Factors Influencing Smartphone Adoption: A Study in the Indian Bottom of the Pyramid Context

Abstract

The usage of smartphones is increasing with each passing day. The growth of wireless subscription in India in the last 5 years is very high. However, the penetration rate of smartphones among low-income people is significantly low. Looking at the near saturation and high competition of markets at the higher side of the economic pyramid, companies are trying to explore the segment of low-income people termed ‘Bottom of the Pyramid’ (BOP). In order to explore the BOP market for smartphones, one needs to realize the factors influencing the adoption of smartphones at the BOP. Our study explores these factors with reference to the existing literature of technology adoption. A theoretical framework is proposed and tested with 266 valid data points. Structural equation modelling (SEM) is used to test the proposed framework. The empirical analysis revealed that ‘performance expectancy’, ‘effort expectancy’ and ‘perceived monetary value’ have a positive impact on the ‘behavioural intention’ of using smartphones at BOP. ‘Smartphone anxiety’ and ‘smartphone self-efficacy’ are found to have an impact on ‘effort expectancy’.

Keywords

Smartphone adoption at BOP performance expectancy effort expectancy anxiety self-efficacy perceived monetary value

Introduction

With the advent of telephony, the mobile phone has become one of the most used technological devices across the globe. In India, there are 1170.18 million wireless telephone subscribers (Telecom Regulatory Authority of India, 2017). However, there is a significant difference in the penetration rates of mobile subscriptions in rural and urban India (Gupta & Jain, 2015). The wireless teledensity in urban India is found to be 167.97 while it is just 57.31 in rural India (Telecom Regulatory Authority of India, 2017).

The proliferation of smartphones has helped in reducing the digital divide (Sung, 2016) and social isolation (Cho, 2015). The lower strata of the economic pyramid have a stronger feel for social isolation (Prahlad, 2004) as compared to the upper strata, and therefore, they tend to increase consumption of newer products and technologies in order to reduce this alienation (Alwitt, 1995; Hill & Stephens, 1997). With the growing popularity and reducing price, the smartphone has the potential to be penetrated into the segment of low-income people termed ‘Bottom of the Pyramid (BOP)’ (Prahlad, 2004) as this type of technology-based consumption may help them satisfy their aspirational needs (Gupta & Srivastav, 2016). Therefore, it is interesting to explore the factors influencing the adoption of smartphones at the BOP. BOP represents the economically weaker segment of the world’s population with a daily per capita income of US$2 or less and they are characterized by low literacy, poor health conditions, limited access to media, striving to meet basic needs and geographical isolation (Prahlad, 2004). In India, 29.5 per cent of the total population falls in extreme poverty line with a daily income of US$2.44 (in terms of purchasing power parity or PPP) or less (Planning Commission, 2014). For this research, we have set the upper threshold value for BOP at ₹13,152, considering five members per family with a living expense of US$5 per person per day (Heeks, 2012) with a purchasing power parity value of 17.536 (OECD, 2017). While setting this upper threshold value of income to qualify for the BOP, we have considered Prahlad’s proposition of US$2 per person per day to be too outdated. On the other hand, the upper cut-off per capita value of US$3,000 or more per year (Hart & London, 2011; Yurdakul, Atik, & Dholakia, 2017) is considered to be too high for a developing country.

In order to understand the adoption of smartphones at the BOP, our study initially explores the important factors of technology adoption to build a theoretical framework and then proceeds ahead to test the framework empirically. The following sections provide the review of literature, objectives, rationale of the study, theoretical framework, methodology, analysis, discussion, conclusion, managerial implications and limitations of the study.

Review of Literature

Adoption of technology is dependent on multiple factors. Adoption refers to the user’s individual decision-making process regarding the purchase and use of new commodities (Gupta & Jain, 2014). Previous research considered ‘perceived usefulness’ and ‘ease of use’ as key factors of technology adoption (Davis, 1986, 1989; Gangwar & Date, 2016; Venkatesh, Morris, Davis, & Davis, 2003). Venkatesh et al. (2003) used terms ‘performance expectancy’ (PE) and ‘effort expectancy’ (EE) for ‘perceived usefulness’ and ‘ease of use’, respectively, in the Unified Theory of Acceptance and Use of Technology (UTAUT). PE has been formally defined as the degree to which a person believes that using the system or technology will help him or her achieve gains in job performance (Venkatesh et al., 2003). Multiple constructs from previous studies pertain to PE including ‘perceived usefulness’ (Davis, 1986), ‘extrinsic motivation’ (Davis, Bagozzi, & Warshaw, 1992), ‘job fit’ (Thompson, Higgins, & Howell, 1991), ‘relative advantage’ (Moore & Benbasat, 1991) and ‘outcome expectations’ (Compeau, Higgins, & Huff, 1999). EE has been defined as the degree of ease associated with the use of the system or technology (Venkatesh et al., 2003). Venkatesh et al. (2003) adopted EE from the Technology Acceptance Model (TAM) where it was termed ‘perceived ease of use’. The usage of mobile services can be increased by making it easy to use and learn (Kim, Mirusmonov, & Lee, 2010). ‘Perceived ease of use’ is conceptualized as an essential factor in various technology adoption researches encompassing a variety of domains (Kim et al., 2010; Liébana-Cabanillas, Sánchez-Fernández, & Muñoz-Leiva, 2014).

Various studies have found that adoption is closely associated with intention to use (Carter & Belanger, 2005). A positive behavioural intention to use a system or technology leads to actual adoption. Warshaw and Davis (1985) formally defined ‘behavioural intention’ (BI) as the degree to which an individual has framed conscious plans to execute or refuse a future behaviour. There is a strong belief in the social cognitive theory that human ‘BI’ is largely determined by both anxiety and self-efficacy (Bandura, 1986). Anxiety has been defined as evoking anxious or emotional reactions when it comes to performing a behaviour (Bandura, 1986; Venkatesh et al., 2003).

‘Self-efficacy’ has been defined as individuals’ beliefs about their capabilities of managing their behaviours to produce desired outcomes. This construct was formally introduced by Bandura (1977). Paglis (2010) reported that self-efficacy has been greatly emphasized in recent times. In a study, Malliari, Korobili, and Togia (2012) viewed self-efficacy as one’s expectations and confidence of achieving something in a specific state of affairs. Self-efficacy has been introduced to the studies of Management Information Systems (MIS) as a significant research construct. For instance, Simmering, Posey, and Piccoli (2009) established that self-efficacy has a positive impact on learning from online courses. Similarly, Johnson, Hornik, and Salas (2008) found that self-efficacy is positively related to the perceived effectiveness of an e-learning system. Chien (2012) showed that computer self-efficacy moderates the impact of system functionality on training effectiveness.

Actions of the people belonging to the BOP segment are always impacted by financial constraints. This financial challenge compels them to look for greater ‘value for money’. They tend to search for better quality products at affordable prices (Prahlad, 2004; Rahman, Hassan, & Floyd, 2013). Perception of monetary value plays an intensive role especially for the BOP in the purchasing decision-making process as this segment of people expects a better ‘perceived monetary value’ (PMV). Kim, Park, and Oh (2008) formally defined PMV as the degree to which an individual perceives the appropriateness of the cost in relation to one’s perceived benefits and preference of the service.

