Dissataion - Ranga Perera (CB002688)
-
Upload
ranga-perera -
Category
Documents
-
view
49 -
download
6
Transcript of Dissataion - Ranga Perera (CB002688)
MSc. in TECHNOLOGY MANAGEMENT
DISSERTATION
ON
DEVELOPMENT OF A CONSUMER TECHNOLOGY ADOPTION MODEL
FOR MOBILE DATA SERVICES WITH UTILITARIAN AND HEDONIC
VALUE PROPOSITIONS
BY
Ranga Perera
CB002688
7TH SEPTEMBER 2009
1
DISSATATION
Student Name: Ranga Naresh Perera
Student Number: CB002688
Email Address: [email protected]
Award Name: MSc in Technology Management
Site Name: APIIT Sri Lanka
Title of Project: DEVELOPMENT OF A CONSUMER TECHNOLOGY
ADOPTION MODEL FOR MOBILE DATA SERVICES WITH UTILITARIAN
AND HEDONIC VALUE PROPOSITIONS
Supervisor: Professor Kennedy Gunawardena
External supervisor: Associate Professor Ms. Geetha Kanaparan
2
Abstract
This research investigates the adoption of two mobile data services with utilitarian and
hedonic value propositions in Sri Lanka. The proposed Technology adoption model has
been built based on empirical academic research into consumer motives of cognition,
hedonics, social influences and studies into consumer behavior attitude and intension.
The model attempts to explain 40%-54% of the Sri Lankan consumer behavior in the
context of selected mobile data services. There are a number of important findings from
this research, including identification of key determinants of technology adoption in
mobile data service, the importance of variables such as perceived usefulness, perceived
ease of use and comparative advantage in the adoption decisions. The research further
explores the relationship of hedonic motives and their influence of attitude towards
adoption and adoption intension.
3
Acknowledgement
The writing of this thesis has been one of the most significant academic challenges I have had to
face. Without the patience, support and guidance of the following people I would have not been
able to complete this journey.
Professor Kennedy Gunawardena who undertook to supervise my Dissertation with short notice
in June. Had it not been for the wisdom, knowledge and commitment of Professor Kennedy, I
doubt that I would have been able to present this dissertation. In June when Professor took the
supervision of my research, I was lost and confused. His knowledge and experience guided me,
inspired me and motivated me. I hope this research justifies the support and confidence you
extended to me.
Professor Gordon C. Bruner II from the Southern Illinois University, USA who was kind enough
to provide me research papers and advise on the Consumer Acceptance of Technology model,
which I used as the foundation of this dissertation. Professor Herbjørn Nysveen from Norwegian
School of Economics and Administration for research papers on Mobile Data Services.
The volunteers and provincial coordinators who helped in mammoth task of distributing and
collecting the questionnaires nationally. Thanks to your friendship and interest I was able to
undertake one of the first national surveys on mobile data services adoption and usage in Sri
Lanka.
To Eranga, Sepali, Nayomi, Bashini, Harshini, Chethani, Priyanwada, Janaki, Sadani, Harsha and
Eureka who entered the 450+ questionnaires painstakingly by working day and night, I am
humbled at your friendship and dedication. All the analysis in this research was possible due to
this effort.
To my friends Eranjan and Leshani who extended their valuable support and time to ensure that
this dissertation was a success.
And Finally to my Mom and brother
4
Table of Contents
1. Introduction ............................................................................................................... 11
1.2 Problem overview ............................................................................................. 13
1.2.1 Problem statement ..................................................................................... 13
1.2.2 Aims .......................................................................................................... 13
1.2.3 Objectives ................................................................................................. 13
1.3 Justification for selection of Mobile Data Services for research .................. 14
1.4 Significance of the study ................................................................................... 17
1.4.1 Theoretical significance ............................................................................ 17
1.4.2 Significance to other stakeholders ............................................................ 19
1.5 Scope and Limitations....................................................................................... 20
2. Background ............................................................................................................... 21
2.1 Mobile telecommunication industry overview ................................................. 21
2.2 Mobile Technology Evolution .......................................................................... 24
2.3 Mobile Data Services ........................................................................................ 27
3. Literature review ....................................................................................................... 32
3.1 Overview of the selected research area ............................................................. 32
3.2 Review of literature on research subject ........................................................... 33
3.2.1 Motives – Utility vs Hedonics .......................................................................... 33
3.2.2 Technology adoption models and Mobile Data Services adoption .................. 35
3.3 Literature review on selected independent variables ............................................ 41
3.3.1 Independent variable 1 - Perceived usefulness ......................................... 41
3.3.2 Independent variable 2 - Perceived ease of use ........................................ 42
3.3.3 Independent variable 3 - Relative advantage ............................................ 43
3.3.4 Independent variable 4 - Pleasure ............................................................. 44
3.3.5 Independent variable 5 - Arousal .............................................................. 44
3.3.6 Independent variable 6 - Dominance ........................................................ 45
3.3.7 Independent variable 7 - Social Influences ............................................... 46
3.3.8 Attitude and Intention ............................................................................... 47
3.3.9 Short Message Service – Mobile Data Service used to test the cognitive
utilitarian value proposition ...................................................................................... 48
5
3.3.10 Mobile Ringtone – Mobile Data Service used to test the hedonic value
proposition ................................................................................................................ 48
3.3.11 Utilitarian Motives .................................................................................... 50
3.3.12 Hedonic Motives ....................................................................................... 51
4. Solution ..................................................................................................................... 52
4.1 Solution overview ............................................................................................. 52
Proposed model for mobile services adoption in Sri Lanka (Sri Lanka Consumer
Acceptance of Technology Model – SLCAT) .............................................................. 53
4.2 List of developed hypothesis ............................................................................ 54
5. RESEARCH METHODOLOGY.............................................................................. 60
5.1 Research Philosophy ......................................................................................... 61
5.2 Research Approach ....................................................................................... 61
5.3 Research Strategy.............................................................................................. 62
5.4 Pilot study ......................................................................................................... 63
5.5 Time Horizon .................................................................................................... 63
5.6 Determining the Sample and Sample Size ........................................................ 64
5.7 Questionnaire design – Likert scales used ........................................................ 66
5.8 Treatment of data .............................................................................................. 67
6. Deliverable ................................................................................................................ 68
6.1 Descriptive Analysis ......................................................................................... 68
6.1.2 Respondents by Gender ............................................................................ 68
6.1.3 Respondents by Age ................................................................................. 69
6.1.4 Respondents by Province of residence ..................................................... 70
6.1.5 Respondents by Education level ............................................................... 72
6.1.6 Respondents by Employment status ......................................................... 73
6.1.7 Respondents by monthly income level ..................................................... 74
6.1.8 Mobile Data Services Awareness ............................................................. 76
6.2 Statistical analysis of data ................................................................................. 77
6.2.1 Utilitarian model testing using SMS ......................................................... 78
6.2.2 Hedonic model testing using Mobile Ring tone ....................................... 78
6.3 Hypothesis Testing............................................................................................ 79
6
6.4 Simple liner model building.............................................................................. 89
6.5 Model building .................................................................................................. 95
6.5.1 Utilitarian Product of SMS ....................................................................... 95
6.5.2 Attitude towards adoption ......................................................................... 95
6.5.3 Intension to adopt ...................................................................................... 98
6.5.4 Hedonic Product of Mobile Ringtone ..................................................... 101
6.6 Data Analysis Summary ................................................................................. 104
6.6.1 Utilitarian product – SMS adoption model testing ................................. 104
6.6.2 Hedonic product – Mobile Ringtone adoption model testing ................. 114
7. Discussion ............................................................................................................... 122
8. Recommendations ................................................................................................... 134
10. Future researchReferences .................................................................................. 144
10. References ........................................................................................................... 145
7
List of Tables
Table 1: Mobile Telephony systems ................................................................................. 26 Table 2: Mobile Data Services classification .................................................................... 29 Table 3: Summary of Litreture review - Utilitarian motives ............................................ 50 Table 4: Summary of literature review - Hedonic motives ............................................... 51 Table 5: : Literature review summary - Attitude and intension ........................................ 51 Table 6: Literature review summary - MDS with utilitarian and hedonic propositions ... 51 Table 7: Hypothesis for utilitarian motives in SMS ......................................................... 54 Table 8: Hypothesis for hedonic motives in SMS ............................................................ 55 Table 9: Hypothesis of social influences in SMS ............................................................. 56 Table 10: Hypothesis attitude and intension in SMS ........................................................ 56 Table 11: Hypothesis for utilitarian motives - M-Ringtone.............................................. 57 Table 12: Hypothesis for hedonic motives in M-Ringtones ............................................. 58 Table 13: Hypothesis for Social influences - M-Ringtones .............................................. 58 Table 14: Hypothesis Attitude and intesion - M-Ringtone ............................................... 59 Table 15: Questionnaire distribution ................................................................................ 64 Table 16: Respondents by Gender .................................................................................... 68 Table 17 : Respondents by Age ........................................................................................ 69 Table 18: Respondents by Province of residence ............................................................. 71 Table 19: Respondents by Education level ....................................................................... 72 Table 20: Respondents by Employment status ................................................................. 73 Table 21: Respondents by monthly income level ............................................................. 74 Table 22: Colour Display vs Black/White display ........................................................... 75 Table 23: Mobile Data Services Awareness ..................................................................... 76 Table 24: Test values for internal consistency – SMS ...................................................... 78 Table 25: Test values for internal consistency - M-Ringtones ......................................... 78 Table 26: Correlation Matrix for SMS.............................................................................. 79 Table 27: Utilitarian model testing using SMS................................................................. 82 Table 28: List of accepted hypothesis (alternative) – Utilitarian product ........................ 83 Table 29: List of Accepted Null Hypothesis..................................................................... 83 Table 30: Correlation Matrix for hedonic motives ........................................................... 84 Table 31: Hypothesis testing for Hedonic model ............................................................. 87 Table 32: List of accepted hypothesis – Hedonic Product ................................................ 88 Table 33: Simple liner model building – SMS ................................................................. 91 Table 34: Simple liner model building - Mobile Ringtones ............................................. 94 Table 35: Variable ranking based on correlation to Attitude towards adoption ............... 95
8
List of Figures
Figure 1: World mobile subscribers .................................................................................. 22 Figure 2: Cellular subscriber growth rate in Sri Lanka ..................................................... 23 Figure 3: Evolution of GSM Technologies ....................................................................... 25 Figure 4 ............................................................................................................................. 28 Figure 5: Proposed classification of MDS ........................................................................ 29 Figure 6: Techno-centric MDS classification ................................................................... 30 Figure 7: Four tiered MDS classification.......................................................................... 31 Figure 8: Classification of consumer value ...................................................................... 34 Figure 9: Proposed model for mobile services adoption in Sri Lanka .............................. 53 Figure 10: Research Onion (Saunders et al, 2007a) ......................................................... 60 Figure 11: Respondents by Gender ................................................................................... 69 Figure 12: Respondents by Age ........................................................................................ 70 Figure 13: Respondents by Province of residence ............................................................ 71 Figure 14: Respondents by Education level...................................................................... 72 Figure 15: Respondents by Employment status ................................................................ 73 Figure 16: Respondents by monthly income level ............................................................ 74 Figure 17: Colour Display vs Black/White display 75
9
Abbreviations
10
1. Introduction
Mobile technology has taken rapid strides in its diffusion across the global. These
quantum leaps in penetration are not only global phenomenon but one also experienced in
the local context of Sri Lanka. In 1992 Sri Lanka had 2,644 mobile phone subscribers.
Today 17 years later the number stands at 11 million (TRC-SL 2008). While mobile
penetration rates are impressive, with 50%-60% average annual growth rates experienced
in Sri Lanka, the strategic prospects of the mobile telecommunication industry are up for
discussion. What comes after you have sold every one a mobile phone?. Signs are
ominous. Across the globe the average revenue per unit (ARPU) are significantly
depreciating (ABI Research 2009; Mälarstig et al. 2007). These issues are compounded
with increase competitive structures and global market competition. The industry seized
on an emerged opportunity in the early 1990 with a new application called Short
Messaging Service. The mobile phone and its use were viewed in a different light than a
simple communication device, rather the gateway to a plethora mobile data services. The
industry spent the next decade investing in high bandwidth, high capacity and new
mobile data services product lines, awaiting the next killer application (C. Carlsson et al.
