Technology Acceptance in Rural Udupi

By
Abhay Velayudhan

Under the Guidance of

Dr. SAVITHA BASRI
 ABSTRACT
Technology has become one of the major factors in driving development of the
country. With infusion of technology the importance of smartphone and internet
banking has gone up. With regards to recent crackdown on black money by GOI,
there was sudden resurgence of e wallets or e-banking services. This was successful
in urban areas, compared to rural areas. The objective of this project is to identify
factors affecting adoption of technology in Rural India. Three villages in Udupi
District was chosen for conducting a survey regarding the same.

The researcher has done a door to door survey and Google docs was also spread
among the rural folk so as to collect data. The researcher also did an informal
interview with respondents so as to understand their mentality towards technology.
Most of the aged group members had a phobia towards usage of technology. After
going through various factors in previous studies regarding the topic, the researcher
identified few key factors and done a study based on how these factors effect in the
decision-making process of adopting technology.

After doing various hypothesis testing the study concludes that PU has a very high
influence on adaptation of smartphone and internet banking service. Therefore,
companies have to focus on making Perceived Usage easier to rural customers’ in
order to gain Intention to Adopt technology from those customers.

i
 Contents
ACKNOWLEDGEMENT ............................................................................................ iv
CHAPTER 1: INTRODUCTION .................................................................................. 1
1.1 ............................................................................................................................... 2
1.3 Scope of study ...................................................................................................... 3
CHAPTER 2: ................................................................................................................. 4
LITERATURE REVIEW .............................................................................................. 4
2.1 Adoption of mobile telephony in rural India ........................................................ 5
2.2 Understanding the use of information by poor female mobile users in rural India
.................................................................................................................................... 6
2.3 Digital Preparedness versus Digital Divide ......................................................... 7
2.4 Empowering rural people to use ITS.................................................................... 9
CHAPTER 3: ............................................................................................................... 11
RESEARCH METHODOLOGY................................................................................. 11
3. 1 Constructs of the study: Level of analysis ........................................................ 13
3.1.1 PU- Perceived use ....................................................................................... 13
3.1.2 PEOU- Perceived Ease of Use .................................................................... 13
3.1.3 EOSA- Ease of Service Accessibility ......................................................... 13
3.1.4 SI- Social Influence..................................................................................... 13
3.1.5 SQ- Service Quality .................................................................................... 14
3.2 Sampling design: Population of the study .......................................................... 14
3.3 Determining the sample size .............................................................................. 14
3.4 Data characteristics ............................................................................................ 16
3.4.1 Data collection ............................................................................................ 16
3.4.2 Description of the sample ........................................................................... 16
3.4.4 Data distribution.......................................................................................... 17
3.5 Method ............................................................................................................... 18
3.6 Scale of variables ............................................................................................... 19
3.7 Structural Model ................................................................................................. 19
3.9 Measurement model: Constructs and items. ...................................................... 19
3.10 Bootstrapping: The algorithm settings ............................................................. 21
3.11 Measurement model assessment: Reliability ................................................... 21
3.12 Measurement model assessment: convergent validity ..................................... 22

ii
CHAPTER 4: ............................................................................................................... 23
DATA ANALYSIS AND INTERPRETATION ......................................................... 23
4.1 Data Analysis ..................................................................................................... 24
4.2 Structural model assessment .............................................................................. 24
4.3 Reliability: .......................................................................................................... 26
4.3.1 Internal consistency reliability .................................................................... 27
4.3.2 Indicator reliability...................................................................................... 27
4.3.3 Outer loadings of indicators ........................................................................ 27
4.4 Hypothesis Testing: ............................................................................................ 30
CHAPTER 5: ............................................................................................................... 32
5.1 Research findings ............................................................................................... 32
5.2 Discussions ......................................................................................................... 32
5.3 Contributions of the study .................................................................................. 33
5.4 Practical implications ......................................................................................... 33
5.5 Research implications ........................................................................................ 33
Bibliography ................................................................................................................ 34
Appendix ...................................................................................................................... 35

iii
 ACKNOWLEDGEMENT

This project report bears the imprint of many persons without the help of whom I
would not have been able to fulfil the tasks.

It’s my pride and privilege to thank my Director Dr. Ravindranath Nayak, School
of Management, Manipal University for letting me as apart this prestigious
institution and for all the help and assistance extended to me during the course of
this project.

Great expression of thanks to my guide Dr. Savitha Basri who gave me her tired
less support and interest during the course of the project and gave me the valuable
ideas of this project and all the required help and references without which this
project is null.

I also would like to thank Professor Nandan Prabhu for guiding me in usage of PLS
software and for translating questionnaire into Kannada.

A special thanks to my parents, friends, family members and other members of the
faculty of the college for their incredible support, help and good wishes. This
project would have been impossible to complete if not for the help of my friends
Anil Kumar and Sahana Bhat as they helped me in data collection.

