Switching intention to online channel in Vietnam – A case study of consumer electronics goods

Abstract

Online retailing has been growing significantly. However, electronic retailers have still encountered some issues related to consumer psychology and potential risks besides the widely acknowledged advantages of the online channels. This current study extends the PPM (Push-Pull-Mooring) framework by adding customer psychological variables to explain the consumers’ switching intention to online channels. In detail, Push refers to limitations of traditional channel at physical stores, Pull and Mooring present pros and cons of online channel. The relationship of consumer psychology, including Perceived Risk, Attitude towards switching and Habit of buying goods at physical stores were integrated to adapt to the characteristics of buying consumer electronics goods. Conducting a survey of 1104 customers grouped in 3 big cities: Hanoi, Ho Chi Minh City, and Da Nang, the research results show that: consumer skepticism originating from Perceived Risk negatively influences Attitude towards switching (−0.192); Attitude towards Switching channels has the strongest influence (0.374) on Switching intention, while buying Habits at traditional stores lessens the Switching intention (−0.136). This is the new contribution to existing literature regarding switching from traditional channels to online channels for goods requiring consumer experiences.

Keywords:

• buying offline and online
• PPM
• switching intention

Reviewing editor:

• Kaouther Kooli, Bournemouth University, United Kingdom

Subjects:

• Asian Studies; Consumer Psychology; Business, Management and Accounting

1. Introduction

Online retailing is booming both globally and in emerging markets like Vietnam. Despite the challenges posed by the COVID-19 pandemic, the online retail sector saw a remarkable growth of 15% ($13.2 billion) in 2021 (VECOM, 2020), with estimates projecting a growth rate of 29% from 2020 to 2025, resulting in a market size of $52 billion by 2025 (Anh, 2021).

According to websolution (2019), electronics items are also among the top 10 best-selling items online. The figures of prestigious organizations for the years 2019, 2020, 2021 also clearly show this (AI, 2020; Reputa, 2021; VECOM, 2020).The pandemic and the “new normal” have led to a surge in online purchases, particularly among consumer electronics like laptops, desktops, and accessories for various purposes such as work, learning, and entertainment (Khải, 2021; VECOM, 2022). However, it is essential to note that the transition from physical stores to online shopping is not complete, especially for high-priced or experiential products (Nielsen, 2017). Concerns about product quality, lack of accreditation, and the absence of the physical store experience remain significant factors hindering the shift to online channels (VECOM, 2022).

As far as we could reach, there are only 05 studies on Switching intention of from offline to online channel, and moreover, these studies are different contents: intention to switch English learning from offline to online (Chen & Keng, 2019), intention to switch purchases of general items from a physical store to a mobile device (Chang et al., 2017), intention to switch to using cloud-based healthcare services (Lai & Wang, 2015), intention to switch to online grocery shopping (Monoarfa et al., 2023) and switching behavior of buying cosmetics online thanks to AR technology (Nugroho & Wang, 2023). Studies on this channel switching theme are very rare, as mentioned in Nimako (2012). There is a lack of research into a systematic model of significant factors that represent the influences of advantages and limitations of buying in online channel and buying at physical stores on the intention of switching buying channel. Besides, for products that require customer experience like consumer electronics goods, psychological hindrance may make customers more hesitant in their intention to switch to buying on an online channel. In order to fill in these two research gaps, this research aims at investigating factors affecting the Switching intention, which remained quite rare in literature (Chen & Keng, 2019; Lai & Wang, 2015), from the perspective of the PPM framework.

By review studies that contains factors suitable for the essential meaning of each “Push”, “Pull”, “Mooring” group, we build up the model based on the PPM paradigm as follows:

+ Push is the limitations of buying in physical stores: Forsythe et al. (2006), Lester et al. (2006), Chiu et al. (2011) …

+ Pull is the advantages of buying online: Forsythe et al. (2006), Lester et al. (2006), Jiang et al. (2013), Ye and Potter (2011a), Monoarfa et al. (2023)) …

+ Mooring is the limitations of buying online: Forsythe et al. (2006), Lester et al. (2006), Nugroho and Wang (2023), Ye and Potter (2011a), Bustamante and Rubio (2017), … Among those factors is the Perceived risk factor of buying electronic goods online as an antecedent for customers’ hesitating psychology.

+ Such psychological factors as attitude (A) (Pookulangara et al., 2011; Primabudi & Samopa, 2017; Wani & Malik, 2013a) and habit (H) (Limayem et al., 2007) are integrated into the model reflecting customer’s apprehension of potential risks. The linking of perceived risk—attitude – habit—switching intention all have scientific basis. This psychological sequence shows that customers’ reluctance to switch to online shopping for electronic products is due to the awareness of risks that are inherent to online shopping and the habit of buying at physical stores is still somewhat maintained. It is likely that this assumption hinders Switching intention, even though buying online has many advantages and is increasingly popular in the digital transformation trend. This is also a research gap on customer psychology, thus we propose a PPM-H-A variant model to examine the influence of consumer psychology on the intention of switching to online buying channels in the context of experience goods.

2. Literature review and hypothesis development

2.1. Switching intention

A relevant question that confronts business organizations is how to retain consumers. However, more and more offline platform operators now face customer switching. Switching intention is defined as “the possibility of the behavior of switching the transaction of consumers from a previous brand to another competitive brand” (Nimako, 2012).

From perspective of buying channel switching, the intention to switch channels presents the customers’ tendency to switch from the current channel to another one, i.e., from offline channels to online ones (Chang et al., 2017; Chen & Keng, 2019; Monoarfa et al., 2023). Customers may switch channels when they believe online channels might be more useful and popular than offline ones (Nugroho & Wang, 2023). The ability to adopt technology, i.e., traceability or payment, increases the intention to switch (Monoarfa et al., 2023; Singh & Rosengren, 2020). In addition, customers might feel overloaded and fatigued after buying from offline channels for a long time, which motivates them to stop using the current channels and decide to switch (Chaouali et al., 2016; Maier et al., 2015).

2.2. The PPM theory

The theory of Push-Pull-Mooring (PPM) was founded by Lee (1966) and Moon (1995), and was empirically tested in numerous scholar works (Bansal et al., 2005; Haridasan et al., 2021; Monoarfa et al., 2023; Nugroho & Wang, 2023). The PPM theory is applied to represent the linear relationship between a combination of factors that influence the Switching intention/behavior of human being with regards to accommodation, suppliers, or transaction modes. The PPM theory mainly examines factors that can influence the switching intention including push, pull and mooring. Push factors tend to motivate and push people away from current choices; Pull factors refer to advantages of the new method which may encourage people to switch to the new mode; and Mooring effects refer to psychological and social hindrances, conditions, and limitations of the new method when people consider switching. The Figure 1 below is the framework model:

Figure 1. The Push-Pull-Mooring framework.