2005b). However, today after spending billions of dollars into 3G licenses and
sophisticated new services such as MMS, Mobile Internet, Mobile Banking, the “next
killer application” is yet to emerge. SMS still remains the most popular mobile data
service in all markets including the USA(Nielsen Research 2008) and European markets
such as Finland (C. Carlsson et al. 2005b) and Norway(Nysveen et al. 2005b). While
academics and industry in developed countries have focused on studying mobile data
services with new vigor, in developing Countries like Sri Lanka, industry and regulators
seem to be unaware of these global trends and threats.
The aim of this research is analyze the key variables involved in understanding and
predicting consumer behavior of technology adoption. Through this analsys, it is
expected that a behavioral model can be produced which can be identify scientifically the
relationships between the drivers of consumer attitude to adopt and intension to adopt
mobile data services. While there are models researched and developed in countries like
11
Finland (C. Carlsson et al. 2006), Norway (Pedersen et al. 2002), Korea (B. Kim et al.
2009) and USA, development of an indigenous technology adoption model is essential in
the context of Sri Lanka because of the different socio-economic cultural paradigms.
Further due to the regional similarities in South East Asia, the inter-portability of this
model may help diffusion of mobile data services in similar regional countries.
To undertake this study we recommend identifying key research into information
technology adoption including empirically tested models such as the Technology
Adoption Model (Davis 1989; Davis et al. 1989) and diffusion of innovation models
(Rogers 2005). Further as this model involves operations within the consumer context, it
is proposed that research into better understanding the variables that influence the attitude
towards adoption and intension to adopt be researched. Further, the recent research done
on developing a unified theory for technology adoption (Kulviwat et al. 2007; Kulviwat
et al. 2008; Nasco et al. 2008) provides an important starting point. Therefore it was
decided that the study would focus on the logical motives and hedonic motives of ‘Fun
and entertainment’. While motives guide the decision, what nature of value propositions
influence these motives. The second focus of the research would be on value propositions
and their interrelation to technology adoption.
Based on this analysis it was decided that the research would study two mobile data
services products. One which has primarily a cognitive utilitarian value proposition and
another that has primarily a hedonic value proposition. This research would then enable a
better understanding of the behavior of the model in these different context. The balance
of this document will relate to the building of the proposed model based on empirical
research and testing of the model in the context the Sri Lankan consumer through a
market survey. It is expected that this research path would enable the achievement of this
ultimate objective.
12
Problem overview
Problem statement The mobile data services adoption in Sri Lanka remains at a very low rate in comparison
to the penetration of mobile phone technology which is estimated to be at 55% (TRC-SL
2008). Research indicates that the future revenues of mobile telecommunication industry
will depend on the provision of mobile data services rather than on voice calls (Kunin et
al. 2005; C. Carlsson et al. 2005b). The dramatic drops in average revenue per user on
voice calls across the globe are an indication of future trends (ABI Research 2009).
Further in most matured telecommunication markets, where mobile penetration has
exceeded 80% reach of the general population, the industry was compelled to look for
more viable sources of revenue other than voice and new subscriber connection fees
(Mälarstig et al. 2007). While the strategic response of the mobile industry was to invest
in expensive 3G technology, the global adoption rates of mobile services that use this
platform remains very low.
Aims To proposition an analytical model that identifies the key attitudinal influences involved
in the adoption of selected Mobile Data Services in the Sri Lankan market context. This
model could be used by the Telecommunication industry and Mobile Data Services
application vendors to identify key consumer relationship variables that influence the
adoption and diffusion of their products and services.
Objectives
– To analyze the nature and behavior of existing relationships between cognitive
utilitarian motives, hedonic motives, social influences and their impact on the
consumers attitude and intension to adopt key Mobile Data Services in Sri Lanka
13
– To develop a statistical model that analyses the influence of cognitive utilitarian
motives, hedonic motives and social influences to predict the adoption attitude
and intension to adopt the selected Mobile Data Service of Short Message Service
(SMS) which has a dominant utilitarian value proposition in Sri Lanka
– To develop a statistical model that analyses the influence of cognitive utilitarian
motives, hedonic motives and social influences to predict the adoption attitude
and intension to adopt the selected Mobile Data Service of Mobile Ringtone
which has a dominant hedonic value proposition in Sri Lanka
– To analyze the determinant factors that influence the adoption of Mobile Data
Services based on the developed statistical analysis models for utilitarian and
hedonic products
Justification for selection of Mobile Data Services for research
The mobile telecommunication industry has invested billions of dollars in improve the
network bandwidth capacity, mobile phone capacity and overall infrastructure to support
the expanded usage of mobile devices beyond being simple communication devices (.
The anticipation and excitement was that the introduction of 3G would provide mobile
telephone subscribers access to a vast array of mobile data services. However actual
adoption of Mobile Data Services across global markets remain consistently low. On
commenting on this low rate of adoption of Mobile Data Services, states “Our results are
consistent with previous research. Mobile services still have much less users than
envisioned and their usefulness is being questioned by consumers”. While the low
adoption rates of mobile data services are symptoms of consumer perceptions, (Gilbert &
Kendall 2003) outline the need to change the behavioral patterns of consumers to ensure
viable adoption and usage. They state that “MDS are a current example of technology
enabled discontinuous innovation, similar from the economic and behavioral perspectives
to the Internet. Such innovations will succeed only if adopted by a critical mass”. The
researchers highlight the critical need for creating new value and new behavior patterns
14
to ensure sustainable usage of these innovative products. Elaborating on the behavior,
they state that “..such behaviors include acquiring the enabling technology, learning to
use it, applying it to solve problems or adding value in everyday life, and communicating
what one has learned about it to others”. However industry and academia have been only
starting to recognize the need for identify and build viable and empirically tested
consumer adoption models to enhance the overall adoption of the technology.
The lack of research into mobile data services has been a key issue identified by many
researchers. Umino (2004) in a report on OECD countries notes that there is a general
lack of research into the area of Mobile Data Services both by Government and Industry.
The researcher states “Often mobile data is not yet presented separately from aggregate
data. Industry or government sponsored studies focus only on certain markets or
technologies and definitional constraints make it difficult to compare data across studies.
Further research in this area is worth undertaking”. This lack of research and focus may
be stemming from the industries original concern about new connections. Carlsson et al
(2005) commenting on the lack of industry focus on mobile data services comments
“Gartner Inc. in a recent report still focus on the handset market…. It is strange that not
much is reported on the development of mobile services..”. Thus the evolving nature of
the mobile telecommunication industry, which at its inception presented a value
proposition of a simple communication solution, to todays’ mobile data services, which
are value added services may be key reasons for these gaps in research. While the reasons
behind the lack of research may be varied, this lack of understanding of the mobile data
services phenomena has presented the industry with a major challenge. This challenge
has presented itself as an additional risk into the investments towards mobile data
services in general. Carlsson et al. (2006) observers on this increased risk that has
inherited into the Mobile Data Services market as “Year after year the mobile service
market(s) produce(s) new services and applications that due to complexity or lack of
relevance fail to meet the consumers’ expectations”. Therefore the need for research into
understanding the consumer and technology application has emerged as an important of
the overall mix in the product development cycle of mobile data services.
15
Therefore there is a clear need to understand the consumers adoption and usage
preferences towards mobile data service. A number of researchers have explored the
applicability of different psychological models within the context of mobile data services.
These notable researchers include, but are not limited to, Carlsson et al. (2006) in the
Finnish mobile services market, Pedersen et al. (2008) in the Norwegian market, Bina et
al. (2007) in the Greek market. However, when exploring the existing research, it is clear
that research are presented within predominantly from Western and European countries
which have mobile penetration rates exceeding 80%.
There are only few research that has been done to analyze the consumers of South East
Asia including consumers within Sri Lanka, India, Bangaladesh and other East Asian
emerging economies. Commenting on the lack of research into developing countries Gao
& Rafiq (2009) state that “We lack knowledge about the characteristics of mobile
telecommunications transformation in developing countries, and the social and
technological factors that impact this process”. In their litreture review covering a period
of 5 years between 2003 and 2008 they have found eight published articles on mobile
telecommunication industry in developing countries. However the critical features of this
analysis are that these literatures have been prepared based on secondary research and not
primary research.
The huge value of investments made into building 3G+ networks and the ubiquitous (ref)
nature of mobile phone technology presents both a threat and opportunity for a country
like Sri Lanka. Emerging from a three decade old conflict situation, the mobile
technology has a huge potential of enriching and thrusting the rural agrarian economy of
the island rapidly into the 21st century. Mobile Internet, Video calls and other ranges of
mobile data services through 3G + networks would provide the stimulus and hope to our
country. However, understanding the intentions and barriers of the Sri Lankan consumer
in the adoption of this innovative technology is crucial for the penetration of mobile data
services products and services in Sri Lanka. It is based on these reasons that this research
area was selected.
16
Significance of the study
Theoretical significance The research conducted under this project seeks to contribute towards a number academic
interest areas.
Significance of the market data for academic study
The research into adoption of mobile data services is considered by many researchers as a
key gap area in the existing knowledge. Further majority of the available research has
been undertaken in developed markets such as in the USA, Europe and developed
economies in Asia such as Japan and Korea. Therefore this research into the use,
adoption and adoption intension of mobile data services in a country such as Sri Lanka
will be beneficial to understanding the attitudes of a population which has unique
demographic and psychographic characteristics. These characteristics include the high
literacy rate of 96% (G15 2008), the rising Gross Domestic Income of over US$ 1200
which has risen by US$ 150 within the last 3 years, the low computer penetration rate of
and internet penetration of 2% (G15 2008). These characteristics combined with the
estimated mobile phone penetration rate of 54%(TRC-SL 2008) makes this research in to
the study of mobile data services an important and long-term significant study.
The study cover 6 of the 9 provinces and can be used for province wise analysis.
There is currently no available data for academic analysis of the handset types and
capabilities used by the Sri Lankan consumer. This information is particularly important
because the capacity of the mobile phones carried by the consumers in Sri Lanka should
influence decisions on the types of Mobile Data Services that can be promoted in the
island. Further this information should provide a valuable decision and strategic options
consideration tool for mobile telecommunications companies, on whether their current
strategy of not getting involved in the handset market is compatible with their network
17
investment strategy. In order to successfully launch mobile data services that are accepted
by the consumers, the handsets they use have a major influence on the decision making
process. Therefore it is envisioned that this research would initiate a dialog on this issue.
The research focuses on understanding the existing market share of the five
telecommunication services providers of Sri Lanka.
It studies the switching habits of consumers in terms of change of mobile
telecommunications providers, reasons and switching time frame. The cross referencing
of this information with demographic information of users should provide information
vulnerable market segments that are likely to shift to other telecommunications providers.
The research investigates the current usage of native language features in mobile phones
users. This research information would provide information on the popularity and actual
usage of native language features.
The research focuses on consumer awareness of selected mobile data serives, one time
usage and regular usage and the consumers attitude towards the future adoption of the
services.
Significance of the research proposition and hypothesis testing
The research proposition was built on key Information Systems theories of Technology
Adoption Model, Diffusion of Innovation model and PAD model. The significance of
each of these theories towards the adoption intention in the context of the population of
Sri Lanka will be tested through this research.
The applicability of the Consumer Acceptance of Technology model has not been tested
in a wide national study prior to this research. This would be the first occasion the
propositions applicability is tested within a unique market such as Sri Lanka.
18
Significance to other stakeholders Provide the government and regulators insights into the importance of promoting
mobile data services through policy frameworks based on the key influences
identified through this research
Provide the mobile telecommunication industry a better understanding of the
influences of motives and attitude towards adoption of utilitarian and hedonic
value propositioned mobile data services.
Provide software and related technology developers of mobile data services
applications a model to test their product prototypes prior to expensive releases to
market.