Abhay Velayudhan

iv
 CHAPTER 1:
INTRODUCTION
1.1

India has one of the fastest rising internet user population in the World. With
evolution in technology internet has come down from PCs to smartphone or
phablets. According to an article in thought leadership, a journal by CISCO
(Wadhwa, 2015), 10% increase in tele-density contributes to .06% growth of
GDP. Internet not only helps in development of economy by providing modern
workforce, it will also open up immense opportunities to agriculture and
dormant sectors. ITCs e-choupals program has already shown what digitization
can do in removing obstacles that hinder growth among rural folks. If the e-
choupal model is implemented in rest of India the real time weather assist,
market price and transportation cost will help them in better strategizing their
products. The true benefit of technology lies in its application. In order achieve
effective application of technology we need to understand the factors that play
an important role in technology adaptation among rural folk.
Ruchita (2014) and Jain (2014) found that technology adoption is a complex
process majorly influenced by factors such as social, economic, technological
and political factors. As part of recent crackdown on black money by
Government of India, India saw a sudden raise in increase of mobile banking
and e-wallet services. Indian Government removed service charges on e-
transactions so as to provide incentive to public to use mobile banking or e-
banking services. According to Law and IT Minister Ravi Shankar Prasad,
“Digital transactions have surged in range of 400-1000% since demonetization.
The e wallet transaction have increased from Rs17 lakh per day to 63 lakh per
day.” This phenomenon have mostly occurred in urban areas. When we compare
TRAI’s (Telecom Regulatory Authority of India) data we find that this
phenomenon will take time to take off in Rural India.
Wiggins and Proctor (2001) suggested three characteristics of rural area for this
hindrance. They are 1. Abundance of land when compared to urban areas, 2.
Distance between rural settlements and cities which are also obstructed by rivers
and mountains, and 3. Lower average incomes when compared to urban areas.
These characteristics make movement of goods and information slow and costly.

2
Moreover low population density and spread of population and some of the
following attributes that differentiate rural area from urban area makes it
difficult. The attributes are:
 Low purchasing power and limited disposable income (NSS, 2011)
 Low education level (Census of India, 2011)
 Cultural characteristic difference, i.e., collective vs individualistic (Jha &
Singh, 2011)
 Poor infrastructure and connectivity.
Now with these constraints the connectivity in rural areas become a big issue.
When these constraints limit growth. According to World Banks source book
Empowerment and Poverty Reduction, “access to information is one of the key
elements for empowerment of disadvantaged populations, since information
enables people to take advantage of opportunity, access services, exercise their
rights and hold state and non-state actors accountable (Narayan 2002).
The wireless users stood around 155.35 million users whereas the rural users
where merely 51.98 million. This shows the gap in rural and urban growth story
of India. Rural India is expected to account for 40% of the 250 million new
wireless users as per a recent study conducted by FICCI and Ernst and Young.
So, it’s relevant to check the technology acceptance level in Rural India.
1.2 Research Objective
To understand technology acceptance level and perception of people in rural
areas on adaptation smartphone and e-banking.

1.3 Scope of study

The study covers only three villages in Udupi district, namely Herebettu, Chara
and Herga villages. Paper can be used to understand technology acceptance
level of people in these villages. This study can be replicated in other villages
in Udupi District. The study only focuses only five aspects of TAM, i.e.
Perceived Use, Perceived Ease of Use, Social Influence, Ease of Service
Accessibility, Service Quality.

3
 CHAPTER 2:
LITERATURE REVIEW

4
2.1 Adoption of mobile telephony in rural India

Gupta and Jain (2014) propose a model based on various factors and their effects
on mobile adoption in rural India. The authors are trying to link technology with
socio economic development of a country. They also have tried to find out the
behavioral pattern of rural population and their general attitude towards
adaptation of technology. In TAM there are two major factors that build the
model, they are PU (Perceived Usefulness) and PEOU (Perceived Ease of Use).
They also added many other sub-factors into technology acceptance model, so
as to develop a tailor made model for developing countries. The authors have
added another 5 sub-factors such as CH (Cost of Handset), LEOSA (Lack of
Ease of Service Accessibility), LSQ (Lack of Service Quality), LOTSA (Lack
of Transparency of Service Offerings), PHH (Perceived Health Hazards). CH
and LEOSA has been added to PU, and LSQ, LOTSA, PHH has been added to
PEOU. Paper has taken into consideration various other factors such as
geographic condition that add on to challenges for service provider to provide
adequate services in rural areas. But the important fact is that most of the service
providers in India have not focused on rural segment. When you compare the
service variants available in city to that of rural area the plans available are
drastically different. The rural folk has lesser plans when compared to urban
mobile users. Paper also have marked the most important drivers in making a
decision, i.e. family and group. The fact that most of the villagers hesitation to
use technology lies with the fact that they are afraid of the changes and
uncertainty that lingers around adaptation of technology. The paper rightly
marks that nothing has been done by the mobile operators to remove this
uncertainty from the mind of rural folks. Unlike the urban crowd which tries to
learn on their own, rural folk require a certain amount of push.

One major drawback of this paper is that it is focusing only on a particular type
of population in rural areas. It has avoided people who have not used mobile
phones and those who have no experience in using mobile phone. This is a major
drawback as it excludes a significant chunk of population from the study. This
means that the conclusion arrived will be biased version as it includes only

5
people with knowledge or hands on experience in using mobile. The study also
hasn’t focused on user experience and what more do the mobile operators have
to do to get the other part of population that’s backing away from adopting this
particular technology. There is further scope for improvement if also focus more
on what technology can do to remove the divide society due to various socio
economic factors.

2.2 Understanding the use of information by poor female mobile users

in rural India
Potnis (2015) focuses on what digital inclusion has done in helping out women
from rural background. As rightly pointed by author “access to information by
itself is of limited value unless the intended beneficiary has the capacity to use
it.” Without the ability or understanding regarding usage of information, it loses
its value. It’s like having the world’s most advanced AI (Artificial Intelligence)
which collects all information but does not know how to use it. In India, women
always have played an important role in its development. But over the years due
to increasing social backwardness in rural areas, women are not as empowered
as their counter parts in urban locality. Mostly women in rural areas are
discouraged from taking up studies or in any skill development programs. The
paper focuses on a model of human information model that leads to digital
inclusion. The steps in the Global model of human information behavior are:
1.Context of Information need (activating mechanism and intervening
variables), 2. Information seeking behavior, 3.Information processing and use.
The paper suggest 6 steps or information use and digital inclusion. They are 1.
Accessing mobile phones, 2. Identifying context specific information need, 3.
Operating mobile phones and engineering their functionalities, 4. Seeking and
searching information over mobile phones, 5. Overcoming human and technical
barriers to receiving, processing and applying information, 6. Satisfying
information needs. In this particular paper the author has chosen to study
working women to do his study. This is also a limitation in the study as it focuses
only on woman who are enforced to adopt technology due to nature of their
work or due to peer pressure. The study then divides them into two groups G1
and G2. G1 consist of women who have adopted technology on their own accord.