Source: Bansal et al. (2005)

2.2.1. Push factors

Consumers were found to encounter certain disadvantages when buying at traditional stores (Aryani et al., 2021; Gupta, 2015). However, these scholars did not empirically evidence the influence of these disadvantages on motivating customers to switch to new channels. There are many different disadvantages such as difficulty in product search, product comparison, waste of time and effort, and even pressures from sellers (Forsythe et al., 2006; Gupta, 2015; Lester et al., 2006). These disadvantages may push customers to switch to online channels from traditional channels.

Chiu et al. (2011) also considers vicarious experience as one of the Push factors that can influence the intention of using free riding cross-channel. When consumers learn that many of their relatives have made successful electronic transactions, they will have an intention to switch. Even when consumers have not done any transaction previously, they can consider the results of others’ transactions as an indirect experience. Thus, the following hypotheses are posited:

H1(+): Push factors are positively associated with the switching intention.

H1a(+): Disadvantages of traditional channels are positively associated with the switching intention.

H1b(+): Vicarious experience of others is positively associated with the switching intention.

2.2.2. Pull factors

Customers may shop online for a wide variety of reasons such as reasonable price, convenience, perceived ease of use ((Handayani et al., 2020; Jiang et al., 2013; Lester et al., 2006; Monoarfa et al., 2023; Mpinganjira & Studies, 2015).

First, scholars (Elida et al., 2019; Handayani et al., 2020; Hooi et al., 2021; Lester et al., 2006) have reported that when buying goods online, customers benefit from a more reasonable price. Consumers could easily make price comparisons across online channels.

Second, buying online is very convenient because buyers can easily search for information; evaluate the product; save time and effort; have the product delivered at selected location, and so forth (Jiang et al., 2013; Mpinganjira & Studies, 2015). This factor has been indicated to strongly encourage consumers to switch to buying in online channels (Forsythe et al., 2006; Johan et al., 2020).

In addition, perceived ease of use is when a person can use a new technology easily (Davis, 1989; Monoarfa et al., 2023). Users will perceive buying online to be easy when they do not need to make too much effort (Perea et al., 2004). Thanks to current advanced technology, Ye and Potter (2011a) stated that consumers consider switching to online channels because they can use technology on their personal devices. Perceived ease of use has been integrated into Pull factors in the context of buying in online channels (Monoarfa et al., 2023); Bauerová and Braciníková (2021); Kumar and Kashyap (2022)). Thus, the following hypotheses are suggest as follow:

H2(+): Pull factors are positively associated with the switching intention

H2a(+): Price (in online channels) is positively associated with the switching intention

H2b(+): Convenience (in online channels) is positively associated with the switching intention

H2c(+): Perceived ease of use is positively associated with the switching intention

2.2.3. Mooring factors

In the context of switching to online channels, mooring factors are referred to as limitations of buying in online channels, which are also partial advantages of buying in traditional channels at physical stores (Forsythe et al., 2006; Han & Kim, 2017; Ilhamalimy & Ali, 2021; Lester et al., 2006; Neslin et al., 2006; Nugroho & Wang, 2023; Verhoef et al., 2007; Wang et al., 2013).

First, perceived risks in online channels are considered as the biggest challenge. It is defined as the consumers’ uncertainty of experience when they cannot predict outcomes of their purchase decision (Chellappa et al., 2005). For instance, consumers may hesitate to buy online if they perceive risks related to information security and privacy. Consumers are also concerned about the risk of delivery. Risks might also occur when products cannot be tested prior to use. Customers can be irritated when post-purchase benefit does not live up to their initial expectation (Yang et al., 2010). In general, when consumers perceive high possibility of these risks, they hesitate to switch (Bhatti & Ur Rehman, 2020; Marriott & Williams, 2018).

Secondly, subjective norm is another mooring factor that might be negatively associated with the switching intention (Lian & Yen, 2013; Ye & Potter, 2011a). In a study about the psychology of shoppers, prudence and collectivism of oriental people affect online shopping behaviors which might delay their behavior of switching channels (Lian & Yen, 2013; Zheng, 2013).

Thirdly, experience at physical stores has a significant impact on consumer satisfaction and loyalty (Bascur & Rusu, 2020; Bonfanti et al., 2020; Bustamante & Rubio, 2017; Forsythe et al., 2006; Lester et al., 2006; Silva et al., 2021; Verhoef et al., 2009). Thus, a lack of physical store experience in online channels might discourage customers from switching channels, thereby being a mooring factor. Therefore, when considering switching to buying in online channels, the lack of in-store experience will discourage customers from switching to online channels. Thus, the following hypotheses can be infered that:

H3(-): Mooring factors are negatively associated with the switching intention.

H3a(-): Perceived risks (online) are negatively associated with the switching intention.

H3b(-): Subjective norms are negatively associated with the switching intention.

H3c(-): (A lack of) Physical store experience is negatively associated with the switching intention.

H3d,3e(-): Mooring factors negatively moderate the impact of Push and Pull factors on the switching intention.

2.2.4. Attitudes towards switching to online channels

According to Pookulangara et al. (2011), Attitude toward switching buying channel is defined as consumers’ evaluation of outcomes of their buying behaviors in the channel they choose. Scholars have provided different explanations about what leads to attitudes toward switching buying channels (Madahi & Sukati, 2014; Primabudi & Samopa, 2017; Sinha & Kim, 2017; Wani & Malik, 2013b). Results revealed that perceived risks directly impact consumers’ attitude while risks related to finance, products, convenience, and delivery negatively influence attitude towards buying behaviors in online channels. Specifically, scholars have found that buying consumer electronics goods in online channels certainly poses a variety of perceived risks for consumers (Harn et al., 2006; Levin et al., 2003; Zhang, 2008). Therefore, the following hypothesis is:

H4(-): Perceived risks (in online channels) are negatively associated with attitude towards switching to online channels.

People tend to behave consistently with their Attitude. Bansal and Taylor (1999) state that the more favorable Attitude toward switching, the more likely the Switching intention. Findings of Madahi and Sukati (2014), Nikseresht (2016) and Palau-Saumell et al. (2021) also infer that Attitude towards switching positively influences the Switching intention to buying in online channel. Therefore, the following hypothesis is:

H5(+): Attitude towards switching to online channels is positively associated with the switching intention.

