (information), H6a (price) and H7a (self-efficacy) positively affected PBC towards channel-switching intention from the Internet to brick-and-mortar stores.

Attitude, SN and PBC were the exogenous constructs for the endogenous construct channel-switching intention. The results show that attitude (t-value = 4.43 and β = 0.22) and SN (t-value = 6.35 and β = 0.41) positively and significantly affected consumers’ channel-switching intention. The path coefficient between these three variables and channel-switching intention was significant at 0.01. However, PBC (t-value = 0.68 and β = 0.03) did not predict consumers’ channel-switching intention. Therefore, H8a (attitude) and H9a (SN) positively affected channel-switching intention from the Internet to brick-and-mortar stores (Table 4 and Figure 2).

OPEN IN VIEWER

Table 4.

Results of Hypotheses Testing: Internet

Predicted Variables	Predictor Variables	Hypothesis	Path Weight	Standard Error	t-values
Attitude	Relative Advantage	H1a	0.09	0.06	1.40	Not Supported
	Compatibility	H2a	0.37***	0.08	4.90	Supported
	Complexity	H3a	0.36***	0.05	7.77	Supported
Subjective Norm	Normative Belief	H4a	0.59***	0.04	16.12	Supported
Perceived Behavioural Control	Information	H5a	0.19**	0.07	2.51	Supported
	Price	H6a	0.11*	0.07	1.71	Supported
	Self-efficacy	H7a	0.36***	0.05	7.79	Supported
Channel-switching Intention	Attitude	H8a	0.22***	0.05	4.43	Supported
Channel-switching Intention	Subjective Norm	H9a	0.41***	0.06	6.35	Supported
Channel-switching Intention	Perceived Behavioural Control	H10a	0.03	0.04	0.68	Not Supported

Source: Researcher’s own construction.

Notes: * Significant at p < 0.1; ** significant at p < 0.05; *** significant at p < 0.01.

OPEN IN VIEWER

Table 5.

Results of Hypotheses Testing: Brick-and-mortar Stores

Predicted Variables	Predictor Variables	Hypothesis	Path Weight	Standard Error	t-Values
Attitude	Relative Advantage	H1b	0.44***	0.08	5.70	Supported
	Compatibility	H2b	0.36***	0.08	4.45	Supported
	Complexity	H3b	0.03	0.02	1.37	Not Supported
Subjective Norm	Normative Belief	H4b	0.80***	0.02	39.48	Supported
Perceived Behavioural Control	Information	H5b	0.30***	0.06	5.14	Supported
	Price	H6b	0.30***	0.05	5.57	Supported
	Self-efficacy	H7b	0.22***	0.04	5.20	Supported
Channel-switching Intention	Attitude	H8b	0.33***	0.07	4.57	Supported
Channel-switching Intention	Subjective Norm	H9b	0.29***	0.06	4.71	Supported
Channel-switching Intention	Perceived Behavioural Control	H10b	0.01	0.04	0.08	Not Supported

Source: Researcher’s own construction.

Notes: * Significant at p < 0.1; ** significant at p < 0.05; *** significant at p < 0.01.

Discussion and Conclusion

In addition, the present study provides evidence of consumers’ channel-switching intention focusing on the effect of attitude, SN and PBC on switching intention. Based on the data analyses and findings, it can be notified that exogenous variables in both Internet and brick-and-mortar stores channels have differences as well as similarities while predicting channel-switching intention. In the current study, relative advantage, compatibility and complexity are three exogenous constructs which differently predicted attitude towards channel-switching intention in both channels (Internet and brick-and-mortar stores). The significant effect of relative advantage on attitude towards channel-switching intention from brick-and-mortar stores to Internet is not surprising given the fact that the extrinsic benefits of using Internet channels are numerous for those consumers who prefer online shopping. Some of the benefits are faster and convenient execution of online transactions, lower economic cost (reduced commuting and time saving), convenient online access to product information, etc. The results of this study imply that individuals form positive attitude towards channel-switching intention from stores to use Internet because of these benefits. The significant effect of attitude on intention found in this study and also in other studies (Ajzen & Fishbein, 1985; Pookulangara et al., 2011; Shim et al., 2001) implies that before individuals start using the Internet channel, a positive attitude towards the technology needs to be formed. The benefits, such as convenient and economic gains as well as time saving, can be highlighted as positive features of the Internet channel (Jepsen, 2007; Morton, Zettelmeyer & Silva-Risso, 2001; Ratchford, Lee & Talukdar, 2003; Zettelmeyer, Morton & Silva-Risso, 2006). Retailers and marketers in Malaysia should continue publicizing these benefits so that customers and potential customers will develop a positive attitude towards the Internet channel.