Help in influencing the technology infrastructure investments done by the
government as a part of developing the information communication technology
infrastructure of urban and rural Sri Lanka.
Provide greater insights to brand and marketing managers in investing their
marketing budgets and understanding of societal influences on adoption.
19
Scope and Limitations
The research is undertaken within the geographical boundries of Sri Lanka and
may be unique in its findings
The consumers surveyed were primarily from urban and rural areas of Western,
North Western, Southern and Central Provinces
Consumers from other provinces including Northern, Eastern, Uva and North
Central province have not contributed
Due to the small sample size cannot provide analysis at provincial level
implications of the model
20
2. Background
2.1 Mobile telecommunication industry overview The global penetration of mobile phones reached a new height by the end of 2008 when
the International Telecommunication Union (ITU) declared that it estimates the global
telecommunication subscriber base to be 4 billion ITU (2008). This estimated figure is an
increase of over 1 billion mobile subscribers within a period of one year (ITU, 2007). In
late 2007 the global mobile subscriber base was estimated to be around 3 billion
subscribers and was equivalent to 50% of the global population. The year-on-year
average growth of the global mobile telecommunication subscription between the years
2000 to 2008 has been at an average of 24%. While these figures would indicate that the
global penetration of mobile phones are at 61% and that on average every other person
should have a mobile phone, the information needs to be qualified. Its is noted that the
figures represent subscriptions and not actual persons, an individual may have multiple
subscriptions and mobile phone operators methods of counting the prepaid and post paid
consumer may create duplication. Noting this point, it is estimated that over 30 countries,
predominantly in Europe have mobile penetration rates exceeding their country
populations, the highest being Italy at the rate of 151 subscribers for a population of 100
in 2009(ITU, 2009). While these qualifications are valid, Ratan et al. (2007) in their
research of the Bangladesh mobile market note that through ‘Village Phone Program’
each village is provided with a single mobile phone which is shared between a number of
persons.
These impressive mobile phone penetration figures combined with the analysis by ITU
(2009) that Mobile subscriptions accounted for 61% of the total communication
subscriptions, while standard phone line subscriptions were at a low 26% solidifies the
future importance of mobile technology. Further compounding this trend is the increase
in the average usage minutes of mobile phones. The ITU (2009) analysis of average
minute usage suggests that the number of minutes spent by subscribers on mobile phones
are rising while the usage minutes of fixed phones are reducing. Another important
21
observation in this analysis is that users of fixed phone lines are spending an increased
number of minutes communicating with mobile phone subscribers. Other important
global trends are in the dramatic reduction in prices of mobile calls. The estimates
indicate that there is an average reduction of over 20% in call charges associated with
mobile phones.
Figure 1: World mobile subscribers
Extracted from ITU (2008)
While the global penetration rates of mobile phones are impressive, these figures are
sustained primarily through the four BRIC countries of Brazil, Russia, India and China.
Based on estimates the total subscription rates of these economies have an estimated 1.3
billion subscription. ie. One third of the world mobile phone subscribers. While China
has over 600 million subscribers, the Indian subscriber rate is estimated to be 296 million.
This represents a very low penetration rate of 20% in comparison with BRIC countries
and regional countries such as Sri Lanka which has an estimated penetration of 55% by
2008. However, these figures indicate that the mobile penetration and growth will remain
healthy over the next few years in the region.
22
Figure 2: Cellular subscriber growth rate in Sri Lanka
Extracted from TRC-SL (2008)
In Sri Lanka the mobile phone penetration rate has been at a dramatic pace and has
mimicked the global trends closely. In terms of the supply side there are five mobile
telecommunication companies with one new entrant Airtel coming to the market in early
2009 (TRC-SL, 2008). Between the year 2007, where the mobile phone subscribers were
estimated at 7.9 million and 2008 where the figure rose to 11 million, the annual
increment year-on-year has been has been 39%. With the country emerging from a three
decade old conflict situation, Sri Lanka would most likely see the mobile penetration
rates reaching over 80% from the existing rate of 55% of the population within the next 3
years. In comparison to these mobile phone penetration figures the fixed access phone
connection has grown by 20% in 2008 to a figure of 3.4 million phones. It is in the year
2001/2002 that the mobile phone connectivity rate surpassed that of mobile phones in Sri
Lanka. An interesting statistic published is the number of pager connections in the island
which stood at over 10,000 in the year 1996 has seen a complete decline by 2005. While
published research is not available, this may be due to popularity of SMS services.
23
While Sri Lanka macro economic indicators such as Gross Domestic production (average
rate of 5% to 6% REF) and Gross Income (US$) have see improvements over the last few
years, its internet penetration rate remains at very low level of 2%. Further there is no
available information on Mobile Data Services usage and related trends.
2.2 Mobile Technology Evolution In order to thoroughly appreciate the current mobile industry issues, risks and the
implications of Mobile Data Services, it is crucial to understand the underlying
technology, reasons for technological evolution, the technology evolutionary path, factors
that pushed and pulled the evolution, the current point and future evolutionary path.
The second generation of mobile phones also known as 2G started appearing in the early
1990s. Kunin et al. (2005) states that “Most 2G standards are based on circuit-switched
technology, and they have provided the mobile telecommunications industry with an
exponential growth in terms of the numbers of subscribers as well as new types of
services”. Among the most successful technology variants of the 2G included technology
standards such as CDMA (Code Division Multiple Access), TDMA (Time Division
Multiple Access), Global System for Mobile (GSM). The CDMA technology is a digital
wireless technology that has the capability to provide simultaneous access for subscribers
to share radio frequency. The researcher describing some of the distinguishing features of
CDMA “a voice or data call is assigned a unique code that distinguishes it from others
and all of the signals hop and spread over a shared frequency band”. Kunin et al. (2005)
states that as of 2004 CDMA based mobile telecommunication systems were operational
across 63 countries and services an estimated 200 million users. Originally known as the
IS-54 standard, TDMA technical platform has the capability of delivering as much as six
times more information using the same bandwidth than the first generation analog
technology. It is estimated that the TDMA technology which was simultaneously
developed and implemented with CDMA technology serves approximately 113 million
subscribers. The GSM technology is considered the most widely adopted platform in the
2G family. It uses a combination of Frequency Division Multiple Access (FDMA) and
24
Time Division Multiple Access (TDMA). This technology has the capacity to deliver
over eight calls over a single channel.
These underlying technologies supported the deployment of a range of value added
services other than voice. Carlsson et al. (2005) in their analysis of the evolution of
mobile applications identifies that SMS which was available with GSM platform since
the early 1990s started to become unexpectedly popular by 1995. Mobile based internet
browsing services was enabled by 1999 through the deployment of Wireless Application
Protocol (WAP) over the GSM networks. While WAP was introduced aimed at linking
the internet with mobile devices its performance and willingness by subscribers to adopt
the technology was poor (Teo & Pok, 2003).
Figure 3: Evolution of GSM Technologies
Extracted from Carlsson et al. (2005) Continuous technology upgrades to the 2G platform continued since its introduction.
These technology upgrades, which positioned between the 2G (GSM) standard and 3G
(UMTS), included enhancements to GSM in the form of General Packet Radio Service
(GPRS) are noted as 2.5G (Carlsson et al. (2005). GPRS is considered a pivitol
technology enhancement as it introduced the concept of “always-on” capability (Kunin et
al. (2005), which mean that users only had to pay for actual downloads instead of
connectivity. Further the use of packet based data transfer meant that the cost of
operating the service was much cheaper than circuit switched networks (Carlsson et al,
2005). OECD commenting on 2.5G technology platform states that “Many operators are
deploying services with these technologies instead of waiting for 3G since they are
capable of delivering many of the 3G services with higher speeds than 2G.”
25
Table 1: Mobile Telephony systems
Extracted from Kunin et al. (2005)
Kunin et al. (2005) commenting on 3G mobile technology states that “is a generic term
for a set of mobile telephony technologies using a set of high-tech infrastructure
networks, handsets, base stations, switches and other equipment to allow high-speed
Internet access, broadband audio-visual services, and voice and data communications”.
While the 3G technology has a wide bandwidth between 128Kbps to 2 Mbps, the
technology has demonstrated much faster speeds. Among the key distinguishing features
of 3G technology is the wider bandwidth that enables the usages of rich mobile data
services applications such as video calls, mobile internet, high quality audio and visual
services delivery to consumers.
Beyond the 3G technology lies the 4G IP based technology, with an estimated speed 10
time more than that of 3G, with the capability of handling “volatile traffic patterns such
as multiple transmissions of multimedia messages from camera phones” (OECD 2005).
26
2.3 Mobile Data Services
The global Revenue derived from Mobile Data Services have exceeded US$ 200 billion
in the year 2008. This rise in income is an increase of over US$ 43 billion from the
previous year, an estimated increase of over 22% (Cellular-news, 2008). These revenue
figures represent approximately 20% of the total revenue earned by telecommunications
providers. The Filipino telecom provider Smart Communications recorded 50% of their
total earnings from Mobile Data Services. These revenue trends indicate the important
role that Mobile Data Services will play in the future telecom market. While Short
Messaging Services were the initial driver of growth, the industry has been searching for
new “killer applications” which leverage the network capacities setup through the
institution of 3G technology (C. Carlsson et al. 2005). It is therefore anticipated that the
Mobile Data Services would be the driver of growth in the telecommunication industries
where mobile penetration has achieved saturation level.
Bina et al. (2007) in defining Mobile Data Services states that “encompass all non-voice
value-adding services accessible through mobile networks that are designated to augment
end-user experience with mobility and enrich mobile business models for operators,
service providers and other industry constituents”. While this is a general all
encompassing definition, researchers have sought to better define and understand Mobile
Data Services through consistent study. Kunin et al. (2005) in their early study of Mobile
Data Services sought to categorize them into communications, transactional and content
based services. While this classification attempts to identify the Mobile Data Services
from a technology perspective, it lacks the detailed classifications and categorizations
necessary for detailed study of the products and services. Further the classification by the
researcher is based on technology criteria and not a consumer centric perspective. This
classification lacks the depth of analysis and attempts to basket all mobile data services
into one group. However, for the development and positioning of Mobile Data Services it
is crucial that better understanding and analysis of the portfolio be undertaken.
27
Figure 4
Extracted from Kunin et al. (2005)
Carlsson et al. (2005) in their analysis of Mobile Data Services in the Finnish market
have rarely attempted to define Mobile Data Services. Rather their focus has been on the
adoption of the technology and therefore the Mobile Services they have used have been
categorized in a more practical classification. Namely, the MDS have been classified into
Communication, Entertainment, Reservation and purchases, and Information. Into these
four major classifications of the services, they have incorporated a total of six-teen (16)
services. However, the problems associated with the definition of mobile data services
could be highlighted through such classification. Under Communication product ranges
the researcher has included SMS services which are primarily interpersonal in nature.
However, Bina et al. (2007) in their definition of MDS specifically state that “all services
afforded through a mobile network except for voice communication and interpersonal
SMS exchanges”. While the researchers have not commented further on this exception, it
is clear from the analysis that they view MDS in the context of business value creation.
However, not withstanding this interpretation by the researcher, SMS is considered to be
the most popular MDS and the foundation of todays’ recognition and pursuit of killer
applicationsCarlsson et al. (2005a). Further commenting on the popularity of the MDS
products in the United States, Nielsen Research (2008) identified that 53% of the US
consumers were using SMS services as oppose to the next most popular MDS which was
MMS has a subscriber base of 26%. In research done in the Finnish mobile market where
penetration rates have exceeded 80%, over 92% of mobile users regularly use SMS
28
Carlsson et al. (2006) Therefore the exclusion done by Bina et al. (2007) points to the
need to study the context and spatiality of MDS.
Table 2: Mobile Data Services classification Extracted from Gilbert & Han (2004)
A more comprehensive analysis matrix of MDS was presented by Pedersen et al. (2002)
in their analysis of the Norwegian MDS consumer. The matrix attempted to classify MDS
based on the perceived motives and technology characteristics of MDS. The technology
characteristics used by the researchers are communication and transaction. These
dimensions of MDS are cross matched with purpose of usage, where the researchers
introduce the motives of entertainment and utility. This classification is considered by
many researchers as one of the most important cross combinations used in the analysis of
MDS (Nysveen et al., 2005).