Technology adoption has been explored both theoretically and empirically in multiple contexts. Most of the empirical studies of technology adoption are carried in developed countries (Hasan, Lowe, & Petrovici, 2017). Studies related to technology adoption need further attention in developing countries (Rahayu & Day, 2015; Riffai, Grant, & Edgar, 2012; Sharma & Govindaluri, 2014; Tarhini, Hone, & Liu, 2014). Scholars are advocating for the need for studies concerning the adoption of advanced mobile-based services at the BOP (Hossain & Jamil, 2015; Kansal, 2016). In particular, there is no prominent study in our knowledge which explicitly explores the adoption of smartphones at the BOP. Therefore, our study tries to fulfil this gap by exploring and empirically testing the factors influencing the adoption of smartphones at the BOP in the Indian context.

Objectives

Our study has two primary objectives. The first objective is to build a theoretical framework for smartphone adoption at the BOP. This objective is achieved by identifying the important constructs of technology adoption in the context of BOP and then proposing logical and literature-based relationships among these constructs. The subsequent objective of our study is to empirically assess our theoretical framework in the Indian context. This objective is achieved by using statistical techniques on the data collected from the BOP respondents in India.

Rationale of the Study

Technology adoption is contextual in nature (Downs & Mohr, 1976; Jelinek, Ahearne, Mathieu, & Schillewaert, 2006; Kimberly & Evanisko, 1981). Adoption behaviour at the BOP may significantly differ from the rest of the population due to the existence of differences in socio-economic conditions (Jaiswal, 2015; Karnani, 2009; Prahlad, 2004). For instance, Gupta and Srivastav (2016) stated that BOP people may indulge in the consumption of aspirational products such as a smartphone in order to satisfy the assimilationist behaviour. Dey, Binsardi, Prendergast, and Saren (2013) questioned the applicability of dominant technology adoption models in mobile phone adoption at the BOP because of the presence of financial constraints and lack of experts’ help to overcome technical difficulties.

Researchers have recommended the use of information and communication technologies (ICTs) for improving service delivery to the BOP (e.g., Berger & Nakata, 2013; Tarafdar, Anekal, & Singh, 2012). Smartphones, being a relatively newer ICT-based device with growing popularity (Meeker, 2015) and reducing prices, may have the potential to act as a catalyst for improving service delivery to the BOP segment.

When the fact of the contextual nature of technology adoption is coupled with the potentiality of smartphone usage at the BOP, it becomes meaningful to explore the factors influencing the adoption of smartphones at the BOP segment.

Theoretical Framework

In this section, we will propose the theoretical research framework along with research propositions. The framework considers ‘PE’, ‘EE’, ‘smartphone anxiety’ (SA), ‘smartphone self-efficacy’ (SSE) and ‘PMV’ as antecedents to ‘BI’. It also checks if the effect of SA and SSE on BI is mediated by EE. Though the framework shows that BI leads to ‘use behaviour’, this relationship is not tested in our study as it has been universally accepted by all the popular studies (Davis, 1986, 1989; Venkatesh et al., 2003).

Antecedents to BI

Most of the prominent studies in technology adoption have considered ‘PE’ and ‘EE’ as direct determinants of BI (Davis, 1986, 1989; Venkatesh et al., 2003). When a person considers that the use of smartphones will boost his/her performance in day-to-day life, interest for using it is likely to increase. Using this logic, it can be proposed that ‘PE’ is positively related to ‘BI’ to use a smartphone at the BOP. Though the relationship between these two constructs is well established in earlier literature (e.g., Gupta, Dasgupta, & Gupta, 2008; Lallmahomed, Lallmahomed, & Lallmahomedc, 2017; Pynoo et al., 2011) in a variety of contexts, some studies could not establish any significant relationship between them (e.g., Dwivedi et al., 2017; Lee, Park, Cho, & Jin, 2018). The reason behind these adverse results may be the presence of context-specific conditions. As the context of BOP consumers is quite different from other consumers, the magnitude and effect of ‘PE’ may significantly differ. Therefore, we intend to test this relationship between ‘PE’ and ‘BI’ to use a smartphone at the BOP.

Similarly, the self-perception of an individual about the ease of using a smartphone is likely to impact its adoption. An easy-to-use and the user-friendly smartphone will favour its adoption. The perception of less effort requirement may boost the intention of an individual to use a smartphone. Sanakulov and Karjaluoto (2015) carried out a detailed literature review on the empirical studies of mobile technology adoption and found that 63 per cent of the studies supported the relationship between ‘EE’ and ‘BI’. This proposition is empirically established in many dominant studies (e.g., Davis, 1989; Gupta et al., 2008; Venkatesh et al., 2003). On the other hand, some articles could not establish the significance of the proposed relationship between ‘EE’ and ‘BI’ (e.g., Lallmahomed et al., 2017; Pynoo et al., 2011). Pynoo et al. (2011) argued that the effect of ‘EE’ is often inferior to that of ‘PE’ in the case of professional users. However, the users or potential users of smartphones at the BOP are very less likely to be an expert professional in using smartphones. Therefore, we propose that ‘EE’ will have a significant impact on the ‘BI’ to use a smartphone at the BOP.

Negative emotions such as anxiety can be assumed as resistance to technology adoption as they create depression, divert attention and discourage people to use technology, which takes a long time to master the usage (Booker, Detlor, & Serenko, 2012). These negative psychological traits primarily decrease users’ ‘perceived ease of use’ (Venkatesh, 2000). The BOP segment is more impacted by anxiety for a relatively sophisticated entity as this segment is characterized by a low literacy rate. Gutiérrez and Gamboa (2010) stated that less educated people are more prone to anxiety while using mobile handsets. Some medical and psychological studies have already established that socio-economic conditions are highly correlated to anxiety and people in the lower socio-economic strata are likely to be more affected by anxiety (Ansseau et al., 2007; Mwinyi et al., 2017). Anxiety acts as a catalyst in the manifestation of behaviour (Bozionelos, 2004). Bandura (1986) considered ‘anxiety’ as a predictor of BI. Multiple studies have empirically established the direct relationship between ‘anxiety’ and ‘BI’ (Adetimirin, 2015; Alenezi, Karim, & Veloo, 2010; Elasmar & Carter, 1996; Tung & Chang, 2008). Nagar and Gandotra (2016) explored this relationship in the context of Internet shopping and established that ‘anxiety’ has a negative impact on ‘patronage intention’. However, Venkatesh et al. (2003) found the relationship between ‘computer anxiety’ and ‘BI’ to be non-significant. Venkatesh and Bala (2008) found that ‘perceived ease of use’ mediates the effect of ‘computer anxiety’ on BI. From the previous literature, it is evident that the direct relationship between ‘anxiety’ and ‘BI’ is inconclusive. As the impact of anxiety is likely to be more in the BOP segment due to the higher level of frustration elevated from limited income, we intend to test the relationship between ‘anxiety’ and ‘BI’ to use a smartphone at the BOP. As our study is related to smartphones, the term ‘SA’ has been used in place of ‘anxiety’ and we propose that ‘SA’ is negatively related to ‘BI’ to use it at the BOP.