2.2.5. Habit

Habit has been integrated to the PPM theory to develop the combined framework named “PPM-H” to test the impact of habit on the switching intention in some contexts such as cloud application services; and learning English online with real people (Chen & Keng, 2019; Cheng et al., 2019; Lai & Wang, 2015). Results indicate that individuals unconsciously perform a certain behavior because of repeated activities. Thus, habit can reduce the chance that consumers will consider alternatives; and therefore, stay with the current choices (Limayem et al., 2007). This will slightly influence the switching intention (Lai & Wang, 2015; Sun et al., 2017). Other studies have found out that habit has a negative impact on the switching intention, especially in contexts related to technology (Chen & Keng, 2019; Cheng et al., 2019; Lin et al., 2021). Therefore, the following hypothesis is:

H6(-): Habit is negatively associated with switching intention.

According to Tuu (2015b), habit is indicated to negatively moderate the relationship between attitude and intention because the automaticity of behavior diminishes the need to access to consumers’ perception of switching intention. This means among people who have a habit, the predictive power of attitude on intention seems to be weaker. In contrast, for those who do not have a habit of doing things, their attitude may act as a stronger predictor of their intentions. Most studies have found that the attitude—intention relationship is typically weaker when the behavior is habitual than when the behavior is not habitual (De Bruijn et al., 2007; Honkanen et al., 2005). Therefore, the following hypothesis is formally posited:

H6a(-): Habit negatively moderates the impact of attitude towards switching channels on switching intention.

The research model is proposed in this Figure 2:

Figure 2. Research model.

3. Methodology

3.1. Teamwork discussion/interview

Measurement scales are adapted from previous literature (see Index 4). The initial measurement scales consisted of 68 items. After consulting with 04 academic experts in marketing and e-commerce in October 2020, 54 items were finalized to assure the simplicity, the meaningfulness, and the feasibility of the research model. The item scale used is Likert 7, with point from 1–7 expressing the assessment of consumers representing strongly disagree to strongly agree.

Besides, 11 consumers were interviewed in January 2021 to examine the validity of the survey sheet. This helped revising the questions to make them easy to understand for the respondents. After that, 22/54 items were revised to make them more understandable for Vietnamese people but still maintain their meaningfulness. The second purpose is to explore the main hindrance for customers to switch. Among 3 mooring factors (Perceived risks, Subjective norms, lack of Physical store experiences), Perceived risks was voted by 9/11 consumers to be the most concerning issue when they consider switching to online channel. This also suggests Perceived risks as the most influential antecedent variable to set in the hypothesized psychological sequence of Perceived risk—Attitude—Habit. Details of the interview records are available upon request.

3.2. Pilot study

Due to the complexity of the 2-level structure, we conduct a pilot survey to evaluate the validation of the scale by CFA (confirmatory factor analysis). We distributed survey sheets to 400 participants, collected 312 responses, and filtered out (missing information, answered “vertically”) for 289 valid observations. Regarding sample characteristics: 53.63% male (155) and 46.37% female (134); 34.36% students (99) and 65.64% working (190); the number of online purchases of electronic goods in the last 2 years is 10.55% for 0 times (30), 50.65% for 1–2 times (147), 20.35% for 3–4 times (59) and 18.45% for more than 5 times (53) respectively.

The rules for evaluating the quality of the scale are guided in Churchill (1979) and Steenkamp and van Trijp (1991). Use the software SPSS 25, AMOS 22 to assess for both second order structures (Push, Pull, Mooring), and first order ones (Attitude, Habit, Switching intention). Results of CR, AVE for all latent variables are valid measurement (see index 1).

3.3. Official full survey

We continue to use the Likert 7 scale for items, and nominal/ordinal for demographic questions.

Regarding sample size and distribution, in the materials of F Hair et al. (2014), apply the 10× rule of Barclay et al. (1995), and statistical power that Cohen (1992) proposal should also be considered when choosing sample size. With effect size = 0.2, statistical force = 0.8; indicator number is 54 and latent variable number is 15, the recommended sample size is 530.

We decided to distribute the survey sheet to 2000 participants, expecting the number of valid observation to analyze is about 1000 to ensure meaningful results.

According to the report on the e-commerce index by locality in the e-commerce market report of VECOM (2021), we made survey for the 03 cities of Ho Chi Minh city, Hanoi và Da Nang with rankings 1, 2, 3 point (67.63; 55.66; and 19.04 respectively), as well as regional representativeness. Implement geographical stratification to distribute survey sheet in these three cities. The distribution of sheets is distributed proportionally to the percentage of the population of these three cities. Then, with the total number of sheets distributed in each city, re-allocate them correspondingly to the percentage of the population of the each district.

The survey was conducted by snowball distribution through acquaintances with wide social network. This is a “non-sensitive” research issue, because respondents are not pressured to answer to follow social desirability, and the opinion of buying online can be generalized to this modern digital society. The acquaintances are chosen by their living place, then they transfer the sheets to participants nearby and online. The Covid period lasting from 2020 to early 2022, the digital transformation process is widely known, the conditions to access the internet widely … also support this argument.

The actual survey period took place from December 2020 to March 2022 due to the complicated stages of the Covid epidemic affecting the results of this study. Specifically, there are three phases: the Covid epidemic is basically controlled through social distancing (from December, 2020 to before 28 January 2021), social quarantine because of serious community spread (from 28 January 2021 to November, 2021), and adapting to new normal conditions as widely vaccinated (from January 2021 to June, 2022). The reason for surveying in these time intervals is to avoid biased results when synthesizing.

After distributing sheets to 2000 participants through both direct survey and online survey, the total number of sheets collected is 1182 votes (59.1%). After collected, survey sheets that are missing information, disordered, or “abnormally vertical” ticks are discarded (F Hair et al., 2014). The final number of valid observations is 1104. Different demographic characteristics need to ensure a reasonable ratio, avoid bias. Details are presented in the descriptive statistics (Table 1).

Table 1. Demographic information.

Demographic Characteristics	Nominal/Ordinal scale	Number	Percentage
Gender	Male	559	50.63%
	Female	545	49.37%
Age	Under 23	384	34.78%
	23–30	519	47.01%
	30–40	126	11.41%
	Over 40	75	6.79%
Profession	Student	348	31.52%
	Working	756	68.48%
Place	Hanoi	496	44.93%
	Da Nang	80	7.25%
	Ho Chi Minh city	528	47.83%
Online buy times	0	176	15.94%
	1–2	451	40.85%
	3–4	210	19.02%
	Over 5	267	24.18%
Product type	Smartphone	254	23.01%
	Sound devices	437	39.58%
	Television	14	1.27%
	Visual devices	25	2.26%
	Appliance	208	18.84%
	Others	166	15.04%
Buying channel	Pure online	209	18.93%
	Click-and-mortar	775	70.20%
	Facebook	120	10.87%
Expenditure (USD)	Under 216	584	52.90%
	216–432	214	19.38%
	432–2160	279	25.27%
	Over 2160	27	2.45%
Phase	Controlled	405	36.69%
	Widespread	312	28.26%
	Normal	387	35.05%

+ Male accounted for 50.63% (559) compared to 49.37% of females.