6
G2 consist of women who have bought mobile phones, due to their persistence.
The researcher has taken enough steps o as to ensure that data collection was
unprejudiced. To find out how well the subjects have adopted to technology the
researcher has done second round of interview via mobile phone. The study
proves that rural women have more restriction for getting mobile phones when
compared to their contemporaries in the city. G1 had 13 members, and had
purchased mobile understanding the need for the same, whereas G2 were most
of their husbands played key role in getting mobile. The numbers also suggest
the social backwardness still present in our villages. Out of 43 women, only 13
women fell under G1 category. The remaining came under G2. The study has
identified 4 barriers that hinder in receiving, processing and using information.
They are Human barriers in receiving information, Human barriers in applying
information for deriving benefit, Technical barriers in receiving information and
processing information and Technical barriers to processing information. Most
of the ladies uses their mobile for employment purpose. Only 2% of the
respondents or women used mobile for running their own business. This shows
how much of information inclusion is there among the ladies. It has also been
found that once these women got used to mobile phones they have developed
codes as a means to communicate their messages to their loved ones. Another
important finding is that these women are harassed by unknown people who s
peaks in very rude and indecent manner. This particular human behavior creates
lot of hesitation among family members in encouraging women of their families
to use mobile phones. The study needs to further expand its reach on non-mobile
phone users and find out what barriers are keeping them away from adopting
technology. Also a similar study needs to be done on male population so as to
find how much information is being used by the opposite sex in India. If there
is a drastic difference in the sex ratio on digital inclusion further study needs to
be done.

2.3 Digital Preparedness versus Digital Divide

The term 'digital divide' was presented in the mid-1990s and has turned out to
be mainstream topic for interdisciplinary concern. The term has gotten much

7
consideration from numerous analysts and strategy makers. James (2008),
proposes a model on digital divide that includes civic strategy discussion that
incorporates social, monetary and political issues which influences humankind
and the universe at large. ‘Digital divide' may bring about 'knowledge divide' or
'data divide' which mirrors the level of information and data about the universe
and one's environment and additionally its security. Absence of such learning
and data may likewise have particular ramifications for the financial
improvement in various groups. This paper surveys the writing on the advanced
separation as to specialists' diverse perspectives on the idea. The writing
inspected in this paper incorporates diary articles and gathering papers
distributed in the vicinity of 2000 and 2014 from different territories, in
particular data innovation and data science, sociology and training, and financial
aspects and administration science. It is clear from the "review" that the
advanced divide is seen as a contradistinction by the scientists. The distinctive
perspectives of the advanced divide are affected by various variables, for
example, the analyst's field of study, how the specialist comprehends the idea
computerized separation, and issues which the scientist expects to understand.
The paper organized the current writing on computerized isolate as indicated by
perspectives of various writers.

This paper (James, Mar2008) demonstrates that there are different meanings of
the digital divide and that these definitions are educated by specific perspectives
which are viewed by analysts as huge. There is no authoritative definition since
every definition and investigation of the advanced partition mirrors the
perspective of the particular researcher. The distinctive creator's comprehension
of the computerized isolate educated how they saw it. A few researchers trying
to characterize the computerized isolate have a tendency to suit different
perspectives. For instance, a few researchers trust that the digital divide is a
crevice that exists in both the PC itself and the web (Information and gadget
see). Along these lines, the partition can exist inside the two advanced
perspectives, in particular in the information, communication and technology
(ICT) gadgets from one viewpoint and the web on the other. This paper gave a

8
structure of digital divide investigate as per diverse authors' perspectives. By
this, the paper classifies the different comprehension of the digital divide. The
thought imparted in this paper serve to upgrade the aggregate comprehension of
the marvels of the computerized isolate and would likewise fill in as a reason
for future experimental work for the examination think about. Since the issue of
advanced partition can have both positive and negative ramifications on
humanities and in the social orders, this paper recommends a more far reaching
and comprehensive approach towards characterizing digital divide and that the
different basic perspectives ought to be considered. Since every one of the
perspectives specifically or in a roundabout way add to the computerized
crevice, it is suggested that a comprehensive approach be taken in inspecting
the advanced partition. The diverse perspectives concerning the advanced
separation demonstrate that examination on the computerized divergence has
risen above the innovative get to approach.