‘Self-efficacy’ is a concept of self-perception about self-capabilities. It controls individuals’ psychological adjustments and behaviours (Maddux, 1995). A person having high self-efficacy is likely to be open to accept challenges and put more effort in solving a problem (Bandura, 1986). Bandura (1986) considered ‘self-efficacy’ as a predictor of BI. The direct relationship of ‘self-efficacy’ and ‘BI’ is empirically evident in earlier research (Adetimirin, 2015; Alenezi et al., 2010; Chen, Lin, Yeh, & Lou, 2013; Tung & Chang, 2008). At the same time, some literature opposed the direct relationship between these two constructs and claimed that ‘perceived ease of use’ mediates the effect of ‘self-efficacy’ on ‘BI’ (Venkatesh & Bala, 2008; Venkatesh et al., 2003). As this study is related to smartphones, the term ‘SSE’ has been used in place of ‘self-efficacy’. It is evident from existing literature that the direct relationship between ‘self-efficacy’ and ‘BI’ is inconclusive across studies and contexts. Therefore, we strongly feel the need for empirical testing of the effect of ‘SSE’ on ‘BI’ to use it in the BOP context.

The most dominant problem faced by the BOP is the possession of limited disposable income (Prahlad, 2004). Therefore, ‘PMV’ is likely to play a more intensive impact in case of the BOP segment as compared to other segments due to the presence of higher price sensitivity and less affordability in this segment. Dood, Monroe, and Grewal (1991) established a positive relationship between ‘perceived value’ and ‘willingness to buy’. Kim et al. (2008) found that PMV is a direct predictor of ‘continued intention to use’. Kang and Maity (2012) challenged the application of the technology acceptance model at BOP and introduced PMV in the adoption process. As the BOP people are always constrained by the limited income, PMV may have a strong effect on BI of using a smartphone. Therefore, we may propose that ‘PMV’ is positively related to ‘BI’ to use a smartphone at the BOP.

Antecedents to EE

The Theory of Reasoned Action (Fishbein & Ajzen, 1975) and Technology Acceptance Model (Davis, 1986) suggested that the impact of any external variable on ‘BI’ is mediated by key predictors (‘perceived usefulness’ and ‘perceived ease of use’). The negative impact of anxiety on ‘perceived ease of use’ was implicated in classical theories (Phillips, Martin, & Meyers, 1972). Venkatesh (2000) empirically showed that anxiety is a predictor of ‘perceived ease of use’. Bandura (1986) implicated that anxiety is a reciprocal construct to expectancies such as ‘self-efficacy’ and ‘perceived ease of use’. Abdullah and Ward (2016) reviewed the literature on e-learning adoption and found that 59 per cent of the studies established the negative impact of ‘anxiety’ on ‘perceived ease of use’. This is a clear indication that the causal relationship between ‘anxiety’ and ‘EE’ is dependent on contextual factors. Empirical evidence for the negative impact of anxiety on ‘perceived ease of use’ is established in numerous studies (Venkatesh, 2000; Venkatesh & Bala, 2008). However, all the above-mentioned empirical studies are carried out in developed countries’ organizational contexts. Therefore, our study intends to test whether ‘SA’ is negatively related to ‘EE’ in the BOP context.

Similarly, there are multiple studies based on the variety of technologies establishing the positive influence of ‘self-efficacy’ on ‘perceived ease of use’ (Gu, Lee, & Suh, 2009; Ong, Lai, & Wang, 2004; Venkatesh, 2000; Venkatesh & Bala, 2008). Venkatesh (2000) empirically showed that ‘self-efficacy’ is conceptually distinct from ‘anxiety’ and it is a predictor of ‘perceived ease of use’. Self-efficacy of using smartphones may significantly vary at the BOP segment from other consumers due to lack of experts’ help. This reduction of self-efficacy may impact the perception of ease of using a smartphone, that is, ‘EE’. Therefore, we propose to test the causal relationship between ‘SSE’ and ‘EE’ of using a smartphone at the BOP. The theoretical framework is diagrammatically shown in Figure 1.

Methodology

We have used structural equation modelling (SEM) for empirical validation of the theoretical framework. A package called ‘Lavaan’ is used in R to execute SEM. We have calculated Cronbach’s alpha coefficient and composite reliability to confirm that the constructs are reliable. Factor loading scores are used to check the unidimensionality. Convergent and divergent validity of the measurement model is checked through average variance extracted (AVE). The goodness of fit for the model is checked through multiple indices including root mean square error of approximation (RMSEA), standardized root mean square residual (SRMR), chi-square (with regard to the degree of freedom), comparative fit index (CFI) and Tucker-Lewis index (TLI).

Figure 1.

Source: The authors.

A questionnaire was created from items validated from previously established research on technology adoption. Few measurement scales are added to fit the current context. A list of the measurement scale items of the constructs is given in Table 1 with their original source.

Table 1.

Constructs and Measurement Scale Items

Construct	Measurement Scale	Source
Performance expectancy (PE)	Using smartphone enables me to accomplish tasks more quickly (PE1)	Venkatesh et al. (2003)
	I believe that using smartphones would help me to communicate faster (PE2)	Contextual
	Smartphones would be a great device to access public services easily (PE3)	Contextual
	Using smartphones increases my productivity (PE4)	Venkatesh et al. (2003)
	I would find smartphones useful in life (PE5)	Venkatesh et al. (2003)
Effort expectancy (EE)	I would find smartphones easy to use (EE1)	Venkatesh et al. (2003)
	I am confident to learn any application installed in smartphones (EE2)	Venkatesh et al. (2003)
	My interaction with smartphones would be clear and understandable (EE3)	Venkatesh et al. (2003)
	I think smartphones are user friendly (EE4)	Contextual
	It would be easy for me to become skilful at using smartphones (EE5)	Venkatesh et al. (2003)
Smartphone anxiety (SA)	I feel apprehensive about using smartphones (SA1)	Venkatesh et al. (2003)
	I feel nervous to use smartphones (SA2)	Venkatesh and Bala (2008)
	Using smartphones is somewhat intimidating to me (SA3)	Venkatesh et al. (2003)
	I hesitate to use smartphones for fear of making mistakes I cannot correct (SA4)	Venkatesh et al. (2003)
	It scares me to think that I could lose a lot of information using smartphones by hitting the wrong key (SA5)	Venkatesh et al. (2003)
Smartphone self-efficacy (SSE)	I could complete a job or task using a smartphone if there was no one around to tell me what to do as I go (SSE1)	Venkatesh and Bala (2008)
	I have the ability to explore smartphone applications for easing my work (SSE2)	Contextual
	I could complete a job or task using smartphones if I could call someone for help if I get stuck (SSE3)	Venkatesh and Bala (2008)
	I could complete a job or task using smartphones if I had just the built-in help facility for assistance (SSE4)	Venkatesh and Bala (2008)
	I could complete a job or task using smartphones if I had a lot of time to complete the job for which the application was provided (SSE5)	Venkatesh and Bala (2008)
	I could complete a job or task using smartphone applications if I found the application interesting (SSE6)	Contextual
Construct	Measurement Scale	Source
Perceived monetary value (PMV)	I think smartphones are reasonably priced (PMV1)	Dood et al. (1991), Kang and Maity (2012), Kim et al. (2008)
Perceived monetary value (PMV)	I think smartphones offer value for money (PMV2)	Dood et al. (1991), Kang and Maity (2012), Kim et al. (2008)
Behavioural intention (BI)	I intend to use smartphones in future (BI1)	Venkatesh et al. (2003)
	I predict I would use smartphones in future (BI2)	Venkatesh et al. (2003)
	I plan to use smartphones in future (BI3)	Venkatesh et al. (2003)

Source: Measurement scale items based on literature and context.