+ The percentages of age of the respondents are 34.78%, 47.01%, 11.41%, 6.79% for age groups of under 23, 23–30, 30–40, and above 40, respectively.

+ Respondents who are students, and those who are having a job account for 31.52% and 68.48%, respectively.

+ Those who live in Hanoi, Da Nang and Ho Chi Minh account for 44.93%, 7.25%, and 47.83%, respectively. This is appropriate with population size ratio (8; 1.2; 9 million in these cities, respectively) when comparing these three cities.

+ Times of buying goods in online channel account for 15.96%, 40.85%, 19.03% and 24.18% for 0, 1–2, 3–4, above 5 times, respectively.

+ The types of products bought in online channel are mobile phones, sound devices, televisions, visual devices, appliances and others (23.01%, 39.58%, 1.27%, 2.21%, 18.84% and 15.04%, respectively).

+ Types of online channel include pure online retailer, brick-and-mortar, Facebook (18.93%, 70.20% and 10.87%).

+ Expenditure of under 216 USD (equal to 5 million VND), 216–432 2160 USD, 432–2160 USD and above 2160 USD accounts for 52.90%, 19.38%, 25.27% and 2.45%, respectively.

+ The phases that COVID-19 was basically controlled (2020–28/01/2021), COVID-19 was spreading in the community (28/01/2021–01/2022) and the period that people were adapting to new normal conditions in the society (from 01/2022 on) were also surveyed (with responses accounting for 36.69%, 28.26%, 35.05%, respectively.

3.4. Analysis software

SPSS 25 and AMOS 22 software to evaluate the quality of the scale, analyze descriptive statistics, check the distribution of errors extracted from the model estimation; SmartPLS 3.0 is used for model analysis.

Studies applying PPM can use PLS-SEM, which is more suitable for hypothesis exploration among variables that may not have been found in previous studies, rather than to confirm the well-proven model hypothesis, that widely applied like TAM., TPB … of Davis et al. (1989), Ajzen (1991). This study is exploratory, so using PLS-SEM is more appropriate than CB-SEM, according to Hair et al. (2021), p. 4 and 19.

4. Results

4.1. Descriptive statistics

After filtering, 1104 responses were valid for full analysis. In order to avoid bias, information about gender, age, occupation, educational levels, places of living (based on the percentage of population size in 2020) was also collected.

4.2. Measurement model assessment

The measurement model is empirically validated by means of the PLS-SEM to ensure robust results, including addressing common method bias, data distribution, internal consistency, convergence, and discriminant validity.

To address common method bias, we followed the method proposed by Kock and Lynn (2012). We conducted a two-step process: first, we estimated the PLS-SEM model to extract latent variable scores as independent variables, and then we created a randomized-value variable using the “RAND” function in Excel. By regressing the independent variables against the randomized variable as the dependent one, we confirmed that all VIFs were below 3.3, indicating that the model is free of common method bias.

Regarding data distribution, PLS-SEM does not strictly require normality as CB-SEM does, according to Hair et al. (2021). Instead, we assessed skewness and kurtosis for each item, with the criterion −1 < Kurtosis/Skewness <1. Except for item PRON2 (skewness = −1.013), all other items exhibited skewness and kurtosis within the interval (−1, 1). Consequently, we excluded PRON2 from the model.

Next, we evaluated internal consistency, convergent validity, and discriminant validity. Comparing the threshold values introduced in Hair et al. (2021), we found that all latent variables met the requirements. Internal consistency was confirmed with all Cronbach’s Alpha values exceeding 0.6, while Composite Reliability (CR) values were all above 0.7, and Average Variance Extracted (AVE) values were greater than 0.5. See the Tables 2 and 3. Additionally, discriminant validity was established, as the diagonal cells in each variable were higher than the values within the corresponding column (Fornell & Larcker, 1981). See the Tables 4 and 5.

Table 2. Internal consistency and convergent validity (first order latent variables).

	Cronbach’s Alpha	Composite Reliability	Average Variance Extracted (AVE)
Attitude	0.879	0.917	0.735
Convenience (online)	0.945	0.955	0.751
Disadvantages (offline)	0.816	0.879	0.645
Habit	0.894	0.926	0.759
Perceived Risk (online)	0.908	0.927	0.618
Perceived ease of use	0.905	0.933	0.778
Physical store Experience	0.918	0.934	0.671
Price (online)	0.871	0.912	0.721
Subjective Norms	0.887	0.922	0.747
Switching Intention	0.895	0.934	0.826
Vicarious Experience	0.906	0.935	0.782

Table 3. Internal consistency and convergent validity (second order latent variables).

	Cronbach’s Alpha	Composite Reliability	Average Variance Extracted (AVE)
Push	0.863	0.834	0.717
Pull	0.955	0.925	0.805
Mooring	0.941	0.889	0.729

Table 4. Discriminant validity (among first order latent variables with reflective constructs).

	Attitude	Convenience (online)	Disadvantages (offline)	Habit	Perceived Risk (online)	Perceived ease of use	Physical store Experience	Price	Subjective Norms	Switching Intention	Vicarious Experience
Attitude	0.857
Convenience (online)	0.282	0.866
Disadvantages (offline)	0.260	0.329	0.803
Habit	−0.194	−0.035	−0.088	0.871
Perceived Risk (online)	−0.192	0.120	0.039	0.284	0.786
Perceived ease of use	0.280	0.742	0.346	−0.066	0.058	0.882
Physical store Experience	−0.086	0.181	0.093	0.207	0.587	0.155	0.819
Price (online)	0.218	0.715	0.384	−0.057	0.034	0.696	0.121
Subjective Norms	−0.004	0.224	0.158	0.178	0.543	0.197	0.672	0.849	0.865
Switching Intention	0.550	0.418	0.335	−0.274	−0.229	0.435	−0.087	0.164	−0.046	0.909
Vicarious Experience	0.266	0.434	0.439	−0.045	0.051	0.424	0.164	0.397	0.189	0.348	0.884

Table 5. Discriminant validity (second order and first order latent variables with reflective constructs).

	Attitude	Habit	Mooring	Pull	Push	Switching Intention
Attitude	0.857
Habit	−0.194	0.871
Mooring	−0.125	0.267	0.700
Pull	0.292	−0.055	0.178	0.785
Push	0.310	−0.076	0.150	0.511	0.718
Switching Intention	0.550	−0.274	−0.153	0.460	0.403	0.909

Before analyze the linear structural model, need to check for multicollinearity between latent variables in Table 6 below:

Table 6. Colinearity between latent variables.