The linkage between compatibility and attitude has also been found in other studies (e.g., Rogers, 1983; Taylor & Todd, 1995; Tornatzky & Klein, 1982). This finding suggests that a positive attitude towards channel-switching intention in the Internet and brick-and-mortar stores channels can be developed by highlighting the compatibility of the technology as well as traditional stores with individual existing values and needs. Communicating, working and entertaining online and stores shopping reflect the current and future lifestyle. Some of the consumers prefer online shopping due to the fact that they are accessed to more and faster product information through Internet channel. On the other hand, some of the consumers change channel from Internet to store because they are more comfortable with store and the traditional channel is more compatible with most of Malaysian consumers’ lifestyle. In addition, some of the consumers use both Internet and brick-and-mortar stores channels to decrease cost of shopping (e.g., consumers searching for information through online and purchase in a store).

As already mentioned, complexity significantly predicted attitude towards channel-switching intention from the Internet to brick-and-mortar stores, but did not affect attitude towards channel-switching intention from store to the Internet channel. It shows that the Internet channel is still not very easy to use for most of the Malaysian consumers and they preferred to switch channels from the Internet to stores due to the complexity of the Internet channel. As a result, consumers change channel from the Internet to stores because brick-and-mortar store is more compatible with their lifestyle and is easier to use. The findings imply that retailers and marketers need to make the Internet channel easy to use; otherwise, consumers will prefer to use the store channel.

One interesting aspect of this finding is the great effect of relative advantage compared to complexity on attitude towards channel-switching intention from brick-and-mortar stores to the Internet. This suggests the significant role of relative advantage (usefulness) over complexity in influencing individual’s attitude to change channel from stores to the Internet channel. We believe, with respect to increasing a positive attitude towards using the Internet channel, more stress should be put in emphasizing the usefulness of Internet channel. It does not mean that we should abandon the efforts to make the Internet channel easy to use. It indicates a higher need to promote the usefulness (relative advantages) of the online shopping and Internet channel over its complexity.

In the current study, NB (friends, family and co-workers) strongly predicted SN towards channel-switching intention in both the channels (Tables 4 and 5). Previous research works have indicated that NB also directly and positively affects SN in both online and store purchase intention studies (Bhattacherjee, 2000; Evans, Christiansen & Gill, 1996; Lim & Dubinsky, 2005). From a practitioner standpoint, these findings suggest that there are three reference groups who influence an individual to change channel from Internet to brick-and-mortar stores channel or from stores to Internet channel. Thus, taking these reference groups into account, more effective advertising and promotional efforts can be developed. The results underscore the importance of using positive testimonials from these reference groups to promote either Internet or brick-and-mortar store channel.

Perceived behavioural control as an endogenous construct was predicted by self-efficacy and facilitating condition as exogenous constructs. Information was a significant exogenous construct for PBC regarding channel-switching intention in both the channels. Previous studies have shown that consumer information about product, services and channels would lead them to decide in order to choose a channel/multichannel or otherwise (Klein, 1998; Swinyard & Smith, 2003). Besides, prior research works on channel-switching behaviour indicated that consumers’ perceptions of channel switching were significantly affected by their knowledge and information about channels (Mathwick, Malhotra & Rigdon, 2002; Noble, Griffith & Weinberger, 2005; Verhoef et al., 2005). Based on the results, in this research, the information affected PBC more significantly in regards to channel-switching intention from brick-and-mortar stores to the Internet. It shows that if consumers have high availability of information online, this will lead them to change channel from stores to the Internet. It can be also inferred that Malaysian consumers’ knowledge and information about either Internet or brick-and-mortar stores impacted on their decision to switch channels.

Perceived behavioural control was also influenced by price. Multichannel consumers are able to compare price of channels and select the best one (Brooks, Kaufmann & Lichtenstein, 2008; Degeratu, Rangaswamy & Wu, 2000). Previous studies have shown that consumers evaluate price of channels and then choose/switch channel based on their findings (Grewal, Iyer, Krishnan & Sharma, 2003; Ramaprasad, Douglas & Pillai, 2010). In addition, according to prior studies and in regards to associated cost with collecting information of price comparison, the highest and lowest costly channels were brick-and-mortar stores and Internet, respectively (Degeratu et al., 2000; Ramaprasad et al., 2010). However, in the current research, price was positively associated with both the Internet and brick-and-mortar stores channels, but findings have shown that price affected PBC more significantly in regards to channel-switching intention from brick-and-mortar stores to the Internet (Tables 4 and 5). It means that a high product price and difficulty in terms of finding the lowest price in brick-and-mortar stores impacted consumers’ PBC to change channel from stores to the Internet more significantly.