Data

While selecting the respondents, we have put the criteria of household income less than ₹13,152 per month and a minimum age of 18 years. Data are collected from both rural and urban areas of Assam and Delhi, India. Maximum effort is made to cover the areas where people face challenges regarding disposable income, electricity, access to information and higher education.

The rule of thumb in SEM suggests a minimum sample size of 200 (Kline, 2005). Bentler and Chou (1987) mentioned a sample size of ten per measurement scale item is good enough in SEM. In our study, we have 23 valid measurement scale items which necessitate a sample size of 230. Each item is measured on a Likert scale of 1–5. We have administered the questionnaire to 278 respondents out of which 266 data points are found to be valid indicating a valid response rate of 95.68 per cent. A face-to-face method of data collection helped us in achieving a high rate of valid response.

Analysis

Empirical Model

We have tested the proposed model with 266 valid data points. Testing of the proposed model is done in two steps—measurement model testing and structural model testing. A measurement model confirms the reliability, convergent validity and discriminant validity of the constructs and associated measurement scale items. Our measurement model has rejected one item of PE (PE3), one item of SA (SA5) and two items of SSE (SSE2 and SSE4) as the factor loading of these items is found to be below 0.5. All the other measurement scale items are found to satisfy the conditions of reliability, convergent validity and discriminant validity. The structural model checks the specified relationships among the constructs. All the propositions except the third (‘SA’ is negatively related to ‘BI’) and fourth (‘SSE’ is positively related to ‘BI’) are found to have satisfactory ‘p’ values at 95 per cent confidence level.

The empirical model of our study can be analysed in two sections—the measurement model and structural model.

Measurement Model

The measurement model starts with checking the factor loading of each item. Hu and Bentler (1999) proposed a minimum factor loading of 0.5 for each item in order to fit the measurement model. All our scale items except PE3, SA5, SSE2 and SSE4 are found to possess a factor loading of more than 0.5. The next step in the measurement model is to check for reliability of the constructs. A Cronbach’s alpha value of 0.7 or more for the accepted measurement scale items of each construct confirms internal consistency (Churchill, 1979). Composite reliability of all the constructs is found to be higher than the minimum acceptable value of 0.7 (Fornell & Larcker, 1981; Hair, Black, Babin, & Anderson, 2006). Table 2 lists factor loading, Cronbach’s alpha and composite reliability values for the items and the constructs.

In order to confirm the validity of the constructs, we have used AVE of the accepted scale items. Fornell and Larcker (1981) set the minimum value of AVE at 0.5 in order to satisfy the condition of convergent validity. The criteria for confirming discriminant validity is that the square root of AVE of any construct should be greater than the correlation across constructs (Fornell & Larcker, 1981). Both the criteria of convergent and discriminant validity are well satisfied in our dataset. Table 3 displays the values of AVE, square root of AVE and correlation across constructs.

All the scores above are found to have a satisfactory value confirming the convergent and discriminant validity of the model.

To assess the goodness of fit for the measurement model, we have checked RMSEA, SRMR, chi-square (with regard to degree of freedom), CFI and Tucker-Lewis index (TLI). Table 4 shows the goodness of fit indices and their acceptable range.

Table 2.

Factor Loading, Cronbach’s Alpha and Composite Reliability

Construct	Item	Factor Loading	Internal Consistency (Cronbach’s Alpha)	Composite Reliability
Performance expectancy	PE1	0.840	0.87	0.88
	PE2	0.762
	PE3	0.320
	PE4	0.799
	PE5	0.817
Effort expectancy	EE1	0.779	0.88	0.88
	EE2	0.805
	EE3	0.765
	EE4	0.789
	EE5	0.777
Smartphone anxiety (SA)	SA1	0.838	0.86	0.86
	SA2	0.822
	SA3	0.742
	SA4	0.727
	SA5	0.300
Smartphone self-efficacy (SSE)	SSE1	0.762	0.83	0.83
	SSE2	0.210
	SSE3	0.767
	SSE4	0.137
	SSE5	0.719
	SSE6	0.721
Perceived monetary value (PMV)	PMV1	0.854	0.85	0.85
Perceived monetary value (PMV)	PMV2	0.872	0.85	0.85
Behavioural intention (BI)	BI1	0.794	0.87	0.87
	BI2	0.870
	BI3	0.846

Source: The authors.

Structural Model

Once we have found the measurement model to be satisfactory, we have checked the specified relationships among the constructs. The raw data file has been used as the input for this analysis. A ‘p’ value below 0.05 in the regression indicates support of the proposition at a 95 per cent confidence level. Our data support all the propositions except the third (‘SA’ is negatively related to ‘BI’) and fourth (‘SSE’ is positively related to ‘BI’) at a 95 per cent confidence level. Table 5 displays the estimate, standard error, ‘z’ value and ‘p’ value of the predictors.

Table 3.

Convergent and Discriminant Validity of Measurement Model (Diagonal Elements Are Square Roots of AVE of the Constructs, off Diagonal Elements Are Correlations Between the Constructs)

Construct	AVE	PE	EE	SA	SSE	PMV	BI
PE	0.64	0.80
EE	0.61	0.31	0.78
SA	0.61	−0.33	−0.58	0.78
SSE	0.55	0.43	0.59	−0.38	0.74
PMV	0.74	0.34	0.30	−0.37	0.35	0.86
BI	0.70	0.53	0.52	−0.38	0.46	0.55	0.83

Source: The authors.

Table 4.

Measurement Model Fit Indices

Sl. No.	Fit-Index	Obtained Value	Acceptable Range	References
1	RMSEA	0.054	<0.07	Hu and Bentler (1999), Steiger (2007)
2	SRMR	0.040	<0.08	Hu and Bentler (1999)
3	Chi-Square/degree of freedom	1.78	<3	Kline (2005), Nagar and Gandotra (2016), Schumacker and Lomax (2004)
4	CFI	95.3	Close to or > 0.95	Hu and Bentler (1999)
5	TLI	94.5	Close to or > 0.95	Hu and Bentler (1999)

Source: The authors.

The goodness of fit of the structural model is checked with the same indices used in the measurement model. Table 6 shows the goodness of fit of the structural model. All the indices are found to exhibit satisfactory value.

The ‘R square’ estimate for BI is found to be 0.533 which means that 53.3 per cent of the variance of BI is explained by EE, SSE and PMV. ‘R square’ estimate for EE is 0.507 which means that 50.7 per cent of the variance of EE is explained by SA and SSE.

Discussion

Our entire study revolved around exploring the drivers of technology adoption with special reference to the BOP. With support from existing literature, we have assumed that ‘BI’ leads to actual usage. We have checked the effects of ‘PE’, ‘EE’, ‘SA’, ‘SSE’ and ‘PMV’ on ‘BI’. Also, the effects of ‘SA’ and ‘SSE’ on ‘EE’ are checked. In the below paragraphs, we will discuss the results of each proposition. Table 7 shows the results of the proposed theoretical framework.