	Switching Intention
Att*Ha	1.003
Attitude	1.208
Habit	1.117
Mooring	1.201
Pull	1.431
Pull*Moor	1.628
Push	1.434
Push*Moor	1.601

It can be seen that the VIFs between the variables are all lower than 5, so there is no violation of multicollinearity.

4.3. Linear structural model analysis results

Next, regression analysis was conducted with latent variables. The results are illustrated in Table 7.

Table 7. Regression analysis results.

	Original Sample	P Values	Hypothesis
Push -> Switching Intention	0.150	0.000	H1: supported
Pull -> Switching Intention	0.294	0.000	H2: supported
Mooring -> Switching Intention	−0.140	0.000	H3: supported
Push*Moor -> Switching Intention	0.002	0.932	H3d: unsupported
Pull*Moor -> Switching Intention	0.018	0.448	H3e: unsupported
Perceived Risk (online) -> Attitude	−0.192	0.000	H4: supported
Attitude -> Switching Intention	0.374	0.000	H5: supported
Habit -> Switching Intention	−0.136	0.000	H6: supported
Att*Ha -> Switching Intention	−0.031	0.311	H6a: supported

The coefficient of Push factors is 0.150. It can be seen that when consumers perceive the Disadvantages of buying at physical stores such as limited information, time, inconvenience, pressure; and when they see that others have successfully bought in consumer electronics goods, they are more likely to be pushed to have Switching intention to buying in online channel, and vice versa. This is consistent with Lester et al. (2006), Forsythe et al. (2006) and Aryani et al. (2021). The reason why customers choose to buy online instead of going to a physical store is to look up a lot of information, convenience without leaving home, flexible time as well as more diverse products. Vicarious experience also promote the switching of old buying methods, similar to the conclusion in the study of Chiu et al. (2011). H1, H1a, H1b are supported.

The coefficient of Pull factors is 0.294. When consumers are attracted and pulled by more attractive Prices; Convenience in terms of time, process, order tracking; simple application of technology in buying in online channel, they are more likely to have Switching intention to online channels.

Coefficient of the effect of Price on online purchases (Handayani et al., 2020) is also consistent with this conclusion (0.291).

The regression coefficient of Convenience also has a strong impact on online repeat purchases in Mpinganjira and Studies (2015), Jiang et al. (2013), Johan et al. (2020), with values of 0.686; 0.670; 0.531 respectively.

The studies of Ye and Potter (2011a) and Monoarfa et al. (2023) all show that Perceived ease of use has positive impact on attraction (0.421 and 0.36) as the variable “Pull” affecting Switching intention. H2, H2a, H2b, H2c are supported.

The coefficient of Mooring factors is −0.140. Besides advantages, buying in online channel also has some limitations. Consumers are concerned about risks related to information, product, delivery, payment; general negative Subjective norms about buying in online channel; and the need for Physical store experiences to minimize the potential risks. This may attenuate the Switching intention to buying consumer electronics goods in online channel.

The negative regression coefficient results also show the opposite effect of Perceived risk on online shopping in terms of product risk, privacy risk (Bhatti & Ur Rehman, 2020); economic risks; product risk, psychological risk (Han & Kim, 2017); general risks (Marriott & Williams, 2018).

Similarly, it is the negative effect of the Subjective norms regression coefficient (belonging to the Mooring group) on the Switching intention (Lian & Yen, 2013; Ye & Potter, 2011a). H3, H3a, H3b, H3c are supported.

Perceived risks are found to negatively influence Attitude towards switching to buying in online channel. This is reasonable because when consumers perceive high potential risks, their Attitude tends to be less favorable when they switch to online channels. The negative influence of Attitude is rather strong (β = −0.192). Regression coefficients of Perceived risk to Attitude in Wani and Malik (2013b), Primabudi and Samopa (2017) are −0.43, −0.15 in detail. It can be seen that this same conclusion is reasonable with the previous results. Hypothesis H4 is supported.

Attitude towards switching buying channels has the strongest impact on the Switching intention (β = 0.374) among the factors studied. The research result of Madahi and Sukati (2014) also support this with the regression coefficient affecting Switching intention, namely 0.33, 0.85. It could be linked from H4 that Attitude towards switching is partially controlled by consumers’ Perceived risk before they perform the switching behavior consistent with their intention formed previously. The more Perceived risks, the more likely the positive impact of Attitude towards switching on the Switching intention is decreased. If consumers’ Perceived risks when buying consumer electronics goods in online channel are high, despite the Push and Pull factors, the Switching intention will be mitigated because of a strong impact of Attitude towards switching buying channels—which is relatively controlled by Perceived risks. H5 is supported.

The Habit of buying at physical stores also lowers the intention of switching buying channels (β = −0.136). Consumers’ Habit of buying at physical stores may lessen the Switching intention to buying in online channel. Vice versa, when consumers buy more in online channel, and are satisfied and thus maintain these behaviors in “conditioned” and predictable online contexts, they are more likely to have a stronger Switching intention in the future. Habit has the effect of reducing Switching intention, which also coincides with the conclusions of previous studies in the context of offline to online switching (Chang et al., 2017; Chen & Keng, 2019), respectively −0.25 and −0.11. H6 is supported.

H4, H5, H6 are supported, affirming the framework of PPM-H-A proposed by the authors. The R² coefficient of the Switching intention is 0.462, indicating that latent variables in the model can explain a variation of 46.2% in the Switching intention.

Hypotheses H3d, H3e, H6a related to moderators are not supported despite having the expected sign in the equation. This can be because the survey was conducted between 2020 and the beginning of 2022, when the pandemic situation and post-pandemic impacts could dilute the moderating variables. In order to give up the old Habit, it is necessary to remove the old environment, establish and maintain a stable new environment; and the new behavior maintained continuously for a long time (Gardner, 2015; Limayem et al., 2001). The isolation levels vary with different Covid periods (distancing, quarantine, new normal), so it is not conditioned enough to change the old habits completely. Probably, during distancing or quarantine time, the old Habit of buying at physical store must be replaced with “new” online buying behavior; but the new normal condition after Covid drags customer back to real store. Perhaps, moderating effect is “weakened”. All are presented in the Figure 3:

Figure 3. Research results.

Finally, consider the predictive performance. We used PSLpredict, with a set of 10 k-folds and 10 repetitions, and obtained the results as shown in the Table 8 below:

Table 8. Prediction performance test.

	PLS	LM	Difference	PLS	LM	Blindfolding
	RMSE	MAE		Q^2	Q^2	Q^2
SI1	1.240	1.241	−0.001	0.274	0.263	0.376 (SI)
SI2	1.317	1.324	−0.007	0.224	0.199
SI3	1.278	1.281	−0.003	0.279	0.267

According to the rule of Shmueli et al. (2019), difference shows that the difference between RMSE of PSL and that of LM is all negative, in other words, the RMSEs of PLS are all smaller than RMSEs of LM. This means that all indicators are predicted well. Combined with PLS and LM’s Q² >0 (Shmueli et al., 2019) and the Blindfolding technique to obtain Q² = 0.376 > 0.35 (Hair et al., 2021), it can be concluded that the model has good predictive capabilities.