2.4 Empowering rural people to use ITS

Nagarajan and Jiji (2012) focuses on the importance of knowledge and

information in increasing efficiency in our day today lives. Knowledge and
information also plays a vital role in progress of a developing nation. As
rightfully pointed out by the paper 75% of India’s population consists of rural
people. If the country needs to progress then it needs to ensure that rural folk
are not left behind. Governments have come and gone still rural areas in India
have lower standard of living when compared to under developing countries.
Though governments have implemented many programs to uplift the standard
of living of rural folks, the situation remains same. Some of the rural areas are
still struggling to get basic necessities. The paper points out merely providing
information is not enough, it’s how one empowers people to overcome barriers
that are restricting their progress. The researcher points out that just providing
technology is not enough, we need to make them realize the power of
technology and make them solve their day to day problems so that can
concentrate on growing further up or developing further. Though many villages
have ICT (Information Communication Technology) that allows them to get

9
abundance of information most of them are not self-sufficient villages. The
paper speaks about 3 essential elements that play pivotal role in empowering
villagers, i.e. 1.Awareness, 2.Vision and 3.Technology. As discussed in the
paper though there are services offered by government existing in the villages,
the villagers do not know how to use the services provided to them. Its human
nature to suspect anything or be afraid of something, which we are not aware
off. The researcher have developed a methodology to help villagers empower
in overcoming the barriers that are obstructing the growth of people. These are
the steps involved in making this particular program, 1. Create and appropriate
learning environment, 2. Develop the community involvement, 3.Frame the
curriculum, 4. Training based on the curriculum, Assessment of the program.
The 3rd step is based on community needs, current curriculum and strategies for
development. The only problem with this model is that, it’s based on an
assumption that rural people will be willing to come together and discuss their
problems and seek solution from each other. If this does not happen then the
entire module or the program cannot start. More over this program would
require very high amount of tie as it will be difficult to very high amount of time
as this require very person in a village to trust each other. Though the particular
village in Tamil Nadu which was taken up for study had many organization
running similar programs, these organizations claim that the situation still
remained same year after year. This shows that people require more amount of
time to be invested. It cannot be a training program that needs to be conducted
on a day to day basis than a quarterly, or annual event. This particular model
needs lot of investment and time in order reach the ultimate goal. The socio
economic divide among the villagers will almost make this program almost
impossible to kick off. But the idea provided in the paper is spot on. If we don’t
empower or village, the road to becoming a developed nation will be a distant
dream.

10
 CHAPTER 3:
RESEARCH METHODOLOGY

11
This chapter has discussed the research methodology adopted for this study. It
has discussed the reasons for the team level of analysis of constructs of this
research endeavor. Further, it has explained the sampling design adopted for
this study. Furthermore, the chapter has narrated the process of determining the
sample size of this study. It has explained, thereafter, the issues such as process
of data collection, characteristics of samples and data distribution. Moreover it
has dwelt on the method, operational definitions, scale of variables, and
measures adopted for this research endeavor. Thereafter, this chapter has
narrated the structural and measurement models of the study. In addition, it has
investigated the reliability and validity of the constructs of this study. Finally,
the measurement model results are presented. The following figure provides an
overview of contents of this chapter.

Figure 3.1: An overview of chapter 3

12
3. 1 Constructs of the study: Level of analysis

The theoretical framework of this research endeavor, based on the review of

prior research, has zeroed-in on five constructs i.e. Perceived Usefulness,
Perceived Ease of Use, Social Influence, Ease of Service availability and Low
Service Quality. All these are behavioral constructs. This section explains these
constructs, in brief, and the reasons for conceptualizing them for analysis.

3.1.1 PU- Perceived use

Most of the literatures regarding this study have indicated that Perceived Use
plays an important role in enhancing Intention to Adapt. In any buying scenario
a customer buys product mainly on this particular construct, i.e. Perceived Use.
From this the research has formed first hypothesis is formed.

Hypothesis 1: Positive PU strengthens IA for smartphone usage in rural India.

3.1.2 PEOU- Perceived Ease of Use

This construct doesn’t have a strong influence over IA as PU and PEOU are
almost similar. PEOU can be replaced completely by PU. PEOU has an indirect
influence on customer’s IA. Many literatures have indicated that PEOU has a
significant effect on PU. From this second hypothesis is formed.

Hypothesis 2: PEOU has a positive effect over PU

3.1.3 EOSA- Ease of Service Accessibility
Previous researchers have identified EOSA also as an important construct which
has influence over target construct IA. This is mainly due to belief that without
proper access to buy something or use some service consumers IA will reduce.
Therefore short coming in EOSA can negatively affect IA.

Hypothesis 3: Negative EOSA leads has a negative impact on IA for smartphone and
e banking services in rural India.
3.1.4 SI- Social Influence
As villagers are believed to have a collectivistic sense of approach or
collectivistic culture, where in they provide importance to family and group.
Hence SI is believed to have important role in creating IA in the mind of rural
customers.

13
Hypothesis 4: SI plays positive role in adaptation of smartphone and e-banking
sector.

3.1.5 SQ- Service Quality

Service quality is usually result of customer expectations versus service
provided. Whenever there is inconsistency between the two it creates a negative
impact on IA. SQ is represented by availability of bank branches, continuity of
data connection etc.

Hypothesis 5: Customer develop negative IA when there is low SQ.

3.2 Sampling design: Population of the study

The target population of the present research endeavor is people from rural area
of Udupi district i.e. Herga, Herebettu, Chara villages. The universe of the
present study is infinite. This is because this research endeavor intended to study
technology acceptance in rural area. The current research intends to investigate
the degree of technology acceptance in rural area. The data are collected
randomly from the residents of those particular villages. Therefore, the
population for current study constitutes of residents of the chosen villages.