Our empirical results show that PE is a strong indicator of ‘BI’ of using a smartphone. This result is consistent with existing literature on a variety of technologies (Bhatiasevi, 2016; Casey & Wilson-Evered, 2012; Venkatesh et al., 2003). ‘Perceived usefulness’ which pertains to ‘PE’ has also proven to be a prominent indicator of ‘BI’ in prominent technology adoption models (Davis, 1986, 1989; Venkatesh & Bala, 2008). When a person believes that the use of a smartphone improves the performance of day-to-day tasks, the person is likely to embrace the use of a smartphone. People who fall in the BOP segment look for usefulness of products as they struggle with their day-to-day life. They are unlikely to invest time or money for something which they think is not useful.

Table 5.

Regression Parameter Values

Independent Variable	Dependent Variable	Estimate	Standard Error	Z value	P value
Performance expectancy	Behavioural Intention	0.335	0.070	4.774	0.000
Effort expectancy	Behavioural Intention	0.380	0.097	3.908	0.000
Smartphone anxiety	Behavioural Intention	0.056	0.073	0.762	0.446
Smartphone self-efficacy	Behavioural Intention	0.024	0.097	0.251	0.802
Perceived monetary value	Behavioural Intention	0.380	0.071	5.376	0.000
Smartphone anxiety	Effort Expectancy	−0.372	0.058	−6.421	0.000
Smartphone self-efficacy	Effort Expectancy	0.473	0.075	6.279	0.000

Source: The authors.

Table 6.

Structural Model Fit Indices

Sl. No.	Fit Index	Obtained Value	Acceptable Range	References
1	RMSEA	0.054	<0.07	Hu and Bentler (1999), Steiger (2007)
2	SMRR	0.040	<0.08	Hu and Bentler (1999)
3	Chi-square/degree of freedom	1.76	<3	Kline (2005), Nagar and Gandotra (2016), Schumacker and Lomax (2004)
4	CFI	95.4	Close to or < 0.95	Hu and Bentler (1999)
5	TLI	94.6	Close to or < 0.95	Hu and Bentler (1999)

Source: The authors.

Table 7.

Results

Sl. No.	Independent Variable	Dependent Variable	Result
1	Performance expectancy	Behavioural intention	Significant
2	Effort expectancy	Behavioural intention	Significant
3	Smartphone anxiety	Behavioural intention	Non-significant because effect is seen to be captured by effort expectancy
4	Smartphone self-efficacy	Behavioural intention	Non-significant because effect is seen to be captured by effort expectancy
5	Perceived monetary value	Behavioural intention	Significant
6	Smartphone anxiety	Effort expectancy	Significant
7	Smartphone self-efficacy	Effort expectancy	Significant

Source: The authors.

The proposed positive relationship between ‘EE’ and ‘BI’ is also supported in our empirical results. The perception of less effort requirement is likely to increase the intention to use a smartphone. The ease of use of technology plays an important role in market penetration and adoption of technology. Previous literature supports this relationship across a variety of technologies (Bhatiasevi, 2016; Casey & Wilson-Evered, 2012; Venkatesh et al., 2003). Multiple prominent studies in technology adoption had found that ‘perceived ease of use’ which is a root construct of ‘PE’ is a primary predictor of ‘BI’ (Davis, 1986, 1989; Venkatesh & Bala, 2008). Since the BOP people are relatively less educated and also have little exposure to new products and technologies, the perception of more effort requirement may make them reluctant to using the smartphone.

We could not find any significant negative relationship between ‘SA’ and ‘BI’. Though Baishya, Samalia, and Joshi (2017) found similar results in the study of e-governance adoption, this result is contradictory to some of the previous empirical research (Adetimirin, 2015; Alenezi et al., 2010; Elasmar & Carter, 1996; Nagar & Gandotra, 2016; Tung & Chang, 2008). Our result of the relationship between ‘SA’ and ‘BI’ is even contradictory to the general perception that anxiety is supposed to impact the BOP segment more as compared to other segments while choosing a product or technology. However, Venkatesh et al. (2003) established that ‘anxiety’ and ‘BI’ do not hold a direct relationship due to the mediating effect of EE. Venkatesh and Bala (2008) also established that ‘computer anxiety’ negatively impacts ‘perceived ease of use’ which directly impacts ‘BI’. Lack of information at the BOP segment may lead to lack of confidence to use relatively newer technologies such as a smartphone. This lack of confidence may create anxiety for them while using or planning to use a smartphone. Although this will result in an accumulation of hindrance to using smartphones, the effect of anxiety may be well captured by ‘EE’.

The proposed relationship between ‘SSE’ and ‘BI’ has not been supported in our study. This result is contradictory to multiple empirical results (Adetimirin, 2015; Alenezi et al., 2010; Elasmar & Carter, 1996; Nagar & Gandotra, 2016; Tung & Chang, 2008). However, this result aligns well with few existing literatures. For instance, Venkatesh (2000) established that the impact of ‘self-efficacy’ on ‘BI’ is totally mediated by ‘perceived ease of use’. Venkatesh et al. (2003) did not include ‘self-efficacy’ as a direct predictor of ‘BI’. The underestimated perception of self-capabilities at the BOP may have attributed to this result. Moreover, the assimilationist behaviour of the BOP segment may lead them to a positive intention to use a smartphone even if they are not confident about their capabilities of using a smartphone.

‘PMV’ showed a strong effect on ‘BI’. This result is similar to multiple previous studies (Dood et al., 1991; Kang & Maity, 2012; Kim et al., 2008). Most of the consumers look for high value for money. Price sensitivity becomes stronger in the case of BOP as they are severely constrained by minimal disposable income. Smartphones may not be perceived as an essential good by the BOP people as they struggle for their survival and basic needs. Therefore, a higher perceived price will compel them to keep the purchase of smartphones in lower preference even though they are interested to use it.

Our study supports the negative relationship between ‘SA’ and ‘EE’. The arousal of anxiety while using a smartphone will lead to the perception of more effort requirement which in turn will decrease the intention to use. Our result is consistent with some previous studies (Venkatesh, 2000; Venkatesh & Bala, 2008; Venkatesh et al., 2003). The BOP people may not always get assistance with usage of smartphones if they get stuck as they are less likely to be surrounded by tech-savvy people. Therefore, more anxiety on smartphone usage will lead to the perception of more effort requirement.

The last proposed relationship in our study is between ‘SSE’ and ‘EE’. Our result shows a significant positive relationship between these two constructs. This means that the perception of self-capability leads to a perception of less effort requirement. This positive relationship has already been established in some existing literature (e.g., Venkatesh, 2000; Venkatesh & Bala, 2008). The BOP people are unable to have frequent interactions with people who are at the top of the economic pyramid and are well educated as well as well informed. Therefore, there is little chance that they can learn the usage of smartphone from these educated and informed people on an everyday basis even if they get stuck while using a smartphone. As a result, the BOP segment may not feel a sense of less effort requirement as long as they are not confident about self-capability of using a smartphone.