Figure 5: Proposed classification of MDS
Extracted from Pedersen et al. (2002)
29
In contrast to the Communication Vs Transaction and Utilitarian Vs Entertainment
classifications of MDS of Pedersen et al. (2002), Verkasalo (2006) seeks to classify
services based on a technology based classification. He uses the continuums of
Communication Vs Content and Interactive vs Background traffic dimensionality.
However, this classification is also primarily a technology centric analysis of these
portfolios of MDS. Beyond this classification of MDS Verkasalo (2006) presents a more
detailed classification of MDS operating on symbion operating systems. Here the main
categories for the classification of MDS include Browsing, Config, Games, Infotainment,
Messaging, Multimedia, Personal Information Management, Productivity, Unknown and
Utility. While these classification relate to applications available in mobile phones, the
products included as part of the analysis relate to MDS. The researchers definition of
MDS as “mobile services which are based on the IP architecture” confirms the
concentration on technology in stead of consumer perspective or service delivery
perspectives.
Figure 6: Techno-centric MDS classification
Extracted from Verkasalo (2006)
30
However one of the most comprehensive analysis and classifications of MDS was
presented by Heinonen and Pura (2006). Their complex analysis of MDS attempts to
classify MDS based on type of “consumption, the context, the social setting and
relationship”. Unlike the classifications of MDS by Verkasalo (2006) which was
primarily technology centric, the researchers attempt view MDS from a consumer service
context. In their criticism of the existing literature on MDS classification, they point-out
that no significant effort has been undertaken to study the classifications of MDS, rather
the existing literature have been produced as a part of a specific aspect of study of the
MDS in terms of intension to use, segmentations, sociability etc.
Figure 7: Four tiered MDS classification
Extracted from Heinonen and Pura (2006)
31
3. Literature review
3.1 Overview of the selected research area Significant research and wide body of knowledge has developed over the past years on
the cogitation and ruminative research of Mobile Data Services adoption across the
globe, in relation to identified markets and on specific mobile data services context.
While the nature of research have been multifarious including industry researchers,
behavioral and social scientists contributing their perspectives, the key thrust area of the
research has remained focused on understanding the adoption of these range of
innovative mobile product and services by the consumer of mobile telephony. Primarily
two significant schools of thought have emerged as the benchmarks for these studies,
namely diffusion research (Rogers 2005) and adoption research (Davis 1989). However
when commenting on research paradigms, it should be noted that compelling research
have been also been undertaken on other promising research directions including (Bina et
al. 2007) on the Triandis (1980) model, the application of Uses and gratifications
research and domestication research by (Pedersen et al. 2002), fit-viability model
proposed by Tjan (2001) which combine the theory of technology and task fit within an
organization, Self-efficacy Theory (Bandura, 2001) were considered during the initial
phased of the LT review.
Excogitating the propositions of the above research, the Consumer Acceptance of
Technology (Kulviwat et al. 2007) distinguishes itself by attempting to build the model
by balancing the logical utilitarian elements of the adoption research (Davis 1989; Rogers
2005) with theories of emotion and affect (Mehrabian & Russell, 1974), to present a
unified theory on technology adoption. The application of this unified theory presents a
potentially powerful prediction and consumer explanation model. This chapter of the
literature review focuses on exploring the critical aspects of the conceptual model
propositioned by this dissertation through a through analysis of the key constructs. It is
hoped that this process would further validate the suitability of exploring the adoption of
32
Mobile Data Services based on the Consumer Acceptance of Technology theory and
indigenous industry specific variables.
3.2 Review of literature on research subject
3.2.1 Motives – Utility vs Hedonics The motives of utility and hedonics formulate a significant composition of the
proposition hypnotized by this research into MDS. This section of the Literature Review
attempts to provide an analysis and definition to these terms.
Understanding the ‘perceived value’ or the benefits customers intend to derive by
acquiring a product or service has been one of the most researched areas in marketing
theory. The decision by Marketing Sciences Institute (2006) to earmark the definition of
‘value’ as a priority research area highlights the continuing and evolving importance of
the subject. Fernández & Bonillo (2007) in their review of research on the subject
observe that, ‘perceived value’ is a result of “interaction” between the customer and the
selected the product or service. Therefore understanding the motives that drive and
influence this interaction is essential in the context of any exchange between a customer
and the provisioning of products or services. On motives and the nature of ‘perceived
value’, the researchers indicate that it may be “...preferential, perceptual, and cognitive-
affective”. It should therefore be appreciated that utilitarian and hedonic motives are only
two key motives that are part of a large portfolio of possible motives that underline the
consumers buying decision. Fernández & Bonillo (2007) identify a large body of research
into ‘perceived value’ while categorizing them into uni-dimensional and multi-
dimensional approaches. While they differentiate between the uni-dimensional and multi-
dimensional approaches because the former propositions a single overall measure to
‘perceived value’, while the latter accepts that multiple components may be used to
define value. However a more pertinent observation between these two classifications is
the evolution of importance placed on utilitarian motives in the more classical uni-
dimensional research and the emerging emphasis of hedonics in multi-dimensional
33
research. It is indeed surprising to note this same evolution of emphasis on utilitarian
motives to hedonics in information systems theory. The once dominant theories such as
Technology Acceptance Model (Davis 1989; Davis et al. 1989) which proposition the
importance of utilitarian motives have now started to incorporate and accept hedonics as
boundary conditions (Heijden 2004; Venkatesh 2000) . Indeed as Ayyagari (2006) notes
on the problems raised due to key Information Systems research such as TAM not
incorporating hedonics, “this might undermine the cumulative results of TAM studies
over the past decade”. Therefore Information Systems researchers such as Kulviwat et al.
(2007) and MDS researchers (C. Carlsson et al. 2005a; C. Carlsson et al. 2005b; C.
Carlsson et al. 2006; Bina et al. 2007; Heinonen & Pura 2006; Nysveen et al. 2005b)
have continued to incorporate hedonics to improve the prediction capabilities of their
research constructs.
Figure 8: Classification of consumer value
Extracted from Fernández & Bonillo (2007)
When considering the different motives that influence consumer decisions, the research
undertaken by Sheth et al. (1991) on consumer value, which has been classified by
34
Fernández & Bonillo (2007) as multi-dimensional, identifies five key ‘values’ that
influence the choice of consumers. Namely, functional, conditional, social, emotional and
epistemic values. It is however important to note that the researcher defines functional
value by stating “ ..functional, utilitarian and physical performance”, this statement
underpins the utilitarian motive selected as a part of the MDS research. Further in
defining emotional value, the researcher states “..arouse feeling or affective state”. It
should be noted that hedonic motives are also known as affect and are part of this
research into MDS. It is therefore necessary to appreciate that the two motives of utility
and hedonics considered as part of this research have significant and empirical theoretical
bases.
Fernández & Bonillo (2007) in defining utilitarian value based on Babin et al. (1994)
research as “instrumental, task-related, rational, functional, cognitive, and a means to an
end” and hedonic value as “reflecting the entertainment and emotional worth of shopping
non-instrumental, experiential, and affective”. hedonic value derived from the usage of a
product or service could be identified with fun or entertainment motive. Bina et al. (2007)
in defining affect “the feelings of joy, elation, or pleasure, or depression, disgust,
displeasure, or hate associated by an individual with a particular act”.
3.2.2 Technology adoption models and Mobile Data Services adoption
The Technology Adoption Model (Davis 1989) is one of the most widely used models in
explaining user adoption behavior in relation to innovative technologies especially within
the context of mandatory settings (Pedersen et al. 2008). Technology Acceptance Model
(TAM) proposed by Davis (1989) conjectured that an individuals cognitive behavioral
intent to adopt a given technology is influenced by two main constructs of perceptions,
namely, perceived usefulness and perceived ease of use. TAM further postulates the
significance of behavioral intension on the attitude of the individual towards adoption.
Davis (1989) defines perceived usefulness as “the degree to which a person believes that
using a particular system would enhance his or her job performance” and perceived ease
35
of use as “the degree to which a person believes that using a particular system would be
free of effort”. The significance of these definitions buttress on the individuals
perceptions and not if the system in consideration is actually useful or easy to use.
Although the original postulation of TAM has been used to research and explain users
intention to use in organizational or mandatory context, Davis et al. (1989) describe the
universal adoptability of the TAM variable in computer and information systems by
users. However, the emphasis of cognitive process and its application within mandatory
settings has meant that researchers (Pedersen et al. 2002) have concluded that TAM is a
utilitarian theory on adoption of technology.
The original construct of TAM is primarily based on Theory of Reasoned Action (TRA)
proposed by Fishbein and Ajzen (1975). The application of TRA is general in comparison
to TAM, and focuses on explaining conscious behavior (Davis et al., 1989). Out of the
four variables identified in the TRA model, namely, Attitude towards behavior
(influenced by Beliefs and Evaluations), Subjective Norms (influenced by Normative
Beliefs and Motivations to comply), Behavioral Intension and actual behavior, TAM
focuses on the variables of Attitude toward use and Behavioral Intension. Taylor & Todd
(1995) in their evaluation of this proposition of Davis et al. (1989) suggest that TAM is a
special case of TRA in its application within technology adoption context. While Davis et
al. (1989) invited research into the investigation of influences of social influences, the
exclusion of this variable in the TAM due to lack of evidence of influence, remained a
point of vigorous discussion by researchers. Researches such as Mathieson (1991)
findings supported the assertions of Davis et al. (1989) on the exclusion of the subjective
norm variable within mandatory setting. However, recent studies by the original authors
and other researchers have significantly changed this proposition. Venkatesh & Davis
(2000) in their extension of the Technology Acceptance Model included the variable of
subjective norm. Further Researches such as Venkatesh & Morris (2000), Lucas & Spitler
(1999) have supported this inclusion of the social norms variable as their individual
research has identified strong influences between this variable and attitude towards
adoption within mandatory settings. It should also be noted that Theory of Planned
Behavior (TPB) is an extension to TRA by Ajzen (1991) and proposes the variable
36
“behavioral control” to explain instances where the individuals behavior is influenced by
internal and external constraints. This inclusion of behavioral control variable has
significantly improved the predictive power of TBP considering the fact that Behavioral
intension is explain as a weighted factor of intension to use and behavioral control
(Taylor & Todd 1995).
The original TAM theory has been extensively changed and modified to improve the
validity of its predicting capability. Venkatesh & Davis (2000) included subjective norms
as “peer pressure”, that influence the persons beliefs in using the IS. Venkatesh et al.
(2003) proposed Unified Theory of Acceptance and Use of Technology (UTAUT) claims
to explain over seventy percent of variance in intention of usage behavior in both
voluntary and non-voluntary settings. However, the application of the TAM theory within
mandatory and organizational setting has meant that TAM has been categorized as a
rational, cognitive theory (Pedersen et al. 2002). (Kulviwat et al. 2007) in their construct
of Consumer Acceptance of Technology model, have pointed out that in two research
undertaken by Davis et al. (1992) and (Riemenschneider et al. 2002), the construct of
affect has been deliberately excluded, as the researchers believed that the inclusion of
hedonic variable was inappropriate within organization settings. The consistent
exclusions of affect from the primary proposition of TAM and its various subsequent
flavors have meant that researchers seeking to understand consumer behavior, which is in
many regards voluntary, have supplemented the main TAM construct. Pedersen et al.
(2002) in their analysis of E-commerce and Mobile data services adoption have used
domestication research (Haddon, 2001)(as cited by Pedersen et al. (2002) and uses and
gratification research (Leung & Wei 2000), whereas, the Consumer Acceptance of
Technology theory has used TAM with the Pleasure, Arousal and Dominance theory of
Mehrabian & Russell (1974).