Self-efficacy is defined as ‘concerned with judgments of how well one can execute courses of action required to deal with prospective situations’ (Bandura, 1982). Self-efficacy is a person’s judgement about being able to perform a particular activity (Bandura 1977, 1982). Thus, self-efficacy reflects how confident Malaysian consumers are about changing channel from brick-and-mortar stores to the Internet and vice versa. Findings in this study have shown that self-efficacy was strongly and positively associated with PBC regarding channel-switching intention in both the channels. This finding implies that efficacy or confidence to switch channel and use either Internet or brick-and-mortar stores channel. In the case of Internet channel, among Malaysian consumers, lack of confidence in using a new software may suggest an organization’s need to provide additional training to their employees. The solutions may not be so obvious. However, some suggestions are provided that might be used by retailers and marketers to improve individual’s self-efficacy towards using the Internet channel in Malaysia.

Attitude was an important predictor for both the Internet and brick-and-mortar stores channels. Prior studies also supported that attitude significantly and positively impacted consumers’ intention (Ajzen, 1991; Rhodes & Courneya, 2003; Shih & Fang, 2004; Shim et al., 2001). Malaysian consumers changed channels from the Internet to brick-and-mortar stores and vice versa when a channel is not favourable. The results of the study confirmed the role of attitude towards consumers’ channel-switching intention in regards to both the channels. This result is in line with previous findings of the role of attitude towards online and offline consumers’ shopping behaviour and consumers’ channel-switching behaviour (Abdul-Muhmin, 2011; Gopi & Ramayah, 2007; Kim & Park, 2005; Pookulangara & Natesan, 2010; Pookulangara et al., 2011; Shim et al., 2001; Wang, Chen, Chang & Yang, 2007). As discussed earlier, relative advantage, compatibility and complexity differently affected attitude in the Internet and brick-and-mortar stores channels. Retailers and marketers need to pay attention to these factors as well as on attitude itself to find out how consumers’ behavioural attitude is influenced by these factors and how attitude affects consumers’ channel-switching intention to whether switch channel from the Internet to brick-and-mortar stores and vice versa.

Moreover, research works on relationship between SN and intention have shown different results. Some of the researchers indicated that SN did not predicted intention in their studies and some other investigators claimed that there is a strong and significant association between SN and consumers’ intention (Chau & Hu, 2001; Mathieson, 1991; Venkatesh & Davis, 2000). The statistical results were, in general, consistent with the previous findings related to the stores and online shopping intention of consumers’ channel-switching intention (Nicholson et al., 2002; Pavlou & Fygenson, 2006; Pookulangara et al., 2011; Ramayah et al., 2003). The results showed that SN had a significant influence on consumers’ channel-switching intention. Thus, based on the findings of this research, Malaysian consumers are significantly influenced by judgement of other significant factors in order to switch channel from the Internet to brick-and-mortar stores and vice versa.

Perceived behavioural control is the third component of TPB that predicts intention. Perceived behavioural control refers to people’s perception of the ease or difficulty of performing the behaviour of interest. In other words, it is an individual’s confidence in his/her ability to perform the behaviour intention based on the presence or absence of requisite resources and opportunities (Ajzen & Madden, 1986; Mathieson, 1991). Previous studies have claimed that consumers who perceived more capabilities or ease are more likely to act on an intention and switch from one channel to other or be loyal with a channel (Hsieh & Liao, 2011; Johnson et al., 2004; Shim et al., 2001; Wang et al., 2007). The results of this study revealed that PBC did not predict channel-switching intention in both the Internet and brick-and-mortar channels. Regarding consumers’ channel-switching intention, it can be inferred that when Malaysian consumers perceived power, control beliefs and feel able to change channel from the Internet to brick-and-mortar stores and vice versa, they are less likely to do so. These results are in line with the findings of Pookulangara et al. (2011) (in the case of switching channel intention from the Internet to stores/catalogue channel) and Lim and Dubinsky (2005) (an investigation on online shopping).