Conclusion

Smartphones are becoming popular day by day with a reduction in price. It is observed that all sections of people have gradually started using smartphones. Though the technology adoption rate is higher at the top of the economic pyramid, the market is gradually becoming saturated in those segments. So the marketers are trying to explore new segments at the BOP. If the smartphone can be customized to the needs of the BOP, it has great potential to penetrate into the BOP segment. Furthermore, the adoption of smartphones at the BOP may lead to a social change. If the majority of the people start using smartphones even at the BOP, the government can think of delivering many services to the citizens using applications of smartphones resulting in improved governance. Similarly, financial organizations such as banks can enhance human resource productivity if an effective application such as mobile banking is adopted even by BOP customers. Therefore, the marketers, as well as policymakers, need a better understanding of the inhibitors and enablers of smartphone adoption at BOP.

Our proposed framework contributes to the theoretical base of the technology adoption study at the BOP. We have reviewed the literature related to technology adoption, social cognitive theory and traits of the BOP in order to identify the probable factors which may influence technology adoption at the BOP. It enhances the understanding of technology adoption at the BOP by including constructs related to monetary aspects (PMV) to the framework. Inclusion of this construct itself is a theoretical contribution to the literature as most of the dominant technology adoption models including TRA, TAM and UTAUT did not consider any construct related to monetary aspects. The framework is comprehensive as it considers factors from multiple aspects including technology adoption (PE and EE), social cognitive theory (SA and SSE) and traits of BOP (PMV). The proposed theoretical framework has the potential for guiding future research on technology adoption at the BOP. As this framework is explicit and applicable to the BOP, future studies on technology adoption at BOP can use it as a base for further enhancement and customization.

The current set of literature is in need of empirical evidence on studies related to technology adoption at the BOP. Our study tries to bridge this gap by providing a quantitative analysis of factors influencing the intention to use smartphones at the BOP. Using 266 data points from the Indian BOP segment, we have found that PE, EE and PMV have a direct positive impact on BI of using a smartphone. Contrary to some of the previous studies, our study could not establish a direct impact of SA and SSE on BI of using a smartphone at the BOP. However, the impact of these two constructs on BI is found to be mediated by EE.

Managerial Implications

The results of our study have implications for managers and policymakers. Knowing the fact that there are market opportunities at the BOP, the organizations should concentrate on how to cater to such a market and fulfil the needs of BOP customers. Our study has implications for two types of decision-making procedures for managers of smartphone-producing companies who want to cater to the BOP market. First, the managers should decide on what are the features to be included in the smartphones targeted for the BOP. This decision should consider the traits and expectations of the BOP people. Our results show that EE has a positive direct impact on the BI to use smartphones. This is a clear indication that the effort required to use the smartphone should be minimized in order to increase the adoption rate. Though this indication may be generic to all target markets, the managers should give special concentration for the BOP market as these people are characterized by low literacy and have little exposure to learn new technology. Therefore, the managers should concentrate on how to make smartphones user friendly and easy to use. This requires the collaboration between smartphone designers and target customers. The designers should directly interact with the BOP people to understand their skills and needs. Second, the managers should carefully take decisions regarding the pricing of smartphones targeted for the BOP people as these people are much sensitive to costs due to their restricted disposable income. Our results also show that PMV has a strong impact on the BI to use a smartphone. Therefore, the decision-makers should keep the price of the product as low as possible in order to cater to this market while concentrating more on economies of scale rather than increasing margin per unit.

The second line of the implications of our empirical results is for policymakers. The fact that smartphones have the potential to ease and expedite the process of delivering some of the public services makes this study interesting for the policymakers. Our study has established that PE has a direct positive impact on BI of using smartphones. Therefore, policymakers should put effort to create awareness about the benefits of using a smartphone in the BOP segment. Our model has also established that EE mediates the impact of SA and SSE on BI. Therefore, policymakers can think of implementing programmes such as organizing training programmes on smartphones which will result in reducing anxiety and improving self-efficacy on smartphones among the BOP people.

Limitations

One limitation of our work is that the theoretical model has not included the moderators. While the recent literature of technology adoption has identified multiple moderators such as age, gender, education and income, our study was just confined to exploration of the predictors of ‘BI’. Inclusion of moderators into the framework could have enhanced the implication of the results. Second, our study did not test the relationship between ‘BI’ and ‘user behaviour’. Future researchers on technology adoption at BOP can include these two limitations to improve the understanding.

Footnotes

Acknowledgement

The authors are grateful to the anonymous referees of the journal for their extremely useful suggestions to improve the quality of the article. Usual disclaimers apply.

Declaration of Conflicting Interests

The authors declared no potential conflicts of interest with respect to the research, authorship and/or publication of this article.

Funding

The authors received no financial support for the research, authorship and/or publication of this article.

References

Abdullah

Ward

(2016). Developing a general extended technology acceptance model for E-learning (GETAMEL) by analysing commonly used external factors. Computers in Human Behavior , 56, 238–256.

Adetimirin

(2015). An empirical study of online discussion forums by library and information science postgraduate students using technology acceptance model 3. Journal of Information Technology Education: Research , 14, 257–269.

Alenezi

A. R.

Karim

A. M.

Veloo

(2010). An empirical investigation into the role of enjoyment, computer anxiety, computer self-efficacy and internet experience in influencing the students’ intention to use E-learning: A case study from Saudi Arabian governmental universities. The Turkish Online Journal of Educational Technology , 9(4), 22–34.

Alwitt

L. F.

(1995). Marketing and the poor. American Behavioral Scientist , 38(4), 564–577.

Ansseau

Fischler

Dierick

Albert

Leyman

Mignon

(2007). Socioeconomic correlates of generalized anxiety disorder and major depression in primary care: The GADIS II study (Generalized Anxiety and Depression Impact Survey II). Depression and Anxiety , 25(6), 506–513.

Baishya

Samalia

H. V.

Joshi

(2017). Factors influencing E-district adoption: An empirical assessment in Indian context. International Review of Management and Marketing , 7(1), 514–520.

Bandura

(1977). Self-efficacy: Toward a unifying theory of behavioral change. Psychological Review , 84(2), 191–215.

Bandura

(1986). Social foundations of thought and action: A social cognitive theory . Englewood Cliffs, NJ: Prentice Hall.

Bentler

P. M.

Chou

C.-P.

(1987). Practical issues in structural modeling. Sociological Methods & Research , 16(1), 78–117.

10.

Berger

Nakata

(2013). Implementing technologies for financial service innovations in base of the pyramid markets. The Journal of Product Innovation Management , 30(6), 1199–1211.

11.

Bhatiasevi

(2016). An extended UTAUT model to explain the adoption of mobile banking. Information Development , 32(4), 799–814.

12.

Booker

L. D.

Detlor

Serenko

(2012). Factors affecting the adoption of online library resources by business students. Journal of the American Society for Information Science and Technology , 63(12), 2503–2520.

13.

Bozionelos

(2004). Socio-economic background and computer use: The role of computer anxiety and computer experience in their relationship. International Journal of Human-Computer Studies , 61(5), 725–746.

14.

Carter

Belanger

(2005). The utilization of E-government services: Citizen, trust, innovation. Information Systems Journal , 15(1), 5–25.

15.