3.3 Determining the sample size

Prior research suggests the consideration of desired R^2 levels to determine

sample size. The table provided by Cohen (1988, pp.155-519) considers desired
4eR^2 values also in addition to the other requirements such as the number of
arrows that point towards the endogenous construct, statistical power of 80%
and the significance levels of 1%, 5% and 10% for determining the sample size.
Cohen’s (1988) table is given below:

14
 Table 3.2: Sample size recommendation in PLS-SEM for obtaining power of
80%

Maximum
Number
of Arrows Significance Level
Pointing 1% 5% 10%
at a Minimum R^2 Minimum R^2 Minimum R^2
Construct 0.10 0.25 0.50 0.75 0.10 0.25 0.50 0.75 0.10 0.25 0.50 0.75

2 158 75 47 38 110 52 33 26 88 41 26 21
3 176 84 53 42 124 59 38 30 100 48 30 25
4 191 91 58 46 137 65 42 33 111 53 34 27
5 205 98 62 50 147 70 45 36 120 58 37 30
6 217 103 66 53 157 75 48 39 128 62 40 32
7 228 109 69 56 166 80 51 41 136 66 42 35
8 238 114 73 59 174 84 54 44 143 69 45 37
9 247 119 76 62 181 88 57 46 150 73 47 39
10 256 123 79 64 189 91 59 48 156 76 49 41
Source: Cohen, J.A. Power Primer. Psychological Bulletin, 112, pp.155-519.

From the above mentioned table, the significance level chosen is 10%. There
were five arrows that pointed at the target construct of the study i.e. intention to
adopt. For achieving the minimum R-squared level of 0.25 at 10% level of
significance, while five arrows are pointing at the target construct, the expected
sample size would be 58 respondents. This study could get 102 responses. Out
of 102 respondents, 62 respondents confirmed that they were using smartphone.
Hence, according to the above-mentioned sample size recommendations, the
sample size works out to be 58 to produce a minimum R^2 value of 0.25 at 10%
level of significance.

15
3.4 Data characteristics
The following section describes the process of data collection, description of
the sample responses collected, data cleaning process and the characteristics of
data distribution.

3.4.1 Data collection

The survey instrument, that was administered to survey respondents i.e.
residents of chosen villagers contained items to tap the five constructs of the
study. Perceived Usefulness, Perceived Ease of Use, Social influence, Ease of
service accessibility, Low Service Quality.

Perceived Usefulness was measured using four questions in the final

questionnaire. Three out of four questions were about smartphone and
remaining one question was about mobile banking.

Perceived Ease of Use had seven question in the final questionnaire, of which
three questions represented about smartphone usage and remaining four
represented about e-banking. Social influence was represented by three
questions in the final questionnaire. All of the three questions where regarding
e-banking.

Ease of service accessibility was represented by four questions in the final set
of questions formed. Low Service Quality was represented by two questions
and Intention to adopt had a representation of four questions.

3.4.2 Description of the sample

The questionnaire was handed out to residents in village through both hard and
soft copies. The researcher provided printed forms of Google docs and soft copy
of the same in mobile phone. The villagers had the option of filling
questionnaire in either one of the format. Around 60 responses where collected
in hard copy and remaining where collected in soft copies. The questionnaire
consisted of 24 questions. The factors were mapped from existing study. All the
questions were measured in 5-point Likert scale. The questionnaire was
translated to Kannada (Official language of Karnataka State), so as to make the
questions understandable to local folks. Intention to Adapt is the dependent
variable in the questionnaire.

16
The age group of the sample was divided into 5 categories, they are
under 25 years, 25-35 years, 35-45 years, 45-55 years and above 55
years. The first age group respondents are 21 in number, i.e. 20.6% of
the sample. The second age group had a representation of 27 members
which translates into 26.5% of data. The third age group had around 14
respondents. 13.7% of the sample involved the third age group. The
fourth age group were 8 in number, i.e. 17.6% of total sample. And the
final age group had a representation of 21.6% in sample, which
translates to around 22 respondents. Off the 102 respondents, 47 were
female. They constitute around 46.1% of the sample. The remaining of
55 respondents were male which set up remaining 53.9% of the data. 62
people, i.e. 60.8% of the respondents said that they were currently using
smartphones. The remaining 40 said that they weren’t using smartphone.
3.4.4 Data distribution
The values of skewness and kurtosis of all variables ranged between the
acceptable range +1 and -1.

Table 3.3: Descriptive statistics of the constructs

Descriptive Statistics

N Range Mean Skewness Kurtosis

Statistic Statistic Statistic Statistic Std. Error Statistic Std. Error

PU1 62 4 2.05 1.172 .304 1.162 .599

PU2 62 4 2.05 1.238 .304 2.247 .599

PU3 62 4 2.06 1.264 .304 1.824 .599

PU4 62 4 2.34 .785 .304 .522 .599

PEOU1 62 4 2.26 .992 .304 .840 .599

PEOU2 62 4 2.08 1.182 .304 2.105 .599

PEOU3 62 4 1.95 1.229 .304 1.778 .599

PEOU4 62 4 2.37 .627 .304 .596 .599

PEOU5 62 4 2.42 .180 .304 -.203 .599

17
 PEOU6 62 4 2.53 .257 .304 .128 .599

PEOU7 62 4 2.56 .650 .304 .841 .599

SI1 62 4 2.52 .550 .304 .338 .599

SI2 62 4 2.50 .062 .304 -.264 .599

SI3 62 4 2.63 .064 .304 -.615 .599

EOS1 62 4 3.37 -.359 .304 -.674 .599

EOS2 62 4 3.47 -.503 .304 -.682 .599

EOS3 62 4 3.48 -.681 .304 -.669 .599

EOS4 62 4 3.23 -.226 .304 -1.066 .599

LS1 62 4 2.97 .181 .304 -1.118 .599

LS2 62 4 2.85 -.129 .304 -1.115 .599

IA1 62 4 2.03 1.145 .304 1.183 .599

IA2 62 4 2.00 1.146 .304 1.334 .599

IA3 62 4 2.16 1.139 .304 1.366 .599

IA4 62 4 2.18 .954 .304 .136 .599

Valid N (list wise) 62

The skewness values of the construct ‘Intention to adopt ‘range between

1.264and -0.681, and the kurtosis values of the construct range between -
1.115 and 2.105

3.5 Method
This research endeavor has used the PLS-SEM approach for its model
specification and measurement (Hair, Ringle, &Sarstedt, 2011). The critical
characteristics of PLS-SEM approach are discussed below among the categories
such as (i) data characteristics, (ii) model properties (iii) algorithm properties of
PLS-SEM, and (iv) model evaluation against the context of which an
explanation is provided on the reasons for the choice of PLS-SEM approach to
analyze data of the present study.