This study presents an investigation on consumers’ channel-switching intention in regards to two channels (i.e., the Internet and brick-and-mortar stores) in Malaysia. In view of this, an imperative theoretical contribution of this research is to study and develop the understanding of Malaysian consumers’ channel-switching behaviour by applying TPB. Academically, this study extends the application of the TPB model to consumers’ channel-switching behaviour in Malaysia. This research is also important because the study on consumers’ channel-switching behaviour is still new and needs further investigation (Choi & Park, 2006; Kim & Park, 2005; Pookulangara et al., 2011). On the other hand, the numbers of multichannel consumers are rising; hence, it is necessary to study consumer channel-switching behaviour with regard to the Internet and/or brick-and-mortar stores channels (Pookulangara et al., 2011). In addition, since this study is carried out in Malaysia, the current study extends the existing body of knowledge related to the TPB. Indirectly, the current research examines the robustness of the theory in its capability to measure adoption intentions within different sampling frames. Finally, the study has utilized a well-grounded theory and, therefore, contributed to our understanding of factors that are relevant to the acceptance of consumers’ channel-switching behaviour in Malaysia.

This research decomposed the attitudinal structure of individual’s intention to adopt consumer’s channel-migrating behaviour by considering two channels (traditional stores and especially Internet as a new technology). This provides more precise understanding of the antecedents of adoption (Taylor & Todd, 1995). Better understanding of these factors assist retailers and marketing executives in determining which ones are important to their customers’ intention to adopt consumers’ channel-switching behaviour in regards to Internet and brick-and-mortar stores channels. This also helps retailers and marketing executives to formulate strategies that could significantly affect channel-switching adoption among their consumers. Many multichannel consumers among these two channels (i.e., Internet and brick-and-mortar stores) in developing countries might share the same exposure, experience or go through the same phase of progress in their channel-switching behaviour endeavours as consumers that would change the channel from Internet to brick-and-mortar stores or vice versa as consumers’ channel-migrating behaviour in Malaysia. Since consumers in other developing countries may share the same issues faced by the Malaysian consumers, it is expected that the finding from this research will help retailers and marketing executives in other developing countries in understanding the consumers’ channel-migrating behaviour by considering Internet and brick-and-mortar stores channels in the current study as well.

Implications and Recommendations

From a theoretical perspective, the results of this dissertation have an important implication for the theory of planned behaviour. Theory of planned behaviour (Ajzen, 1991) was chosen as the guiding framework for this study. Theory of planned behaviour was decomposed for several reasons. By decomposing attitudinal structure, it is expected to have a higher explanatory power and a more precise understanding of the antecedents of behaviour (Taylor & Todd, 1995). Limitations of the TPB influenced researchers to decompose the TPB model. In the TPB, the belief structures are typically combined into unidimensional constructs. This integration of beliefs has been subject to criticisms. Combining multidimensional beliefs structures into a unidimensional construct can cause difficulty in interpretation specificity (Bagozzi, 1981; Shimp & Kavas, 1984). For instance, based on a review of literature on innovation adoption, the three key determinants of the adoption decision are perceptions of relative advantage, compatibility, and complexity (Rogers, 1983; Tornatzky & Klein, 1982). Internet is one of the innovation adoptions that is used by consumers for online shopping. Thus, as one of the channels of this study was the Internet, three dimensions (relative advantage, compatibility and complexity) were implied to evaluate the effect of attitude towards channel-switching intention.

Additionally, results of the study demonstrated once again the robustness of the decomposed TPB for helping to explain consumers’ channel-switching behaviour. Other studies have also successfully used the TPB as a theoretical framework from which to explain intention towards multichannel shopping behaviour or channel-switching behaviour (Chau & Hu, 2002; Chiu et al., 2011; Lohse & Spiller, 1998; Stone et al., 2002; Vijayasarathy, 2002; Yang et al., 2007). As more and more studies of Internet purchasing behaviour and its antecedents are done within the TPB framework, we are more able to discover and confirm which antecedents are most important, helping us build a robust theory of channel-switching behaviour.

Attitude is anchored to relative advantage, compatibility and complexity and SN was influenced by NB. Normative belief is defined as the influence of reference group in order to perform behaviour (Amin & Ramayah, 2009; Fishbein & Ajzen, 1975). Previous studies of consumer shopping behaviour and channel-switching behaviour indicated that SN is influenced by NB (George, 2004; Lim & Dubinsky, 2005; Ryan & Bonfield, 1980). Findings of this research also confirmed that NB is a very strong predictor of SN towards channel-switching intention in both the channels.