Casey

Wilson-Evered

(2012). Predicting uptake of technology innovations in online family dispute resolution services: An application and extension of the UTAUT. Computers in Human Behavior , 28(6), 2034–2045.

16.

Chen

Y.-C.

Lin

Y.-C.

Yeh

R. C.

Lou

S.-J.

(2013). Examining factors affecting college students’ intention to use web-based instruction systems: Towards an integrated model. TOJET: The Turkish Online Journal of Educational Technology , 12(2), 111–121.

17.

Chien

T.-C.

(2012). Computer self-efficacy and factors influencing E-learning effectiveness. European Journal of Training and Development , 36(7), 670–686.

18.

Cho

(2015). Roles of smartphone app use in improving social capital and reducing social isolation. Cyberpsychology, Behavior,and Social Networking , 18(6), 350–355.

19.

Churchill

G. A.

(1979). A paradigm for developing better measures of marketing constructs. Journal of Marketing Research , 16(1), 64–73.

20.

Compeau

Higgins

C. A.

Huff

(1999). Social cognitive theory and individual reactions to computing technology: A longitudinal study. MIS Quarterly , 23(2), 145–158.

21.

Davis

F. D.

(1986). A technology acceptance model for empirically testing new end-user information systems: Theory and results (Doctoral dissertation). Massachusetts Institute of Technology, Sloan School of Management, Massachusetts Institute of Technology, Cambridge, MA.

22.

Davis

F. D.

(1989). Perceived usefulness, perceived ease of use, and user acceptance of information technology. MIS Quarterly , 13(3), 319–340.

23.

Davis

F. D.

Bagozzi

R. P.

Warshaw

P. R.

(1992). Extrinsic and intrinsic motivation to use computers in the workplace. Journal of Applied Social Psychology , 22(14), 1111–1132.

24.

Dey

B. L.

Binsardi

Prendergast

Saren

(2013). A qualitative enquiry into the appropriation of mobile telephony at the bottom of the pyramid. International Marketing Review , 30(4), 297–322.

25.

Dood

W. B.

Monroe

K. B.

Grewal

(1991). Effects of price, brand, and store information on buyers’ product evaluation. Journal of Marketing RESEA , 28(3), 307–319.

26.

Downs

G. W.

Mohr

L. B.

(1976). Conceptual issues in the study of innovation. Administrative Science Quarterly , 21(4), 700–714.

27.

Dwivedi

Y. K.

Rana

N. P.

Janssen

Lal

Williams

M. D.

Clement

(2017). An empirical validation of a unified model of electronic government adoption (UMEGA). Government Information Quarterly , 34(2), 211–230.

28.

Elasmar

M. G.

Carter

M. E.

(1996). Use of e-mail by college students and implications for curriculum. Journalism & Mass Communication Educator , 51(2), 46–54.

29.

Fishbein

Ajzen

(1975). Belief, attitude, intention, and behavior: An introduction to theory and research . Reading, MA: Addison-Wesley.

30.

Fornell

Larcker

D. F.

(1981). Evaluating structural equation models with unobservable variables and measurement error. Journal of Marketing Research , 18(1), 39–50.

31.

Gangwar

Date

(2016). Critical factors of cloud computing adoption in organizations: An empirical study. Global Business Review , 17(4), 886–904.

32.

J.-C.

Lee

S.-C.

Suh

Y.-H.

(2009). Determinants of behavioral intention to mobile banking. Expert Systems with Applications , 36(9), 11605–11616.

33.

Gupta

Dasgupta

Gupta

(2008). Adoption of ICT in a government organization in a developing country: An empirical study. The Journal of Strategic Information Systems , 17(2), 140–154.

34.

Gupta

Jain

(2014). Adoption of mobile telephony in rural India: An empirical study. Decision Sciences , 45(2), 281–307.

35.

Gupta

Jain

(2015). Adoption behavior of rural India for mobile telephony: A multigroup study. Telecommunications Policy , 39(8), 691–704.

36.

Gupta

Srivastav

(2016). An exploratory investigation of aspirational consumption at the bottom of the pyramid. Journal of International Consumer Marketing , 28(1), 2–15.

37.

Gutiérrez

L. H.

Gamboa

L. F.

(2010). Determinants of ICT usage among low-income groups in Colombia, Mexico, and Peru. The Information Society , 26(5), 346–363.

38.

Hair

J. F.

Black

W. C.

Babin

B. J.

Anderson

R. E.

(2006). Multivariate data analysis . Upper Saddle River, NJ: Pearson Prentice Hall.

39.

Hart

S. L.

London

(2011). Next generation business strategies for the base of the pyramid: New approaches for building mutual value . Upper Saddle River, NJ: FT Press.

40.

Hasan

M. R.

Lowe

Petrovici

(2017). Antecedents of adoption of pro-poor innovations in the bottom of pyramid: An empirical comparison of key innovation adoption models-an abstract. In Rossi

(Ed.), Developments in marketing science: Proceedings of the academy of marketing science (pp. 1081–1082). Cham: Springer.

41.

Heeks

(2012). IT innovation for the bottom of the pyramid. Communications of the ACM , 55(12), 24–27.

42.

Hill

Stephens

(1997). Impoverished consumers and consumer behavior: The case of AFDC mothers. Journal of Macromarketing , 17(2), 32–48.

43.

Hossain

M. M.

Jamil

M. R.

(2015). Consumer acceptance of more-than-voice (MTV) services: Evidence from the bottom of pyramid in Bangladesh. The Journal of Developing Areas , 49(5), 25–39.

44.

L.-T.

Bentler

P. M.

(1999). Cutoff criteria for fit indexes in covariance structure analysis: Conventional criteria versus new alternatives. Structural Equation Modeling: A Multidisciplinary Journal , 6(1), 1–55.

45.

Jaiswal

A. K.

(2015). The influence of marketing on consumption behavior at the bottom of the pyramid. Journal of Consumer Marketing , 32(2), 113–124.

46.

Jelinek

Ahearne

Mathieu

Schillewaert

(2006). A longitudinal examination of individual, organizational, and contextual factors on sales technology adoption and job performance. Journal of Marketing Theory and Practice , 14(1), 7–23.

47.

Johnson

R. D.

Hornik

Salas

(2008). An empirical examination of factors contributing to the creation of successful e-learning environments. International Journal of Human-Computer Studies , 66(5), 356–369.

48.

Kang

Maity

(2012). Texting among the bottom of the pyramid: Facilitators and barriers to SMS use among the low-income mobile users in Asia . Retrieved from http://lirneasia.net/wp-content/uploads/2010/07/Texting-among-the-Bottom-of-the-Pyramid-Facilitators-and-Barriers-to-SMS-Use-among-the-Low-income-Mobile-Users-in-Asia.pdf

49.

Kansal

(2016). Factors affecting adoption of mobile banking at the bottom of the pyramid in India. International Journal of Marketing & Business Communication , 5(3), 8–19.

50.

Karnani

(2009). Romanticizing the poor harms the poor. Journal of International Development , 21(1), 76–86.

51.

Kim

Mirusmonov

Lee

(2010). An empirical EXAMINATION of factors influencing the intention to use mobile payment. Computers in Human Behavior , 26(3), 310–322.

52.