5. Discussion and conclusion

5.1. Discussion

Firstly, from the perspective of the PPM theory, Push factors and Pull factors positively influence the Switching intention to buying in online channel while the intention is negatively influenced by Mooring factors. This corroborates with research findings in previous literature (Chen & Keng, 2019; Lai & Wang, 2015; Lester et al., 2006; Ye & Potter, 2011b). While most previous literature about the temporal intention of/behavior of switching and partial switching is limited to preliminary theoretical base (Nimako, 2012), this research has extended the PPM theory by proposing a systematic research framework and empirically investigating factors that can influence the intention of switching buying channels.

Secondly, Perceived risks negatively influences Attitude towards switching, which is the strongest predictor of the switching intention among actors studied. In other words, Perceived risks partially “draw back” Attitude. These relationships are also evidenced by other researchers (Madahi & Sukati, 2014; Primabudi & Samopa, 2017; Wani & Malik, 2013b). Besides, this research has contributed to literature by integrating psychological factors to develop a new variant model of PPM-H-A to explain the hesitation of consumers when considering switching to buying consumer electronic goods online.

Thirdly, the Habit of buying at physical stores is also found to lessen the switching intention. This is in line with research findings of multiple researchers (Agag & El-Masry, 2016; Chen & Keng, 2019; Tuu, 2015a). Besides, the time span of the research survey in which consumers were expected to adapt to the new safe condition was only in a three-month period from 01/2022 to 06/2022. Therefore, it is not statistically significant enough to reach a conclusion about the moderating effect of habit on the relationship between Attitude and Switching intention. Hence, it is proposed by this research that changes in the buying habits of consumers throughout the Covid pandemic should be further investigated to avoid bias. The process of digital transformation is continuing, and this research framework can also be applied to evaluate changes in Habit and psychology of consumers in the future.

5.2. Limitations

Future research could be conducted when the post-Covid influences are no longer serious, which can be used to compare the impacts of different variables in the research framework, thereby attaining a deeper understanding of consumer psychology.

In the context of digital transformation, which is increasing significantly, many shops or distribution channels of businesses are being restructured. Multi-channels are still being applied in Vietnam for consumers. This is supported by many researchers who claim that multi-channels are appropriate in which social distancing is not forced (Dong, 2022; Liu et al., 2021; Shi et al., 2020). Thus, consumers’ perception of multi-channel distribution will not be misunderstood, and research results would not be biased. Therefore, because of multiple Covid periods of social distancing in Vietnam in the period between 2020 and the beginning of 2022, it was impossible to conduct a survey on multi-channel. This research does not conflict with the multi-channel model but could support future multi-channel research surveys to identify the rate of “digitalized” buying. The relationship between Risks—Attitude—Habit will always change to adapt to contexts.

Furthermore, future research may also consider investigating the Switching intention-behavior gap when electronic business is booming (Hino, 2017). However, optimising the benefits of multi-channels for consumers should be taken into consideration.

5.3. Conclusion

From a theoretical perspective, the study has a different approach than the commonTPB and TAM models applied by previous studies (Croome, 2006; Pookulangara et al., 2011). This research assess the positive impact of offline buying limitation and online buying advantage, as well as the negative impact of online buying limitations on Switching intention. This is the first contribution.

Research also empirically test the proposed variant model (PPM-H-A), compared with the previous PPM paradigm, for the intention to switch buying consumer electronics from offline to online.

The role of psychological factors (Perceived risk, Attitude, Habit) and the relationship among these variables is another contribution. The results prove that: Perceived risk dilutes the Attitude towards switching, and the Habit also diminishes Switching intention.

Regarding practical perspective, the research results imply recommendations for business to minimize the factors belonging to the “Mooring” group, as well as take advantage of the elements of the “Push”, “Pull” group by promoting communication, online buying convenience and adapting digital transformation progress in the future.

Traders can consider to reduce the Risks of information, delivery, product, after-sales services; thereby encouraging the Attitude towards switching.

Entrepreneurs can also propose strategies to form new habit of customers when shopping with new technologies such as virtual reality, online interactive sales, online promotion, mobile commerce marketing, meaning to gradually change the old habit.

Acknowledgments

This research paper is funded by National Economics University, Hanoi, Vietnam.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

The work was supported by the National Economics University, Hanoi, Vietnam .

Notes on contributors

Phan Duy Hung

Phan Duy Hung is a Lecturer at Electric Power University, Hanoi, Vietnam. He completed his Ph.D in Business Administration from National Economics University, Vietnam. His research interests include e-business, distribution channels, consumer behaviors and AI-application marketing. Other things: Classical arts, role of the Catholic Church in modern political history, George Orwell’s visionary novels are his curiosity.

Vu Huy Thong

Assoc. Prof., Dr. Vu Huy Thong, Dean, Faculty of Marketing, National Economics University (NEU), Hanoi, Vietnam. He received MBA from Boise State University, Idaho, USA (1995) and PhD from NEU (2004). He has been participating in series of research programs and consultancy projects for organizations, companies, ministries and local authorities. He published a number of research papers, articles and textbooks on marketing and business administration topics.

Appendix

1) Pilot study to assess measurement quality

Push

Table

Item <—Latent variable	Factor 1 (EFA)	Factor 2 (EFA)	Loading (CFA)	CR (>0.7)	AVE (>0.5)
DIS1<—Disadvantages	0.825		0.724	0.894	0.685
DIS2<—Disadvantages	0.823		0.818
DIS3<—Disadvantages	0.701		0.688
DIS4<—Disadvantages	0.741		0.635
VE1<—Vicarious Experience		0.887	0.908	0.809	0.517
VE2<—Vicarious Experience		0.909	0.919
VE3<—Vicarious Experience		0.875	0.860
VE4<—Vicarious Experience		0.696	0.578
KMO = 0.841, Bartlett = 1218.176 (sig 0.00); Total variance = 69.698%, Eigen values 2 components
Fornell-Larcker
	Vicarious Experience
Disadvantages
Vicarious Experience	0.82806
Disadvantages	0.474	0.71935