Kulviwat et al. (2007) et al contend in their analysis that theories such as diffusion of
innovation (Rogers, 1995) and TAM ((Davis et al.,1989)) in their application within
consumer adoption of innovations have not considered the impact of affect, rather depend
on cognition to fully explain behavior. Heijden (2004) and Venkatesh (2000) have
37
attempted to incorporate non-utilitarian aspects into TAM, their main problem has been
that they have been built on the cognitive model. Bina et al. (2007) criticize these
developments by pointing out that “they do not differentiate the affective from the cognitive
dimension and further assume that a person is located on an affective and cognitive bipolar
evaluative dimension”. Kulviwat et al. (2008) et al highlight the implications in identifying
the moderating influence of the nature of task the individual engaged in, whether it be
hedonic or utilitarian on the acceptance of technology. An individuals cognitive process will
be influenced by either utilitarian motives or hedonic based on the intension and experience
they may have derived prior to adopting the technology. Thus, the intension of individuals
may be equally influenced by hedonic and utilitarian motives. Therefore, in voluntary
settings the exclusion of either motive may not provide a strong construct of evaluating
consumer acceptance of technology. Kulviwat et al. (2008) in defining the utilitarian task
identifies that the task orientation primarily problem solving. This cognitive process
therefore influenced by logical, reason based approach. The need for including affect in
predicting the behavior of consumers was proposed by a number of theories such as the
Triandis (1980) and propositioned by Bina et al. (2007) in relation to Mobile Data
Services. In defining affect “the feelings of joy, elation, or pleasure, or depression,
disgust, displeasure, or hate associated by an individual with a particular act”. Bina et
al. (2007) uses the triandis theory to propose an alternative approach to analyzing the
adoption of mobile data services. Further leading researchers on mobile data services
such as Carlsson et al. (2005) use hedonic factors such as enjoyment to identify consumer
motives, while using TAM as the main construct of the research. Pedersen et al. (2002)
look to the hedonic variables of uses and gratification research to partially explain the
adoption of MDS.
Researchers on the adoption of mobile data services have been using a variety if variables
to construct the influence of hedonic variables on MDS. These variety of constructs to
monitor hedonics have not been limited to MDS but researchers in variety of fields such
as Electronic commerce, telecommunications etc. have been focusing on this regard.
Carlsson et al. (2005) uses two hedonic variables of “Enjoyment” and “new possibilities”
as the basis of evaluating potential user preferences for adoption of mobile data services.
38
Bina et al. (2007) in incorporating the hedonic variable assessment criteria identify “fun”,
“enjoyment”, “killing time” as the potential candidate emotions towards adoption. In
contrast to these simple approaches, Pedersen et al. (2002) incorporate the uses and
gratification research to correlate the hedonic variable with adoption. While gratification
research is capable of identifying a wide range of gratifications such that was identified
by Leung & Wei (2000) including “fun-seeking”, “entertainment”, fashion and status”,
both these research point to the fact that the emotion continuum of humans are wide and
need to be captured within model that can present it within a parsimonious and
manageable content. Kulviwat et al. (2007) propositions the Consumer Acceptance of
Technology by combining the three dimensions of the Pleasure, Arousal and Domination
model (PAD) by Mehrabian & Russell (1974) to fill the vacuum in the monitoring
construct for affect. The methodology proposed by Kulviwat et al. (2007) to analyze the
affect is through the environmental psychology theory of pleasure, arousal, and
dominance (PAD) by Mehrabian and Russell’s (1974). These researchers contend that the
emotional response signaled by an individual the physical environment and social
environment can be measured within the dimensions of pleasure, arousal, and dominance.
The emotional response of the individual is mapped as a point within the three
dimensions of the PAD variables. The main basis of the Consumer Acceptance of
Technology (CAT) theory is the premise that “Consumers may adopt high-technology
products not only to obtain useful benefits but also to enjoy the experience of using
them”(Kulviwat et al. 2007). Thus unlike the TAM and its related TRB and TPB, CAT
the prediction of consumer adoption of an innovation, especially in the context of
consumer items, the role of affect should be taken into consideration.
The incorporation of relative advantage as a variable that influences the cognitive
utilitarian decision is another distinctive features of the CAT model. The theory focuses
on improving the cognitive conceptualization of belief by introducing the variable –
relative advantage. Kulviwat et al. (2007) in describing relative advantage as “relative
advantage means that the innovation is believed by the adopter to be superior in some
way to what it is intended to supersede”. This is an interesting integration of the diffusion
research with that of the TAM.
39
Researchers on innovation and hi-technology adoption have acknowledge the causal
relationship that exists between the recognition of society and impact on attitude to adopt.
Rogers (1995) identified social systems variable including, social system norms,
tolerance of deviancy, communication integration, as one of the key groups of variables
that influence the knowledge variable/dimension of consumer. Venkatesh & Davis (2000)
identified the variable of subjective norms in their extension to the TAM model.
Kulviwat et al. (2008) in their theory recognizes the role of social influences on adoption
behavior. In their research on “private” and “public” consumption and the influence on
attitude, they observe “It seems, therefore, that adoption decisions regarding
technological innovations are more susceptible to social influence when consumption of
the product is visible to others”. This observation has major implications on the
communication strategy of firms towards their products and services.
40
3.3 Literature review on selected independent variables
3.3.1 Independent variable 1 - Perceived usefulness
Davis (1989) in defining perceived usefulness states that it is the degree to which using
an information system is thought to improve the activities they are performing. In the
context of TAM, perceived usefulness is considered to be the most powerful predictor of
behavioral intent (Taylor & Todd 1995). In its original application within organizational
context, this variable represented the individual belief that its adoption and usage would
result in an increased performance of the job (Davis et al. 1989). In fact Davis (1989)
suggests that the variable of perceived usefulness is more important than that of
perceived ease of use, this contention was supported by Hu et al. (1999). However, as
MDS represents the consumer context, the validity of the variable may be debatable.
Bruner II & Kumar (2005) in their research on the applicability of TAM in consumer
context found that usefulness could be considered an important variable even in
consumer context. However, Pedersen et al. (2002) have identified that the prediction
capability of the usefulness variable is more strong based on the task context.
Specifically, that the usefulness is more important in relation to utilitarian MDS such as
text messaging and payment than entertainment services which are more hedonic
dependant. These research findings of stronger prediction capability of utilitarian motives
in relation to perceived usefulness rather than hedonics were confirmed by Nysveen et al.
(2005b) and Nysveen et al. (2005a). Kulviwat et al. (2007) supports this finding on the
nature of influence of perceived usefulness in the context of products used for utilitarian
motives rather than for hedonic purposes.
Within regional settings research done by Kim et al. (2007) into the Singapore Mobile
Internet usages market, Kim et al. (2009) in to the Korean wireless pay-per-view market
and Hong et al. (2006) into the Mobile internet market of Hong Kong have empirically
accepted the role played by perceived usefulness.
41
3.3.2 Independent variable 2 - Perceived ease of use
Jenson (2006) in his critique of the MDS industry points to “default thinking” in
designing and implementing products and services. In his example of the MMS, Jenson
points to the failure of MDS industry to comprehend the value proposition and
complexity of using MMS, and instead proposing it as a natural extension to SMS. This
suggestion of industry pushing forth technology innovations and line extensions without
considering there practical usability and specifically ease of use is propositioned by him
for the failure of many MDS. Carlsson et al. (2006) supports this proposition in the
Finnish MDS market by point to mismatches of expectations between industry experts
and consumers. The survey identifies that while the industry has been introducing new
and more complex applications for MDS, consumers in general have been slow in their
adoption and continued usage of MDS. Davis (1989) defines perceived ease of use as
“the degree to which a person believes that using a particular system would be free of
effort”. This variable points to expectation of effort involved in using product or service.
Kulviwat et al. (2007) while accepting the importance of perceived ease of use as a
determinant in influencing attitude, considers the influence as indirect. They note that
rather than directly influencing the attitude of the user, it has a direct impact on the
perceived usefulness rather than intension directly. This conclusion is empirically tested
by Bruner II & Kumar (2005) who note that perceived ease of use indirectly influences
both the usefulness and fun variables. However the direct influence capability and
indirect influence capability of ease of use in the context of four mobile data services is
recognized by (Nysveen et al. 2005b) and mobile chat services (Nysveen et al. 2005a).
Here too the ease of use was noted to directly influence both attitude toward use and
usefulness.
Significantly Kim et al. (2009) rejects the influence of ease of use in the context of
inexperienced Mobile internet users and experienced users in Korea. However, the
influence of this variable is identified as important in the context of continuing usage
42
intension. This finding of strong influence of the ease of use variable in post adoption
was confirmed by empirical research done by Hong et al. (2006) in Hong Kong.
3.3.3 Independent variable 3 - Relative advantage
Rogers (2005) included relative advantage as a part of the product variables that
influence the diffusion of innovation. In analyzing the characteristics of innovation to
include relative advantage, compatibility, complexity, triability and observability, Rogers
(2005) note that an innovation being “better” than its existing alternatives is essential.
The decision by Kulviwat et al. (2007) to incorporate relative advantage as a variable in
their research model is interesting because of very few new research literature on the
empirical testing of this variable. Plouffe et al. (2001) in testing the Perceived
Components of Innovation model which Moore & Benbasat (1991) proposed, states that
relative advantage is the most important predictor of adoption intension. The focus of the
formers’ research is comparing the prediction capability of Technology Adoption Model
with Perceived Components of Innovation model. They note that there is similarity
between the variable of perceived usefulness and relative advantage variables. It is stated
in their analysis that dependence on perceived usefulness alone may be misleading as this
variable has a number of sub-classifications – including relative advantage. Kulviwat et
al. (2007) acknowledges the lack of empirical research into the influence of relative
advantage to adoption intension in the context of information systems research, especially
in mandatory settings where uses lack the options of comparing information systems.
However, this decision by the researchers to incorporate the relative advantage variable
was important, as this variable emerged as the most important influencer of intension,
less influential than perceived usefulness and more influential than ease of use.
43
3.3.4 Independent variable 4 - Pleasure
While there is a wide body of research that acknowledges hedonic motives (Bina et al.
2007; C. Carlsson et al. 2005; Childers et al. 2001; Heijden 2004; Heinonen & Pura 2006;
Hong et al. 2006; Kim et al. 2009) they do not attempt to proceed beyond motives of fun,
entertainment. The research proposition of Kulviwat et al. (2007) is unique in that they
attempt to develop a more deeper analytical model towards hedonic motives by
incorporating Mehrabian & Russell (1974) empirically tested Pleasure-Arousal-
Dominance scales. Lee et al. (2003) describes the pleasure emotion as “the extent to
which a person feels good”. They note of a number of research which indicate that the
emotion of pleasure, in combination with arousal and dominance, have been identified as
a stimulus in increasing purchasing behavior of customers. The research conducted by
Lee et al. (2003) confirmed the validity of pleasure in the context of online shopping. Wu
et al. (2008) in their research into the influence of pleasure and arousal in the context of
online shopping note the validity of these measures in predicting consumer buying
behavior. Wulf et al. (2006) have developed a comprehensive website evaluation model
using pleasure as the key boundary condition between the evaluation of websites and
their success.
3.3.5 Independent variable 5 - Arousal
Arousal formulates the second bipolar variable in assessing hedonic motives as proposed
by Mehrabian & Russell (1974). This bipolar nature is represented within the continuum
of feeling of being aroused to that of un-aroused. Kulviwat et al. (2007) notes that the
state of arousal is a result of a reaction of an individual to presented stimuli, influenced
primarily by the emotion of excitement. Wu et al. (2008) have identified and incorporated
arousal as an essential element in combination with pleasure to influence use buying
behavior in the online shopping and website designing context. These findings were
confirmed in an earlier research into stimulating consumer buying behavior in internet
shopping malls undertaken by Lee et al. (2003). While these research identify the
44
variable of arousal and its influence and interplay in the consumers buying decision, a
more unique approach to appreciate arousal was proposed by Wirtz et al. (2000). They
introduce the concept of target level arousal as a moderating variable in the satisfaction of
consumers. They proposition that the satisfaction felt by the consumer is based on their
expectation of a given situation or environment. For example, the expectation of the
consumer is selecting a restaurant is for a low arousal experience vs. deciding to go to a
disco is has an embedded high arousal experience. Therefore the level of satisfaction felt
by the consumer is based on the expectation vs actual experience. They empirically
validate this proposition using dinning experience in the Singapore market.