Kim

G. S.

Park

S.-B.

(2008). An examination of factors influencing consumer adoption of short message service (SMS). Psychology and Marketing , 25(8), 769–786.

53.

Kimberly

J. R.

Evanisko

M. J.

(1981). Organizational innovation: The influence of individual, organizational, and contextual factors on hospital adoption of technological and administrative innovations. Academy of Management Journal , 24(4), 689–713.

54.

Kline

(2005). Principles and practice of structural equation modeling (2nd ed.). New York, NY: The Guilford Press.

55.

Lallmahomed

M. Z.

Lallmahomed

Lallmahomedc

G. M.

(2017). Factors influencing the adoption of e-government services in Mauritius. Telematics and Informatics , 34(4), 57–72.

56.

Lee

Park

E.-A.

Cho

Jin

(2018). Factors affecting tablet computer users’ intention to purchase mobile applications. Social Behavior and Personality: An International Journal , 46(1), 25–38.

57.

Liébana-Cabanillas

Sánchez-Fernández

Muñoz-Leiva

(2014). The moderating effect of experience in the adoption of mobile payment tools in virtual social networks: The m-payment acceptance model in virtual social networks. International Journal of Information Management , 34(2), 151–166.

58.

Maddux

J. E.

(1995). Self-efficacy theory: An introduction. In Maddux

J. E.

(Ed.), Self-efficacy, adaptation, and adjustment: Theory, research and application (pp. 3–33). New York, NY: Plenum Press.

59.

Malliari

Korobili

Togia

(2012). IT self-efficacy and computer competence of LIS students. The Electronic Library , 30(5), 608–622.

60.

Meeker

(2015). Internet trends 2015—Code conference . San Francisco, CA: Kleiner Perkins Caufield Byers.

61.

Moore

G. C.

Benbasat

(1991). Development of an instrument to measure the perceptions of adopting an information technology innovation. Information Systems Research , 2(3), 192–222.

62.

Mwinyi

Pisanu

Castelao

Stringhini

Preisig

Schiöth

H. B.

(2017). Anxiety disorders are associated with low socioeconomic status in women but not in men. Women’s Health Issues , 27(3), 302–307.

63.

Nagar

Gandotra

(2016). Exploring choice overload, internet shopping anxiety, variety seeking and online shopping adoption relationship: Evidence from online fashion stores. Global Business Review , 17(4), 851–869.

64.

OECD. (2017). Purchasing power parities (PPP) (indicator) . Paris: OECD.

65.

Ong

C.-S.

Lai

J.-Y.

Wang

Y.-S.

(2004). Factors affecting engineers’ acceptance of asynchronous e-learning system in high-tech companies. Information & Management , 41(6), 795–804.

66.

Paglis

L. L.

(2010). Leadership self-efficacy: Research findings and practical applications. Journal of Management Development , 29(9), 771–782.

67.

Phillips

B. N.

Martin

R. P.

Meyers

(1972). Interventions in relation to anxiety in school. In Spielberger

C. D.

(Ed.), Current trends in theory and research (pp. 410–465). New York, NY: Academic Press.

68.

Planning Commission. (2014). Report of the expert group to review the methodology for measurement of poverty . New Delhi: Government of India. Retrieved from http://planningcommission.nic.in/reports/genrep/pov_rep0707.pdf

69.

Prahlad

C. K.

(2004). Fortune at the bottom of the pyramid: Eradicating poverty through profits . Pennsylvania: Wharton School Publishing.

70.

Pynoo

Devolder

Tondeur

Braak

J. V.

Duyck

(2011). Predicting secondary school teachers’ acceptance and use of a digital learning environment: A cross-sectional study. Computers in Human Behavior , 27(1), 568–575.

71.

Rahayu

Day

(2015). Determinant factors of e-commerce adoption by SMEs in developing country: Evidence from Indonesia. Procedia—Social and Behavioral Sciences , 195, 142–150.

72.

Rahman

Hasan

M. R.

Floyd

(2013). Brand orientation as a strategy that influences the adoption of innovation in the bottom of the pyramid market. Strategic Change , 22(3–4), 225–239.

73.

Riffai

Grant

Edgar

(2012). Big TAM in Oman: Exploring the promise of on-line banking, its adoption by customers and the challenges of banking in Oman. International Journal of Information Management , 32(3), 239–250.

74.

Sanakulov

Karjaluoto

(2015). Consumer adoption of mobile technologies: A literature review. International Journal of Mobile Communications , 13(3), 244–275.

75.

Schumacker

R. E.

Lomax

R. G.

(2004). A beginner’s guide to structural equation modeling (2nd ed.). Mahwah, NJ: Lawrence Erlbaum Associates.

76.

Sharma

S. K.

Govindaluri

S. M.

(2014). Internet banking adoption in India: Structural equation modeling approach. Journal of Indian Business Research , 6(2), 155–169.

77.

Simmering

M. J.

Posey

Piccoli

(2009). Computer self-efficacy and motivation to learn in a self-directed online course. Decision Sciences Journal of Innovative Education , 7(1), 99–121.

78.

Steiger

J. H.

(2007). Understanding the limitations of global fit assessment in structural equation modeling. Personality and Individual Differences , 42(5), 893–898.

79.

Sung

(2016). A study of the digital divide in the current phase of the information age: The moderating effect of smartphones. Information Polity , 21(3), 291–306.

80.

Tarafdar

Anekal

Singh

(2012). Market development at the bottom of the pyramid: Examining the role of information and communication technologies. Information Technology for Development , 18(4), 311–331.

81.

Tarhini

Hone

Liu

(2014). The effects of individual differences on e-learning users’ behaviour in developing countries: A structural equation model. Computers in Human Behavior , 41, 153–163.

82.

Telecom Regulatory Authority of India. (2017). The Indian telecom services performance indicators . New Delhi: Telecom Regulatory Authority of India. Retrieved from http://www.trai.gov.in/sites/default/files/Performance_Indicator_Reports_28Sep2017.pdf

83.

Thompson

R. L.

Higgins

C. A.

Howell

J. M.

(1991). Personal computing: Toward a conceptual model of utilization. MIS Quarterly , 15(1), 125–143.

84.

Tung

F.-C.

Chang

S.-C.

(2008). Nursing students’ behavioral intention to use online courses: A questionnaire survey. International Journal of Nursing Studies , 45(9), 1299–1309.

85.

Venkatesh

(2000). Determinants of perceived ease of use: Integrating control, intrinsic motivation, and emotion into the technology acceptance mode. Information Systems Research , 11(4), 342–365.

86.

Venkatesh

Bala

(2008). Technology acceptance model 3 and a research agenda on interventions. Decision Sciences , 39(2), 273–315.

87.

Venkatesh

Morris

M. G.

Davis

G. B.

Davis

F. D.

(2003). User acceptance of information technology: Towards a unified view. MIS Quarterly , 27(3), 425–478.

88.

Warshaw

P. R.

Davis

F. D.

(1985). Disentangling behavioral intention and behavioral expectation. Journal of Experimental Social Psychology , 21(3), 213–228.

89.

Yurdakul

Atik

Dholakia

(2017). Redefining the bottom of the pyramid from a marketing perspective. Marketing Theory , 17(3), 289–303.