Pull

Table

Item <—Latent variable	Factor 1 (EFA)	Factor 2 (EFA)	Factor 3 (EFA)	Loading (CFA)	CR (>0.7)	AVE (>0.5)
PEOU1<—PEOU	0.770			0.614	0.839	0.570
PEOU2<—PEOU	0.779			0.692
PEOU3<—PEOU	0.782			0.885
PEOU4<—PEOU	0.790			0.803
Pr1<—Price		0.827		0.768	0.849	0.584
Pr2<—Price		0.700		0.735
Pr3<—Price		0.757		0.762
Pr4<—Price		0.790		0.792
CO1<—Convenience			0.682	0.743	0.912	0.601
CO2<—Convenience			0.691	0.658
CO3<—Convenience			0.760	0.795
CO4<—Convenience			0.735	0.750
CO5<—Convenience			0.776	0.797
CO6<—Convenience			0.772	0.841
CO7<—Convenience			0.830	0.827
KMO = 0.915, Bartlett = 2599.009 (sig 0.00); Total variance = 68.212%, Eigen values 3 components
Standard Fornell-Larcker
	Price	PEOU	Conveni-ence
Price	0.76452
PEOU	0.553	0.75563
Convenience	0.626	0.628	0.77515

Table

Item <—Latent variables	Factor 1 (EFA)	Factor 2 (EFA)	Factor 3 (EFA)	Loading (CFA)	CR (>0.7)	AVE (>0.5)
SN1<—SN	0.815			0.833	0.901	0.504
SN2<—SN	0.783			0.857
SN3<—SN	0.725			0.781
SN4<—SN	0.719			0.640
PRON1<—Perceived risks		0.808		0.781	0.861	0.612
PRON2<—Perceived risks		0.633		0.628
PRON3<—Perceived risks		0.734		0.687
PRON4<—Perceived risks		0.629		0.642
PRON5<—Perceived risks		0.802		0.788
PRON6<—Perceived risks		0.779		0.774
PRON7<—Perceived risks		0.623		0.693
PRON8<—Perceived risks		0.685		0.694
PRON9<—Perceived risks		0.652		0.687
PsE1<—PhysicalStoreEXP			0.637	0.743	0.901	0.566
PsE2<—PhysicalStoreEXP			0.764	0.782
PsE3<—PhysicalStoreEXP			0.694	0.70
PsE4<—PhysicalStoreEXP			0.766	0.763
PsE5<—PhysicalStoreEXP			0.769	0.732
PsE6<—PhysicalStoreEXP			0.759	0.732
PsE7<—PhysicalStoreEXP			0.820	0.812
KMO = 0922, Bartlett = 3438.71 (sig 0.00); Total variance = 62.627%, Eigen values 3 components
Fornell-Larcker
	Perceived Risks	Subjective Norms
PhysicalStoreExp
Perceived Risks	0.78229
Subjective Norms	0.606	0.71043
PhysicalStoreExp	0.535	0.618	0.75278

Table

Item <—Latent variable	Loading (EFA)	Loading (CFA)	CR (>0.7)	AVE (>0.5)
AT1<—Attitude	0.842	0.842	0.867	0.62
AT2<—Attitude	0.817	0.833
AT3<—Attitude	0.791	0.768
AT4<—Attitude	0.818	0.702
HA1<—Habit	0.882	0.850	0.885	0.659
HA2<—Habit	0.856	0.796
HA3<—Habit	0.849	0.790
HA4<—Habit	0.861	0.810
SI1<—Switching intention	0.827	0.768	0.798	0.569
SI2<—Switching intention	0.773	0.716
SI3<—Switching intention	0.825	0.778

2) PLS SEM Validation

Table

Perspective	Validation	Thresholds
Reflective construct	- Internal consistency (Cronbach’s alpha, CR) - Convergent validity (AVE) - Discriminant validity (Fornell-Larcker)	- Cronbach’s alpha > 0.6; 0.6 < CR < 0.95 - AVE > 0.5 - $\sqrt{\mathrm{A}\mathrm{V}\mathrm{E}}$> correlation coefficient between a latent variable and other
Formative construct	- Convergent validity (AVE, w) - Colinearity between indicators (VIF) - Significant and relevance of outer weight (p-value and hypothesis)	- AVE > 0.5, w > 0.8 (path between formative and reflective constructs of the same latent variable) - VIF < 5 - p-value < 0.05 and reasonable with hypothesis
Inner model (among latent variables)	- Colinearity between variables - P regression path coefficient - Determination coefficient	- VIF < 5 - P suitable to hypothesis and p-value <0.05 - 0 < R^2 < 1
Predictivity	- Apply the blindfolding and PLSpredict	- Q^2: weak (<0.02), moderate (>0.15 and <0.35), strong (>0.35); RMSE(or MAE) of PLS-SEM < those of LM

3) Common method bias

Table

	VIF RandomVar
AttitudeVar	1.498
ConvenienceVar	2.845
DisadvantagesVar	1.373
HabitVar	1.155
Perceived RiskVar	1.795
Perceived ease of useVar	2.648
Physical store ExperienceVar	2.120
PriceVar	2.474
RandomVar
Subjective NormsVar	2.024
Switching intentionVar	1.871
Vicarious ExperienceVar	1.484

4) Questionaire

Part 1. Demographic information

Gender□ Male  □ FemaleAge □ Under 18  □ 18–23  □ 23–30  □ 30–40  □ Over 40

Profession□ Student  □ Staff  □ Start-up  □ Others

Place□ Hanoi  □ Da Nang  □ Ho Chi Minh city

For recent 24 months, How many times have you bought electronics online?□ 0  □ 1–2  □ 3–4  □ More than 5 times

Which of the following electronic goods do you buy the most online?□ Smartphone         □ Television □ Laptop      □ Sound (speakers, mp3 player, headphones …)□ Desktop         □ Visual (camera)□ Game console  □ Appliance (air conditioner, oil-free fryer, washing, refrigerator …)□ Other (please write down): … … … … … … … … … … … … … … … … … … …

You most often buy online through□ Pure online retailer (non physical store)□ Click and Mortar (eg: Đien May Xanh, PICO, Nguyen Kim …)□ Facebook

Spending for online purchases of electronic goods/2 years (average exchange: 1 USD = 23,165.5 VND)□ Under 21.6 USD (under 0.5 million VND)   □ 216–432 USD □ 21.6–43.2 USD                □ 432–1296 USD □ 43.2–129.6 USD                   □ 1296–2160 USD □ 129.6–216 USD (3–5 million VND)      □ More than 2160 USD (50 million VND)Covid phase (depending on the Covid timeline in Vietnam): … … … … .