3.3.6 Independent variable 6 - Dominance
The variable of dominance was posited by Russell & Mehrabian (1974) as the third axis
of the PAD dimensional analysis of affect. This bipolar continuum extends from
emotional state of Dominance in which the individual feels greater control over the
innovation to Submissiveness. During the emotional state of Submissiveness the range of
emotion experienced by the individual include those of anger, fear, frustration, confusion
(Russell & Mehrabian, 1977). Kulviwat et al. (2007) when incorporating dominance as
part of the Consumer Acceptance of Technology model noted that there has been
significant debate among researchers on the validity of this variable. This issue of validity
was once again raised when dominance was rejected based on it weak influence on
attitude towards adoption. However, the researchers who propositioned the Consumer
Acceptance of Technology model further investigated the dominance dimension (Nasco
et al. (2008). The researchers note the empirical findings of Yani-de-Soriano & Foxall
(2006) on the role of dominance in the context of consumer setting and their forceful
argument on the validity of this variable. Based on their research into the role of
dominance, they note that in many instances the direct effect of the variable is masked.
45
3.3.7 Independent variable 7 - Social Influences Glotz et al. (2005) in their international review and research into Mobile phones and their
social and cultural usage note that “enabler of social interactions, hierarchies and
communication”. Bina et al. (2007) incorporate the social influences as part of the
research into the Greek market. In defining the social factor “social factors try to capture
the congruency between social norms and individual beliefs and how the human part of
an individual’s environment affects one in performing a specific behavior”. Venkatesh &
Davis (2000) have incorporated ‘subjective norms’ as an extension to the Technology
Adoption Model, in recognition of influence from the cultural and norms. The lead
researchers of Technology Adoption Model also made further research on the moderating
effects of public and private consumption and adoption of technology. The importance of
social influences have also been highlighted in Rogers (2005) Diffusion of innovation
theory. Here social systems variable is a key variable that influences the inception stage,
identified as Knowledge stage, of the adoption process. Among the sub-variables that
have been identified by the researcher are social systems norms, tolerance of deviancy,
communication integration. López-Nicolás et al. (2008) have also validated the social
influences variable in the Dutch consumer market setting. However, they make an
important observation that communication media has a positive influence effect on social
norms. The identification of this relationship is considered important because MDS
vendors can influence the attitude of society enabling greater adoption. Wei (2008) notes
that the general perception in the USA market for mobile phones is primarily as a
communication device. This general perception has created a social barrier towards the
adoption of MDS. Therefore the researcher suggests the usage of media to change this
perception of society.
46
3.3.8 Attitude and Intention
The relationship between attitude and intension information systems research was
established as a result of the work done by Davis (1989); Davis et al. (1989) in the
technology adoption model based primarily on the findings of theory of reasoned action
Fishbein & Ajzen (1975). Kim et al. (2009) defines attitude in terms of Information
Systems as “a psychological tendency expressed by evaluating a particular entity in terms
of the degree of positiveness about IS”. Kulviwat et al. (2007) in analyzing the attitude
toward the act in the context of Technology Adoption Model, identifies the cognitive
dimension of the variable by stating that this “refers to the evaluative judgment of
adopting a piece of technology”. Therefore attitude in the context of its role in
influencing the intension of consumers could be viewed as decision or ‘judgment’.
However, Cohen & Areni (1991)(as cited by Kulviwat et al. (2007)) also point to the fact
that like all human emotions, instead of being completely cognitive, hedonics also play a
role in this judgment. This interplay between cognition and hedonics has also been the
basis of the subsequent proposition of Consumer Acceptance of Technology model by
Kulviwat et al. (2007). However, there has been significant debate among researchers of
information systems research on the validity of the influence of attitude to behavioral
intension. Due to the weak empirical evidence to support the influence of attitude on
intension Venkatesh & Davis (2000) decided to remove this motive from their extended
version of Technology Adoption Model. Researchers such as Adams et al. (1992) proved
that the Technology Adoption Model was robust even without the inclusion of attitude.
While it would have been an easier proposition to drop attitude towards adoption from
the research model, it was noted that majority of the research which decided to remove
attitude from the model mix were done in Information System mandatory settings. This
research focuses on consumer behavior and removing an important predictor could
potentially reduce the explanatory power of the model. It was further noted that through
the work of Bruner II & Kumar (2005) pointed to the mediating role between attitude and
intension. Further researchers such as Ayyagari (2006) have started to identify the
47
presence of hedonics even within the Technology Adoption Model. Therefore both
variable were included as part of this research study.
3.3.9 Short Message Service – Mobile Data Service used to test the cognitive utilitarian value proposition
This research intend to study the behavior and influence of utilitarian and hedonic
motives, social influences on selected mobile data services with utilitarian and hedonic
value propositions. The mobile data service selected to represent utilitarian value
proposition is the popular Short Message Service. This service is known as the “killer
application” which was responsible for the identification of the mobile data services
industry and subsequent investments into 3G technology (C. Carlsson et al. 2005b; Kunin
et al. 2005).Pedersen et al. (2002) in their research into mobile data services classified
SMS as having predominantly utilitarian value propositions. This was primarily due to
the recognition that SMS was used to achieve a specific purpose and the intended value
derived from using the product is task oriented. However, Pedersen et al. (2002) notes
that there is “..potential for entertainment in addition to utility. This classification of SMS
within the value propositions of utility has been confirmed by researchers such as (C.
Carlsson et al. 2005a; C. Carlsson et al. 2005b; Nysveen et al. 2005b). Heinonen & Pura
(2006) in their complex analysis of mobile data services based on consumption, context,
social setting and relationship agree on the classification done by Pedersen et al. (2002)
and Nysveen et al. (2005). It was therefore decided to use SMS as the basis for testing the
behavioural model on utilitarian mobile data services.
3.3.10 Mobile Ringtone – Mobile Data Service used to test the hedonic value proposition
The original selection for hedonic value proposition mobile data services was Mobile
Gaming. However, due to the very low adoption rates among the pilot survey respondent,
it was decided to find a mobile data service which had similar characteristics of value.
Based on the responses from the pilot study it was noted that mobile ringtones were wide
used by the respondents (45%) and therefore used as the basis of testing the research
48
model. Carlsson et al. (2005) in their classification of mobile data services identified that
mobile ringtones are of the entertainment category. Further using the classifications
matrix used by Pedersen et al. (2002) to categorized data services, mobile ringtones can
be considered to fall within the category of “Entertainment/ Transaction” the transaction
classification is relevant to mobile ringtone as they are used based on fee. Further the
hedonic motives of mobile ringtones were accepted by Heinonen & Pura (2006).
However in Verkasalo (2006)’s techno-centric mobile data service classification mobile
ringtone were not recognized. However, it was decided to use this product as the basis of
assessing hedonic value proposition because users adopt this product primarily due to its
entertainment value.
49
Summary of Literature Review
3.3.11 Utilitarian Motives
Variable
Literature
Finding
Perceived Usefulness
Kulviwat et al. (2007)
Taylor & Todd (1995)
Davis et al. (1989)
Davis (1989)
Bruner II & Kumar (2005)
Pedersen et al. (2002)
Nysveen et al. (2005b)
Nysveen et al. (2005a)
Kim et al. (2007)
Kim et al. (2009)
Hong et al. (2006)
Positive Relationship
Perceived Ease of use
Davis (1989)
Carlsson et al. (2006)
Kulviwat et al. (2007)
Bruner II & Kumar (2005)
Nysveen et al. (2005b)
Nysveen et al. (2005a)
Kim et al. (2009)
Hong et al. (2006)
Positive Relationship
Relative Advantage
Rogers (2005)
Kulviwat et al. (2007)
Plouffe et al. (2001)
Moore & Benbasat (1991)
Positive Relationship
Table 3: Summary of Litreture review - Utilitarian motives
50
3.3.12 Hedonic Motives
Variable
Literature
Finding
Pleasure
Kulviwat et al. (2007)
Lee et al. (2003)
Mehrabian & Russell (1974)
Wu et al. (2008)
Wulf et al. (2006)
Positive Relationship
Arousal
Kulviwat et al. (2007)
Wu et al. (2008)
Lee et al. (2003)
Wirtz et al. (2000)
Positive Relationship
Dominance
Nasco et al. (2008)
Yani-de-Soriano & Foxall (2006)
Positive Relationship
Table 4: Summary of literature review - Hedonic motives
Attitude and Intension Variable
Literature
Finding
Attitude and Intension Kulviwat et al. (2007)
Bruner II & Kumar (2005)
Positive relationship
Table 5: : Literature review summary - Attitude and intension Mobile data services with Utilitarian and Hedonic value propositions Mobile Data Service
Literature
Finding
Short Message Service Cognitive utilitarian
value proposition
Mobile Ringtone Hedonic value
proposition
Table 6: Literature review summary - MDS with utilitarian and hedonic propositions
51
4. Solution
4.1 Solution overview The research focuses on two significant issues associated with mobile data service
technology. First is the analysis of existing behavioral relationships between cognitive
utilitarian motives, hedonic motives and social normative influences and attitude towards
adoption and adoption intension. The key variables identified to represent utilitarian
cognitive motives were perceived usefulness, perceived ease of use and comparative
advantage. These motivational influences were selected primarily based on the core
propositions of the Consumer Acceptance of Technology (Kulviwat et al. 2007) which
was based on the Technology Adoption Model (Davis 1989; Davis et al. 1989) and
Diffusion of innovation (Rogers 2005). The variables of Pleasure, Arousal and
Dominance were selected based on the Consumer Acceptance of Technology model. The
second element of the research is the study of the behavior of the selected variables in the
context of Short Message Services (SMS) which has an established utilitarian value
proposition(Pedersen et al. 2002; Nysveen et al. 2005b; C. Carlsson et al. 2006; Heinonen
& Pura 2006) and Mobile Ringtones (C. Carlsson et al. 2005b; C. Carlsson et al. 2005a;
Pedersen et al. 2002; Nysveen et al. 2005a) which has an established hedonic utilitarian
value proposition.
Based on these variables the following conceptual module will be used as the basis of the
two studies.