Furthermore, in the pure TPB model, the path from PBC to intention has failed to achieve significance. Ajzen and Madden (1986) claimed that the PBC is less likely to be related to intention. This suggests that the relationship of PBC to intention is tenuous and merit further investigation. According to the measurement of this study, PBC also did not predicted channel-switching intention in both the channels. Additionally, two components were encompassed in the decomposed TPB model. Both facilitating conditions (price and information) and self-efficacy were significant determinants of PBC in this study. Price and information are known as facilitating conditions which predict PBC (Taylor & Todd, 1995; Triandis, 1980). Previous studies have shown that facilitating conditions affected PBC (Noble & Phillips, 2004; Park, Klein, Smith & Martell, 2009; Stigler, 1961; Yang et al., 2007). This is also conformed in the current research. Taylor and Todd (1995) found that self-efficacy predicted intention to use a wide range of technologically advanced products. Thus, an individual with a confident command of computer skills and familiarity with the Internet is more inclined to adopt Internet banking. Findings in this study indicated that self-efficacy affected PBC with respect to the Internet and brick-and-mortar channels.

The results of this study show that attitude and SN significantly affected consumer’s channel-switching intention regarding Internet and brick-and-mortar stores channels. The successful validation of these constructs on channel-switching intention demonstrates that the TPB model is well founded. In addition, from the viewpoint of consumers’ channel-switching behaviour in Malaysia, this study contributes positively to research utilizing a well-grounded theory.

Limitations and Future Research

As with any study, there are limitations to this research. First, in the current investigation, attitude was measured by three components (relative advantage, compatibility and complexity). As previous researchers contend with respect to multichannel consumer behaviour and channel-switching behaviour by using TPB, attitude can be measured by other dimensions, such as the hedonic and utilitarian behavioural belief scale (Pookulangara & Natesan, 2010; Voss, Spangenberg & Grohmann, 2003). As such, future research should incorporate hedonic and utilitarian constructs into a broader TPB model on analyzing the impact of attitude on channel-switching intention.

Second, in this research, Malaysian consumers’ channel-switching behaviour was evaluated only in two channels (Internet and brick-and-mortar stores channels). Technology is in constant progress; new devices, such as tablets, are available to browse for products and mobile apps to shop online are becoming popular among consumers; as technology is progressing and mobile online sales are increasing, consumers’ shopping habits are also changing. Younger generations have great technology assimilation and are growing with an online culture; therefore, understanding mobile commerce and its potential is fundamental. Catalogue is the other suggested channel that can be examined. Therefore, it is recommended that future researchers study Malaysian consumers’ channel-switching behaviour by examining each of the individual channels (e.g., catalogue, mobile phone, tablet and brick-and-mortar stores). In addition, future studies are suggested to examine less pair similar channels (e.g., catalogue and brick-and-mortar stores; brick-and-mortar stores and Internet by applying new online shopping devices, such as mobile phone as well as tablets).

Third, in this study, the questionnaire was collected from Malaysian consumers in two regions of the country (i.e., Klang Valley and Penang), and this could lead to differences in the parameters under study (Safiek & Hayatul, 2009). Besides, as one of the main channels in this study is the Internet and according to Sulaiman, Ng and Mohezar (2008), students are more familiar with the Internet and computer usage compared to other groups. Thus, it is suggested that future researchers examine Malaysian consumers’ channel-switching behaviour by using students as the respondents for each region separately.

Fourth, to measure channel-switching search and purchase behaviour, respondents were asked to recall how many times during the last 6 months they had changed channel while searching for a product or service or purchased goods or services from Internet to brick-and-mortar stores channels and vice versa. These measures are subject to recall errors and do not require subjects to list actual purchases through these channels or products/services that they searched for. Additionally, the scaling for these variables ranged from ‘never’ in the last 6 months to ‘more than 15 times’ in the last 6 months which might have caused a ceiling effect given that consumers might have searched for a product/service more than 15 times in a day. Future researchers in this area might try to collect a time component to these measures by asking respondents to indicate how much time, on average, they have spent searching for products within a specific channel and then changed channel in the last 6 months (a similar question could be asked for purchasing behaviours). These measures also did not measure search or purchasing effort, which should be included in future research endeavours.

Finally, in an attempt to understand consumers’ channel-switching behaviour at a general level, consumers were not directed to respond in relation to a specific product. Consumers are often presented with a product decision prior to channel selection/switching, which presents a limitation to the work. Although Mathwick, Malhotra and Rigdon (2002) indicated that many consumers select a retail channel first, which eventually results in a shopping decision, research focusing on the product/channel decision-making process would significantly enhance the understanding of this critical issue. It can be noted that the nature of the product could determine channel selection as well as channel-switching behaviour. Products that consumers feel need to be seen, touched, tasted, tried on, etc., prior to purchase are more likely to be purchased through different channels than products that are electronically conveyable or have limited distribution. As such, future research should explore consumers’ channel-switching behaviour as it relates to specific purchases and products.