18
3.6 Scale of variables
A 5-point Likert scale is used to capture the responses, and the variation of
responses is captured across the continuum of responses between ‘strongly
agree’ and ‘strongly disagree’. Accordingly, this study has captured data only
in metric scale.

3.7 Structural Model

The Structural theory is expected to show the relationships among the constructs
of a research endeavor that adopts PLS-SEM approach. The current research
adopts reflective type of model.

Figure 3.2 Structural Model

3.9 Measurement model: Constructs and items.

The Table 3.10 outlines the constructs and the items with which this researcher
has measured them:

19
 Table 3.10: The structure of constructs and items used in the construct of
the study
SL.NO LOC CODE Indicator
PU1 Smartphone makes me more efficient.
Perceived PU2 Smartphone allow me to do things better.
1
Use PU3 Smartphone allows me to act in emergency.
PU4 Mobile Banking allow me to transact better.
I believe it is easy to do what I want to do in my
PEOU1
smartphone.
PEOU2 I find smartphone easy to use.
Learning to operate a smartphone and data services
PEOU3
is easy for me.
I believe it is easy to do what I want to do in mobile
Perceived PEOU3
2 banking service.
Ease of Use
I find mobile banking or internet banking easy to
PEOU4
use.
Learning to operate internet banking and
PEOU5
transactions is easy for me.
My interaction with internet banking is clear and
PEOU6
understandable.
People around me find I should use internet
SI1
banking.
Social People around me think it is good idea to use
3 SI2
Influence internet banking.
People around me have simulated me in using
SI3
internet banking.
There is poor availability of the latest handsets of
ESOA1
mobile in my area.
Ease of ESOA2 Prepaid recharge are not easily available in my area.
4 Service Poor electrification in my area causes problem in
ESOA3
accessibility charging my phone.
There are no bank branches in my area to provide
ESOA4
net banking.
Service LS1 There is poor coverage of mobile network.
5
quality LS2 There is no continuity for mobile data connectivity.
IA1 I intend to use smartphone someday.
Intention to IA2 I intend to use internet from my phone someday.
6
adopt IA3 I intend to use mobile banking services someday.
IA4 I intend to transact with my net banking someday.

20
 Figure 3.3 Reflective model formed from the above constructs

3.10 Bootstrapping: The algorithm settings

As there are no distributional assumptions in PLS-SEM approach, we can
conduct only the non-parametric tests. This requires invoking the non-
parametric bootstrapping procedure so as to test the significance of coefficients.
3.11 Measurement model assessment: Reliability
PLS-SEM approach recommends the measurement of reliability by using the
measure of composite reliability as it measures internal consistency reliability
far better than Cronbach's Alpha. In this regard, composite reliability has two
distinct advantages over Cronbach's Alpha. First, the measure of composite
reliability does not consider all indicators to be the equal contributors to the
construct as does the measure of Cronbach's Alpha. Second, composite
reliability does not underestimate the internal consistency reliability by not
demonstrating the tendency to increase the value of internal consistency even
as the number of items in the scale increases, a tendency that we find in the
measure of Chronbach's Alpha.

21
 Figure 3.4: The results of the measurement models

3.12 Measurement model assessment: convergent validity

Average variants extracted (AVE) is the measure through which we evaluate
convergent validity. AVE is also the communality of a given construct. It is
measured by finding out the average of the squared outer loadings of all
indicators of a construct. The expected threshold value of AVE is 0.50, which
implies that a given construct is able to explain more than half of its indicators-
variances.

22
 CHAPTER 4:
DATA ANALYSIS
AND
INTERPRETATION

23
4.1 Data Analysis
This chapter has assessed the structural model of this research endeavor.
Assessment of collinearity, finding out coefficient of determination, and assessing
the significance of relationships. This chapter has also reported the results of
hypotheses testing through PLS-SEM approach.

4.2 Structural model assessment

The following procedure has been adopted to assess the structural model. First, this
study has made the collinearity assessment of constructs included in the structural
model. Second, path coefficients of structural model are obtained by running the
PLS-SEM algorithm. Third, the implications of coefficient of determination are
assessed.

Figure 4.1 The PLS path diagram that shows the direct effect of constructs
on IA
The figure 4.1illustrates the path estimates of the relationship between constructs
and IA including the mediator PU. Coefficient of determination is the criterion
through which we assess the structural model (Henseler, Ringle,& Sarstedt, 2012).
The value of R2, the coefficient of determination, stands reasonably moderate at
0.397 for the target endogenous construct of the study i.e. IA. This implies that the
structural model has predictive validity.

24
 Figure 4.3: Model showing the significance of relationship after running
PLS
Endogenous construct R2 t– value
IA 0.397 2.885

Bias corrected 95%

Relation Path coefficient p-value
confidence interval
PU 0.785 0.00 (0.112, 0.606)

Table 4.1: Assessment of PLS path model without the inclusion of

mediator
Note: It may be noted that this study chose the algorithm settings of 58 cases, 5000
samples, and the option of ‘no sign changes’ in order to assess the path coefficients’
significance (Hair et al., 2014).