5) Customer response (using Likert 1–7: totally disagree – totally agree)

Table

Code	Items	Source
Push – Disadvantages (store buying channels)
DIS1	Information at physical stores does not always meet my requirements	Lester et al. (2006); Gupta (2015); Sultan and Uddin (2011); Verhoef et al. (2007)
DIS2	There are fewer options at physical stores than through online channels	Santana and Loureiro (2010); Harn et al. (2006)
DIS3	I spend more time buying products at physical stores	Lester et al. (2006); Santana and Loureiro (2010); Gupta (2015)
DIS4	I am hesitant if I don’t buy	Santana and Loureiro (2010)
Pull – Vicarious experience (of other people)
VE1	My colleagues buy electronics consumer goods online successfully	Chiu et al. (2011)
VE2	My friends buy electronics consumer goods online successfully
VE3	My family members buy electronics consumer goods online successfully
VE4	My acquaintances buy electronics consumer goods online successfully
Pull – Price (online channels)
Pr1	Price in online channels is more reasonable than at physical stores	Swinyard and Smith (2003); Lester et al. (2006); Harn et al. (2006); Gupta (2015); Elida et al. (2019); Elida et al. (2019)
Pr2	I can compare prices through many online channels	Lester et al. (2006); Rahman et al. (2018);
Pr3	Online channels offer many promotional offers	Verhoef et al. (2007); Gupta (2015)
Pr4	Buying products online saves my money	Harn et al. (2006); Chang (2003); Elida et al. (2019); Yeo et al. (2017)
Pull – Convenience (online channel)
CO1	I can find good information about the products online	Forsythe et al. (2006); Santana and Loureiro (2010); Jiang et al. (2013); Duarte et al. (2018)
CO2	I can easily compare products and sellers	Santana and Loureiro (2010); Sultan and Uddin (2011); Duarte et al. (2018)
CO3	I have more alternatives of electronics consumer goods to make a choice thanks to searching online	Forsythe et al. (2006); Santana and Loureiro (2010); Swinyard and Smith (2003); Lester et al. (2006); Jiang et al. (2013)
CO4	I don’t need to take much effort (leaving home, travelling, queuing, waiting to be served …) when buying in online channels	Forsythe et al. (2006); Santana and Loureiro (2010); Rahman et al. (2018); Jiang et al. (2013)
CO5	Buying products through online channels saves my time.	Santana and Loureiro (2010); Sultan and Uddin (2011); Harn et al. (2006)
CO6	I can order products online any time.	Forsythe et al. (2006); Santana and Loureiro (2010); Sultan and Uddin (2011); Rahman et al. (2018); Lester et al. (2006); Uzan (2014); Jiang et al. (2013); Duarte et al. (2018)
CO7	I can track the order easily (via email, tracking system …)	Proposed by the authors, based on the research of Rao et al. (2011)
Pull - Perceived ease of use
PEU1	I can easily learn to buy products through online channels although this is the first time	Ho and Chen (2014), Perea et al. (2004), Lin (2007), Ye and Potter (2011a); Bauerová and Klepek (2018)
PEU2	It is easy for me to learn to buy products online
PEU3	I can easily buy products online
PEU4	For me, buying through online channels is easy to perform
Mooring - Perceived risks (online channels)
PRON1	I am concerned about the origin of products	Proposed by the authors
PRON2	My personal information may not be protected (hacked, or provided for other partners)	Forsythe et al. (2006); Santana and Loureiro (2010); Rahman et al. (2018); Wani and Malik (2013b)
PRON3	I may not receive the order I make through online channels	Wani and Malik (2013b); Nguyễn Hoàng Diễm Hương et al. (2016); Hà Ngọc Thắng (2015)
PRON4	I may have to wait for the online delivery longer than expected	Forsythe et al. (2006); Lester et al. (2006); Corbitt et al. (2003)
PRON5	I believe that buying products through online channels has many risks because I cannot touch and feel the product	Forsythe et al. (2006); Lester et al. (2006); Swinyard and Smith (2003); Campo and Breugelmans (2015); Hà Ngọc Thắng (2015); Haridasan et al. (2021)
PRON6	I believe that buying products through online channels has many risks because I cannot try on the products	Forsythe et al. (2006); Lester et al. (2006); Santana and Loureiro (2010); Wani and Malik (2013b); Swinyard and Smith (2003); Nguyễn Hoàng Diễm Hương and cộng sự (2016); Hà Ngọc Thắng (2015); Haridasan et al. (2021)
PRON7	I don’t buy products through online channels if there is no refund policy	Javadi et al. (2012); Lester et al. (2006); Swinyard and Smith (2003); Nguyễn Hoàng Diễm Hương and cộng sự (2016); Verhoef et al. (2007)
PRON8	I feel it is difficult to solve conflicts when buying through online channels
PRON9	I feel worried when buying electronics consumer goods through online channels	Marriott and Williams (2018); Corbitt et al. (2003)
Mooring - Subjective norms
SN1	My influencers think that electronics consumer goods should be bought at physical stores	Bansal et al. (2005), Lian and Yen (2013), Zheng (2013) Ho and Chen (2014), Ye and Potter (2011a)
SN2	People who are important for me encourage me to buy electronics consumer goods at physical stores
SN3	People that I know encourage me to buy electronics consumer goods at physical stores
SN4	Generally, electronics consumer goods should be bought at physical stores
Mooring – (Lack of) physical store experience
PsE1	I find buying at physical stores interesting	Bustamante and Rubio (2017), Terblanche (2018)
PsE2	I find buying at physical stores relaxing
PsE3	I have a positive feeling when I buy at physical stores.
PsE4	Physical stores have attractive product display
PsE5	I can interact and communicate with other consumers at physical stores
PsE6	I need consultation from salespeople
PsE7	I can test products at physical stores	Swinyard and Smith (2003), Rahman and Mannan (2018)
Habit (buying electronics consumer goods)
HA1	I automatically buy at physical stores	Lai (2015); Sun et al. (2017); Cheng et al. (2019); Lin and Wang (2017); Xu et al. (2017); Ye and Potter (2011a)
HA2	Buying at physical stores is obvious to me
HA3	When I need to buy electronics consumer goods, physical stores are my obvious choice
HA4	I choose to buy at physical stores unconsciously
Attitude (toward switching to online channels
AT1	Switching to online channels is a _______ idea (scaling from bad to good)	Bansal et al. (2005), Madahi and Sukati (2014)
AT2	Switching to online channels is a _______ idea (scaling from harmful to beneficial)
AT3	Switching to online channels is a _______ idea (scaling from useless to useful)
AT4	Buying through online channels is a more attractive alternative than buying through traditional channels
Switching intention (from traditional channels to online channels)
SI1	I consider switching to buy electronics consumer goods through online channels	Lai (2015); Chen and Keng (2019); Cheng et al. (2019)
SI2	The possibility that I switch to buy electronics consumer goods in online channels is ______ (scaling from unlikely to likely)	Bansal et al. (2005); Lai (2015); Chen and Keng (2019); Cheng et al. (2019)
SI3	I will switch to buy electronics consumer goods online in the near future	Chen and Keng (2019); Zhou (2015)