52
Proposed model for mobile services adoption in Sri Lanka (Sri Lanka Consumer Acceptance of Technology Model – SLCAT)
Utilitarian motives Social Influences
Perceived Usefulness
Ease of Use
Attitude towards adoption
Adoption Intension
Comparative advantage
Hedonic motives
Pleasure
Arousal
Dominance
Figure 9: Proposed model for mobile services adoption in Sri Lanka
53
4.2 List of developed hypothesis Model for analyzing Utilitarian value proposition – Using SMS 4.2.1 Utilitarian motives in the adoption of Utilitarian Services (SMS)
No Variable tested Hypothesis 1. Perceived usefulness in
the context of
Utilitarian Services
adoption and attitude
towards adoption
Hypothesis 1
H01: There is weak influence of perceived usefulness in the context of attitude towards
adoption of utilitarian Mobile Data Services in Sri Lanka
Ha1: There is strong influence of perceived usefulness in the context of attitude towards
adoption of utilitarian Mobile Data Services in Sri Lanka
2. Perceived Ease of use
in the context of
Utilitarian Services
adoption and attitude
towards adoption
Hypothesis 2
H02: There is weak influence of perceived ease of use in the context of attitude towards
adoption of utilitarian Mobile Data Services in Sri Lanka
Ha2: There is strong influence of perceived ease of use in the context of attitude
towards adoption of utilitarian Mobile Data Services in Sri Lanka
3. Relative advantage in
the context of
Utilitarian Services
adoption and attitude
towards adoption
Hypothesis 3
H03: There is weak influence of relative advantage in the context of attitude towards
adoption of utilitarian Mobile Data Services in Sri Lanka
Ha3: There is strong influence of relative advantage in the context of attitude towards
adoption of utilitarian Mobile Data Services in Sri Lanka Table 7: Hypothesis for utilitarian motives in SMS
54
4.2.2 Hedonic motives in the adoption of Utilitarian Services (SMS)
No Variable tested Hypothesis 4. Pleasure in the context
of Utilitarian Services
adoption and attitude
towards adoption
Hypothesis 4
H04: There is weak influence of pleasure in the context of attitude towards adoption of
utilitarian Mobile Data Services in Sri Lanka
Ha4: There is strong influence of pleasure in the context of attitude towards adoption of
utilitarian Mobile Data Services in Sri Lanka
5. Arousal in the context
of Utilitarian Services
adoption and attitude
towards adoption
Hypothesis 5
H05: There is weak influence of arousal in the context of attitude towards adoption of
utilitarian Mobile Data Services in Sri Lanka
Ha5: There is strong influence of arousal in the context of attitude towards adoption of
utilitarian Mobile Data Services in Sri Lanka
6. Dominance in the
context of Utilitarian
Services adoption and
attitude towards
adoption
Hypothesis 6
H06: There is weak dominance of arousal in the context of attitude towards adoption of
adopt utilitarian Mobile Data Services in Sri Lanka
Ha6: There is strong influence of arousal in the context of attitude towards adoption of
utilitarian Mobile Data Services in Sri Lanka Table 8: Hypothesis for hedonic motives in SMS
55
4.2.3 Role of Social Influences in the context of Utilitarian value proposition
No Variable tested Hypothesis 7. Social influence in the
context of Utilitarian
Services adoption and
attitude towards
adoption
Hypothesis 7
H07: There is weak influence of social influences in the context of attitude towards
adoption of utilitarian Mobile Data Services in Sri Lanka
Ha7: There is strong influence of social influences in the context of attitude towards
adoption of utilitarian Mobile Data Services in Sri Lanka Table 9: Hypothesis of social influences in SMS
4.2.4 Relationship between Attitude towards adoption and adoption intension in the
context of Utilitarian value proposition
No Variable tested Hypothesis
8. Relationship between
attitude towards
adoption of Utilitarian
service and Adoption
Intension
Hypothesis 8
H08: There is weak influence of attitude towards adoption and adoption intension in
the context utilitarian Mobile Data Services in Sri Lanka
Ha8: There is strong influence of attitude towards adoption and adoption intension in
the context utilitarian Mobile Data Services in Sri Lanka Table 10: Hypothesis attitude and intension in SMS
56
4.2.5 Model for analyzing Hedonic value proposition – Using Mobile Ringtone
No Variable tested Hypothesis 9. Perceived usefulness in
the context of Hedonic
Services adoption and
attitude towards
adoption
Hypothesis 9
H09: There is weak influence of perceived usefulness in the context of attitude towards
adoption of hedonic Mobile Data Services in Sri Lanka
Ha9: There is strong influence of perceived usefulness in the context of attitude towards
adoption of hedonic Mobile Data Services in Sri Lanka
10. Perceived Ease of use
in the context of
Hedonic Services
adoption and attitude
towards adoption
Hypothesis 10
H010: There is weak influence of perceived ease of use in the context of attitude
towards adoption of hedonic Mobile Data Services in Sri Lanka
Ha10: There is strong influence of perceived ease of use in the context of attitude
towards adoption of hedonic Mobile Data Services in Sri Lanka
11. Relative advantage in
the context of Hedonic
Services adoption and
attitude towards
adoption
Hypothesis 11
H011: There is weak influence of relative advantage in the context of attitude towards
adoption of hedonic Mobile Data Services in Sri Lanka
Ha11: There is strong influence of relative advantage in the context of attitude towards
adoption of hedonic Mobile Data Services in Sri Lanka Table 11: Hypothesis for utilitarian motives - M-Ringtone 4.2.6 Hedonic motives in the adoption of Hedonic Services (Mobile Ringtone)
No Variable tested Hypothesis 12. Pleasure in the context
of Hedonic Services
adoption and attitude
towards adoption
Hypothesis 12
H012: There is weak influence of pleasure in the context of attitude towards adoption of
adopt hedonic Mobile Data Services in Sri Lanka
57
No Variable tested Hypothesis Ha12: There is strong influence of pleasure in the context of attitude towards adoption
of hedonic Mobile Data Services in Sri Lanka
13. Arousal in the context
of Hedonic Services
adoption and attitude
towards adoption
Hypothesis 13
H013: There is weak influence of arousal in the context of attitude towards adoption of
hedonic Mobile Data Services in Sri Lanka
Ha13: There is strong influence of arousal in the context of attitude towards adoption of
hedonic Mobile Data Services in Sri Lanka
14. Dominance in the
context of Hedonic
Services adoption and
attitude towards
adoption
Hypothesis 14
H014: There is weak influence of dominance in the context of attitude towards adoption
of adopt hedonic Mobile Data Services in Sri Lanka
Ha14: There is strong influence of arousal in the context of attitude towards adoption of
hedonic Mobile Data Services in Sri Lanka Table 12: Hypothesis for hedonic motives in M-Ringtones 4.2.7 Role of Social Influences in the context of Hedonic value proposition
No Variable tested Hypothesis 15. Social influence in the
context of Hedonic
Services adoption and
attitude towards
adoption
Hypothesis 15
H015: There is weak influence of social influence in the context of attitude towards
adoption of hedonic Mobile Data Services in Sri Lanka
Ha15: There is strong influence of social influence in the context of attitude towards
adoption of hedonic Mobile Data Services in Sri Lanka Table 13: Hypothesis for Social influences - M-Ringtones
58
4.2.8 Relationship between Attitude towards adoption and adoption intension in the
context of hedonic value proposition
No Variable tested Hypothesis
16. Relationship between
attitude towards
adoption of hedonic
service and Adoption
Intension
Hypothesis 16
H016: There is weak influence of attitude towards adoption and adoption intension in
the context hedonic Mobile Data Services in Sri Lanka
Ha16: There is strong influence of attitude towards adoption and adoption intension in
the context hedonic Mobile Data Services in Sri Lanka Table 14: Hypothesis Attitude and intesion - M-Ringtone
59
5. RESEARCH METHODOLOGY
Research in what ever domain looks at enriching the sea of knowledge expanding its
horizon. However Uma Sekaran (2006, p.5) provides a complete definition to it as “an
organized, systematic, data-based, critical, objective, scientific inquiry or investigation in
to a specific problem undertaken with the purpose of finding answers or solutions to it”.
Saunders et al (2007b, p.5) too provides a definition to research that says it’s something
that people undertake in order to find out something in a systematic way, thereby
increasing their knowledge.
Both these two authors identify research to be carried out in a systematic of a methodical
manner. This makes it clear that the research need to be carried out in a very structured
and planned manner where you have a clear understanding of what is required to do and
how to do it. A methodology is required that identify methods, practices and procedures
that helps to carryout the research. The following chapter identify the research
methodology used to carryout this research.
Figure 10: Research Onion (Saunders et al, 2007a)
60
The following research onion identifies all facets that need to be identified in a research.
The discussion on the methodology would be carried out in accordance to the research
onion looking at each of its layers and discussing how and what was selected for each.
5.1 Research Philosophy
According to the research onion presented by Saunders et al (2007b, 102) the research
philosophies include; Positivism, Realism, Interpretivism, Objectivism, Subjectivism,
Pragmatism, Functionalist, Interpretive, Radical humanist and Radical structuralist. The
research philosophy used for this research can be identified as positivism. The main
reason for the selection of the following philosophy is due to the fact that the research
involves and looks at social aspects in the society and come up with a framework that can
be generalized to a subset of the society.
5.2 Research Approach
The research approach looks at either deductive research, “where you develop a theory
and hypothesis and design a research strategy to test the hypothesis, or inductive
research, in which you would collect data and develop theory as a result of the data
analysis” (Saunders et al 2007b, p.117). The research approach used in this research is
the deductive method of reasoning. This was due to the fact that the research looks at
explaining casual relationships between variables identified it the framework which used
to develop the hypothesis which would be tested with qualitative data.
Two sets of hypothesis were developed to for the two prediction models developed to test
value propositions in cognitive utilitarian mobile data services and hedonic mobile data
services. These hypothesis were developed based on the identified variables of perceived
usefulness, perceived ease of use, comparative advantage, pleasure, arousal, dominance,
social influences, attitude towards adoption and adoption intension.
61
5.3 Research Strategy In terms of the research strategy several strategies are available such as experimental,
case study, action based, ethnography etc. However due to the nature of this research and
its domain the survey method was selected as the strategy for the research in terms of
collecting the data for the testing of hypothesis deduced from the relationships. The main
reason for selecting this strategy was due to the fact that “survey strategy is usually
associated with the deductive approach” (Saunders et al 2007b, p.138). Another fact for
selecting this was due to the nature of collecting data from a large audience and moreover
due to the fact that it is widely used in social and behavioral science research. Another
reason for choosing survey for this research is due to the fact that it can identify the
“relationships between the data and the unknown in universe” (Kothari, 2002) and also
due to the fact that it is more concerned with formulating hypothesis and testing the
association between the relationships (Kothari, 2002).
Saunders et al (2007b, p. 138) identify the following characteristics of surveys as a
research strategy;
Provides the flexibility to collect large quantities of data from a substantial
population
Considered to be as one of the most economical ways of collecting data from a
large audience
Flexibility of using quantitative analysis using descriptive and inferential statistics
Can be used to propose possible reasons for particular relationships between
variables and to produce models
Provides more control over the research process
These characteristics identified above were some of the reasons for the selection of
survey as the research strategy.
62
5.4 Pilot study To reduce inconsistency of the research findings and to reduce the errors, a pilot study
was carried out using 40 students and 10 co workers. The main objective of having the
pilot study was to check the reliability of the research questionnaire. Saunders et al
(2007b, p.149) identify reliability as the “extent to which the data collection technique or
analysis procedure will yield consistent findings”. Reliability threats can be identified as;
participant error, participant’s biasness, observer error and observer’s biasness (Saunders
et al 2007b, p.149). Hence, the pilot study was able to identify the reliability as well as
the validity of the survey. The pilot study was conducted between the 15th of June and
20th of June, 2009. In terms of the data collection technique a mono method was used, i.e.
the use of questionnaire which was distributed in all three languages English, Sinhala and
Tamil so that the result would not be bias for one level of people.
Based on the responses of the pilot study a number of features in the questionnaire were
changed. Changes undertaken based on response;
It was identified that the general usage of mobile gaming among the selected sample
population including group of between 18-21 was very low. Only
5.5 Time Horizon
The time horizon for a research identify if the research would follow a snapshot or a diary
perspective. Snapshot is identified as cross sectional studies and a diary perspective is
identified as a longitudinal study (Saunders et al 2007b, p.148). Cross sectional study
refer to as “the study of a particular phenomenon at a particular time” and longitudinal
study that looks at change and development over a period of time (Saunders et al 2007b,
p.148). Out of the two it was decided to use cross sectional study that looks at a snap
shot of a particular effect due to the limited time span allocated for the research as well as
63
due to the nature of the study. The research was conducted between 15th July, 2009 and
31st July, 2009.
5.6 Determining the Sample and Sample Size
The research study looks at the total population of Sri Lanka which according to the
Census and Statistics Department stands at 20,010,000. Out of which the effective
population (between the ages of 14-64) has been identified as 13,863,950. Saunders et al
(2007b, p.212) further states that a minimum sample of 384 is needed with a margin of
5% error or in other words with a confidence level of 95% for a population above
10,000,000. Hence, looking at a estimated response rate between 40 – 45%,
approximately 1100 questionnaires were distributed among the different provinces of Sri
Lanka. The below distribution was done according probabilistic sampling method and
cluster sampling. The distribution was done according to the following table.