There are varying rules of thumb that explain whether R2 values are high or not
(Hair et al., 2014, p.20). Prior research (Ibid, 2014) states that R2 values of 0.20 are
considered high in certain research endeavors whereas the values of 0.75, 0.50 and

25
0.25 for endogenous constructs are considered to be high, moderate and weak
respectively in other studies. This study has adopted the above rule of thumb on R2
values to decide whether or not R2 values are high, moderate or weak.

Also there is no significant difference seen between PESOU and PU. So in order
to improvise the model PESOU is eliminated construct from the model. PU was
then considered as a latent variable instead of mediator. The PLS algorithm was run
for 3nd time to arrive at the expected results.

Figure 4.5: Path co-efficient of the model; perceived usage being the mediator

The R2 of the model is 0.397. This means that the construct is moderately
relevant, i.e. This means that all of the contructs doesn’t neccesarly lead to target
construct. From the above run model (Figure 4.3) Its clear that only PU has
significant influence over IA, as the score is above 2.79. The remainder of the
construct does not have as much influence as PU over IA.

4.3 Reliability:
The indicators of constructs demonstrate higher outer loadings than the
threshold value of 0.70 (Table 4.2). Further the indicator reliability values are
also above 0.70 (3.14). All constructs show composite reliability values which
are above 0.85 i.e. much above the threshold value (Table 4.2).

26
4.3.1 Internal consistency reliability
All constructs of the measurement models, in the present study, demonstrate
high levels of internal consistency reliability. This is illustrated by the values of
composite reliability of constructs which are higher than 0.80 (Table 4.2)

4.3.2 Indicator reliability

Indicator reliability is another measure of reliability of items. We find out
indicator reliability by squaring the outer loadings of an indicator. Therefore,
we describe the indicator reliability as communality of an item. It indicates the
extent of variance extracted from an item. The threshold value of indicator
reliability is 0.50. This implies that a given construct explain at least 50% of an
indicator’s variance. Indicator reliability of all lower order constructs is also
above the threshold value of 0.50, in the present study (Table 4.2). Further,
values of indicator reliability of all higher order constructs are also above 0.50
(Table 4.2).

4.3.3 Outer loadings of indicators

Outer loading of indicators in reflective measurement models, indicate the
absolute contributions of indicators of their respective constructs. PLS-SEM
approach calculates outer loadings of indicators by running simple regression
models in which a latent construct is the independent variable and a given
indicator is its dependent variable. Therefore while outer loadings of indicators
signify the absolute contributions of indicators, indicator reliability signifies the
communality of every indicator.

The threshold value of outer loadings is 0.70 and of indicator reliability is 0.50.
Indicators of all lower order construct of the present study demonstrate values
which are above the threshold value of 0.70 (Table 4.2). Similarly, all higher
order constructs’ outer loadings are also above the threshold value of 0.70
(Table 4.2).

27
 Indicato Compos
Outer
Sl. r ite AV
LOC CODE Indicator loadi
No Reliabil Reliabil E
ng
ity ity
Smartphone makes me more
PU1 0.854 0.729
efficient.

Smartphone allow me to do
Percei PU2 0.882 0.778
things better. 0.72
1 ved 0.914
Smartphone allows me to 6
Use PU3 0.876 0.767
act in emergency.
Mobile Banking allow me
PU4 0.793 0.629
to transact better.
I believe it is easy to do
PEOU1 what I want to do in my 0.793 0.629
smartphone.
I find smartphone easy to
PEOU2 0.75 0.563
use.
Learning to operate a
PEOU3 smartphone and data 0.771 0.594
services is easy for me.
Percei
I believe it is easy to do
ved 0.64
2 PEOU4 what I want to do in mobile 0.809 0.654 0.927
Ease of 4
banking service.
Use
I find mobile banking or
PEOU5 0.817 0.667
internet banking easy to use.
Learning to operate
PEOU6 internet banking and 0.867 0.752
transactions is easy for me.
My interaction with
PEOU7 internet banking is clear and 0.806 0.650
understandable.

People around me find I

SI1 0.827 0.684
should use internet banking.

People around me think it

Social
3 SI2 is good idea to use internet 0.901 0.812 0.895 0.74
Influe
nce banking.
People around me have
SI3 simulated me in using 0.852 0.726
internet banking.

28
 There is poor availability of
ESOA1 the latest handsets of mobile 0.722 0.521
in my area.

Ease of Prepaid recharge are not

ESOA2 0.828 0.686
Servic easily available in my area.
0.58
4 e Poor electrification in my 0.846
1
accessi ESOA3 area causes problem in 0.682 0.465
bility charging my phone.
There are no bank branches
ESOA4 in my area to provide net 0.807 0.651
banking.
There is poor coverage of
Low LS1 0.983 0.966
mobile network. 0.83
5 service 0.908
There is no continuity for 3
quality LS2 0.836 0.699
mobile data connectivity.
I intend to use smartphone
IA1 0.895 0.801
someday.
I intend to use internet from
Intenti IA2 0.866 0.750
my phone someday. 0.80
6 on to 0.944
I intend to use mobile 7
adopt IA3 0.909 0.826
banking services someday.
I intend to transact with my
IA4 0.923 0.852
net banking someday.

(Table 4.2): Measurement models results of the constructs

29
4.4 Hypothesis Testing:

95%
Path P confidence
Relation coefficient Values t-values intervals
EOSA_ -> IA 0.362 0.718 -0.299 0.541
LS2 -> IA 0.909 0.364 -0.207 0.343
PU -> IA 2.705 0.007 0.257 1.098
SI -> IA 0.075 0.94 -0.498 0.332
Table 4.3: Path coefficient and p-values after the inclusion of a mediator

Hypotheses Statement of hypothesis Results

No.
Hypothesis 1 Positive PU strengthens IA for smartphone usage Supported
in rural India.
Hypothesis 2 PEOU has a positive effect over PU Removed

Hypothesis 3 Negative EOSA leads has a negative impact on Not

IA for smartphone and e banking services in rural supported
India.
Hypothesis 4 SI plays positive role in adaptation of Not
smartphone and e-banking sector. supported

Hypothesis 5 Customer develop negative IA when there Not

is low SQ. supported

Table 4.4 Results of Hypothesis Testing

The conclusion in the table 4.4 was reached after going through the results in
table 4.3. PU has a path coefficient of 2.705 which is highest among other path
coefficients and has positive t-values. Hypothesis 3, 4 and 5 has negative T-
values and path coefficient values below 1.

30
31
 CHAPTER 5:
CONCLUSION

5.1 Research findings

The main objective of the research was to find out acceptance level of rural folks
of Udupi district towards usage of smartphones and mobile banking. It’s found that
acceptance level towards technology is bit low when compared to urban areas. The
idea that led to emergence of this research was the recent demonetization drive and
the difficulties that were faced by every Indian citizen. In the city due to already
existing e-banking and mobile wallet services that are predominantly depended on
smartphone. In many media houses it was reported that it was the rural folks who
bore the brunt of demonetization. In order to understand what factor drive rural
folks to accept technology and use it in their day to day life this research was
conducted. The study has shown how important PU (Perceived Usage) is important
in reaching target construct, i.e. IA (Intention to adapt) (Table 4.3).

5.2 Discussions
In most of the previous studies on TAM model on rural India Perceived Usage have
always emerged as the top factor that had influence over intention to adopt. Some
of them tried to bring in concept of cost of the handset, role played by marketing,
social influence and few more other factors (Guptha and Jain, 2014). The researcher
wanted to see how far these factor which are very important in urban locale actually
32
mattered in rural areas. The researcher also was able to solidify with the opinion
that Perceived Usage played a major role in adaptation of technology by interacting
with the sample size.

Though the research have included other factor that affect Intention to Adapt
technology most of them do not seem to care what someone else thinks about them
if they are using technology or not. Due to forceful change of environment around
them those who have intention to adapt smartphone and e-banking, do not know
the importance of Service Quality and Ease of Service Accessibility.

5.3 Contributions of the study

This research further strengthens the opinion found in previous studies that
Perceived Usage is the fore front runner that drives technology adaption. The
research also have identified that quality of service also plays a minor role in
technology adoption when compared to Perceived usage. Bad network coverage
and data connectivity discourages people in trying out new technology such as
smartphone and e-banking which depends primarily on internet. The research after
analyzing data has contributed to TAM by its articulation that Perceived Usage is
the potential construct that influence technology adaption.

5.4 Practical implications

Technology is the future for growth of country and economy as well. The study
also provides what the policy makers and service providers need to focus, so as to
include the major population who is hesitant to use smartphones. The government
should give classes or training to rural folk as to how they can use smartphone and
e-banking services. They need to remove the fear that technology can only be
understood by literate people and it’s unlearn able and not user friendly. The
marketing team of smartphone companies have to adapt innovative marketing
campaign so as to remove this phobia from mind of rural folk aged between 30-55
years.

5.5 Research implications

This research can be further enhanced by focusing only on actual smartphone users
so as to understand further as to why they use it. The current study is a reflective
study. Future researchers can focus on developing complex models which can

33
formative in nature. The researchers can also find out why rural folks have
developed this phobia towards technology.

Bibliography

Census of India. (2001). Census of India. Retrieved from

http://www.censusindia.gov.in/.
Davis, D. (1989). Perceived usefulness, perceived ease of use, and user acceptance.
MIS Quarterly, 319–339.
Gupta, R., & Jain, K. (Feb2014). Adoption of Mobile Telephony in Rural India: An
Empirical Study. Decision Sciences, p207-209.
Hair, J., G.T.M.Ringe, & Sarstedt, C. (2014). A primer on partial least squares
structural equation modeling. Sage, Thousands Oaks.
James, J. (Mar2008). Digital preparedness versus the digital divide: A confusion of
means and ends. Journal of the American Society for Information Science &
Technology, p785-791.
Lopez-Nicolas, & Molina-Castillo. (2008). An assessment of advanced mobile
services acceptance: Contributions from TAM and diffusion theory.
Information & Management, 359-364.
Narayan, D. (2002). Empowerment and Poverty Reduction: A source Book. Open
Knowledge Repositry, 50-60.
NSS. (2011). Key indicators of household consumer expenditure in India, 2009–2010.
Press release, Government of India, national sample survey office.
P.Nagarajan, & Jiji, G. (Jan2012). Empowering Rural People to use Information
Technology Services. Journal of Marketing & Communication, p16-20.
34
Potnis, D. D. (2015, Jan/Feb). Beyond Access to Information: Understanding the Use
of Information by Poor Female Mobile Users in Rural India. Information
Society., p83-93.
Singh, K., & Jha, S. D. (2011). An analysis of individualism-collectivism across
Northern India. Journal of the Indian Academy of Applied Psychology, 37(1),
149–156.
TRAI. (2017). Annual report of Telecome Regulatory Authority of India . TRAI.
V, V., & D., D. F. (1996). A model of the antecedents of perceived ease of use:
Development and test. Decision Sciences, 451–481.
Wadhwa, N. (2015). How telecom can empower rural India. Thought Leadership.

Appendix

35