Province No Distributed Number of trained provincial coordinators
Actual responses received by the deadline and used for the research
Actual responses received after the deadline and not used for the research
Western 350 7 230 50 Southern 100 3 25 15 Central 250 7 160 25 North Western 100 3 40 15 North Central 100 3 25 Sabaragamuwa 100 3 20 Uva 100 3 20 1100 29 455 170
Table 15: Questionnaire distribution
In the provinces a total of 29 coordinators were appointed and trained to assist the
questionnaire respondents to coordinates and present clarifications. The questionnaire
was distributed among the volunteers who attended a training workshop conducted by the
researcher. Each volunteer was given five questionnaires to fill with their family and
specific instructions not to distribute the responses among office colleagues. Further it
64
was instructed to the volunteers that any person between the age range of 18-65 could
participate in filling the responses, irrespective of their ownership of mobile phones. The
questionnaire was designed to enable even participant who did not have mobiles to
respond. The objective of this was to understand the demographics of persons who do not
use mobile phones.
Questionnaire distribution and response information
Distributed questionnaires
(1100)
Collected Responses
(600)
Note received (500)
Accepted for data entry
(430)
Rejected prior to data entry (delayed)
(170)
Accepted for data analysis
(409)
Rejected due errors and incomplete entry
(21)
65
5.7 Questionnaire design – Likert scales used
The utilitarian motives of the model, perceived usefulness, perceived ease of use,
comparative advantage and social influences variables were measured using five point
Likert scale of “Strongly agree”, “Agree”, “No comment”, “Disagree” and “Strongly
disagree”. The questions were designed based on similar research done by Bina et al.
(2007); López-Nicolás & Molina-Castilio (2008); Kulviwat et al. (2007). A separate set
of questions were developed for SMS and Mobile Ringtone. Each set of these questions
were intended to address the logical motives of the perceived usefulness, perceived ease
of use and comparative advantage variables. The attitude towards adoption was
developed using a five point likert scale including pleasant/unpleasant, bad/good
etc(Kulviwat et al. 2007) bipolar emotions with no comment scale. While the Intension to
use was measured using three point likert scale.
Unlike the utilitarian motives of perceived usefulness, perceived ease of use and
comparative advantage variables, the construct of questionnaire to monitor Pleasure,
Arousal and Dominance was difficult. This was primarily due to the emotions involved
and their bipolar nature. In reviewing variables identified by Kulviwat et al. (2007) to
measure PAD, it was noted that only two states were identified. For example,
Pleased/Annoyed, Satisfied/Unsatisfied etc. This was deemed unsatisfactory and five
emotional points were developed. For example, Very Pleased, Pleased, No comment,
Annoyed and Very Annoyed. However significant effort and language translation
expertise was involved when developing the questionnaire in Singhala. This was due to
some words such as “In Control” not having the same interpreted meaning in the context
of SMS usage. Due to pilot testing of the questionnaire many users complained on the
difficulties in understanding the meaning of the emotions in Singhala. Therefore, a set of
examples were written above each four group of questions that would help the user in
better understanding the question. The example would indicate two statements, for
example, “1. When Sarath uses SMS in general he feels the emotion of Happiness. 2
When Sunil uses SMS in general he does not feel any emotion”. It was noted by many
66
questionnaire respondents that the states of emotion differ based on the circumstance. It
was based on this input the word “in general” was included.
5.8 Treatment of data
On receiving responses, they were analyzed for completeness and errors. Of the 430
responses 21 were rejected due to incomplete or errors in filling the forms. The
questionnaires were distributed among a group of twelve data entry volunteers. A
Microsoft Access Database was used to enter the data. The group of twelve volunteers
worked in two member teams to ensure quick data entry and quality assurance. One
person reads the responses, while the other enter data. The reader observes the data entry
screen for any typographic errors during the process. Further the data entry operator and
reader changes on a round robin basis to overcome stress in doing the repetitive task. At
the end of the process a printout of the responses are taken and checked randomly to
identify any data entry anomalies. The data entry was completed within two weeks. Data
from the six databases were exported to a main database and readied for analysis after
random checking for quality assurance.
67
6. Deliverable
6.1 Descriptive Analysis The following section relates to the descriptive statistics generated based on the collected
responses. It was decided to generate descriptive statistics on the gender, age, province of
residence, general education level, current employment status, monthly income level,
type of display used and mobile data services awareness. These information provide an
understanding of the participants and their demographic information which are crucial for
the overall understanding of the proposed research model.
6.1.2 Respondents by Gender Gender
Frequency Percent Valid Percent
Cumulative
Percent
Female 225 55.0 55.0 55.0
Male 184 45.0 45.0 100.0
Total 409 100.0 100.0 Table 16: Respondents by Gender
68
Figure 11: Respondents by Gender
The survey sample size was noted at 409 where 225 of the respondents were female and
the male respondents constituted 184. The distribution of gender is noted at 55% for
females and 45% for males. The Sri Lankan population distribution was noted at 50.4 %
for females and 49.6 % for Males. Therefore the sample may consist of higher than
normal rates of responses from females than males.
6.1.3 Respondents by Age Age
Frequency Percent Valid Percent
Cumulative
Percent
No comment 12 2.9 2.9 2.9
18-25 180 44.0 44.0 46.9
26-35 163 39.9 39.9 86.8
36-45 45 11.0 11.0 97.8
46-55 7 1.7 1.7 99.5
56-65 2 .5 .5 100.0
Total 409 100.0 100.0 Table 17 : Respondents by Age
69
Figure 12: Respondents by Age
In 2004 Sri Lankas’ population distribution by age was as follows; between 0-14 years
25%, between 15-64 years 68.3% and above 65 years 6.5%. The age distribution of the
respondents were primarily between the ages of 18-25 representing 44% of the
respondents, ages of 26-35 representing 39% of the respondents. There were 9
respondents between the ages of 46-65 and 12 respondent who did not comment on their
age. This age distribution was considered a valid sample group considering the diffusion
of new technology and their acceptance by younger age groups.
6.1.4 Respondents by Province of residence
Province of Residence
Frequency Percent Valid Percent
Cumulative
Percent
Western Province 207 50.6 50.6 53.5
North Western Province 30 7.3 7.3 60.9
Central Province 142 34.7 34.7 95.6
Southern Province 18 4.4 4.4 100.0
0ther 12 2.9 2.9 2.9
70
Province of Residence
Frequency Percent Valid Percent
Cumulative
Percent
Western Province 207 50.6 50.6 53.5
North Western Province 30 7.3 7.3 60.9
Central Province 142 34.7 34.7 95.6
Southern Province 18 4.4 4.4 100.0
Total 409 100.0 100.0 Table 18: Respondents by Province of residence
Figure 13: Respondents by Province of residence
The distribution of the questionnaires to provinces other than Western province was
aimed at getting a better understanding of the general population of Sri Lanka instead of
the opinions of Western province residents. However, 50% of the respondents were from
Western province, while 34% from Central, 7% from North Western, 4% from Southern
and 2% of respondents were from other provinces.
71
6.1.5 Respondents by Education level
Education Level
Frequency Percent Valid Percent
Cumulative
Percent
GCE Ordinary Level 66 16.1 16.1 16.4
GCE Advance Level 173 42.3 42.3 58.7
Diploma Level 37 9.0 9.0 67.7
Degree and Above
qualification 132 32.3 32.3 100.0
Not responded 1 .2 .2 .2
Total 409 100.0 100.0 Table 19: Respondents by Education level
Figure 14: Respondents by Education level
Majority of the respondents had only completed advance level qualifications, while 32%
had completed University degree and above qualifications.
72
6.1.6 Respondents by Employment status
Current Employment Status
Frequency Percent Valid Percent
Cumulative
Percent
Not responded 1 .2 .2 .2
Employed 277 67.7 67.7 68.0
Unemployed 131 32.0 32.0 100.0
Total 409 100.0 100.0 Table 20: Respondents by Employment status
Figure 15: Respondents by Employment status
67% of the respondents were employed while the balance 32% was unemployed. This
statistic is in line with the gender distribution information where 44% of the respondents
were between the ages of 18-25. Therefore their employability level would be lower than
more matured aged groups.
73
6.1.7 Respondents by monthly income level
Monthly income Level
Frequency Percent Valid Percent
Cumulative
Percent
Less than Rs.5000 62 15.2 15.2 19.8
More than Rs5000 and Less
than Rs.10,000 73 17.8 17.8 37.7
More than Rs. 10,000 and
Less than Rs.30,000 211 51.6 51.6 89.2
More than Rs.30,000 and
Less than Rs. 50,000 35 8.6 8.6 97.8
More than Rs.50000 9 2.2 2.2 100.0
Not responded 19 4.6 4.6 4.6
Total 409 100.0 100.0 Table 21: Respondents by monthly income level
Figure 16: Respondents by monthly income level
Majority of the respondents were with the income range of Rs.10,000 to 30,000 monthly.
This income category represented 51% of the respondent. 17.8% of the respondents were
between the income range of Rs.5,000-Rs.10,000. A significantly small portion of the
respondents had an income level above Rs.50,000.
74
Colour Display vs Black/White display
Frequency Percent Valid Percent
Cumulative
Percent
Valid
Black and White Display 283 69.2 69.2 73.6
Color Display 108 26.4 26.4 100.0
Not responded 18 4.4 4.4 4.4
Total 409 100.0 100.0 Table 22: Colour Display vs Black/White display
Figure 17: Colour Display vs Black/White display
This information was obtained to assess if the preference of the respondents were for
basic phones or more sophisticated ones. The more sophisticated the nature of the phone,
the customer has a better potential to use mobile data services.
75
6.1.8 Mobile Data Services Awareness
No Mobile Data Service
Frequency Percentage
Aware Not aware
No comment Aware
Not aware
No comment
1 SMS 395 14 96.6 3.4 2 MMS 315 54 40 77 13.2 9.8 3 Local language messaging 353 18 38 86.3 4.4 9.3 4 Mobile E-mail 345 24 40 84.4 5.9 9.8 5 Ringtones 365 4 40 89.2 1 9.8
6 Download Icons, Logos and wallpapers 336 29 44 82.2 7.1 10.8
7 Mobile Games 361 10 38 88.3 2.4 9.3 8 Listen to music/radio 343 12 54 83.9 2.9 13.2 9 Ask/ Send credit 312 47 50 76.3 11.5 12.2
10 Book movie tickets through mobile 226 135 48 55.3 33 11.7 11 Mobile Banking 316 48 45 77.3 11.7 11 12 Mobile internet 335 29 45 81.9 7.1 11
Table 23: Mobile Data Services Awareness This information indicates that majority of the respondents are aware of the availability
of mobile data services. This high awareness level may be attributed to the high literacy
rate in Sri Lanka estimated to be 96%. Out of the selected mobile data services surveyed
knowledge of SMS was the highest at 96.6%. While awareness of Book movie tickets
through mobile (55%), mobile baking (77%) and MMS (77) were low in comparison.
These mobile data services are considered technical and require more complex
knowledge. Jenson (2006) commenting on the complexity and stages involved in using
MMS identifies that unlike SMS, using this technology is difficult and there is a high
degree of complexity.
76
6.2 Statistical analysis of data
This section presents a summary of the model testing results for responses on utilitarian
product of SMS and Hedonic product of Ringtone. Data analysis followed the two-step
structural equation modeling approach recommended by Anderson & Gerbing (1988).
Further the recommendations of Sekaran (2006) in analyzing data were used as the basis
of analysis. The software packaged used for analysis was SPSS version 16.
The first step in the model testing and building approach involved the data being tested
for consistency and stability between the variable and responses using Cronbach’s alpha
(Sekaran, 2006). This best fit testing enabled the refining of the model and scales that did
not have high reliability with the individual variables were eliminated. The Cronbach’s
alpha measure of above 0.7 were considered acceptable according to Nunnally &
Bernstein (1994) (cited by Kulviwat et al. (2007)) was maintained in two cycles of this
best-fit model building was undertaken to ensure the quality of the test prior to hypothesis
testing. The descriptive statistics of frequency distribution table and measures of central
tendency and dispersion of maximum, minimum, means and standard deviations,
frequency distribution tables and Cronbach’s alpha values were calculated and tabulated
as indicated in Annex 2. The next stage of the analysis involved testing of the hypothesis.
This testing was done using the Pearson Correlation calculations. These were suited
based in the scaled nature of the data involved (314p, Sekaran,2006).Hypothesis were
accepted or rejected based on 95% confidence levels (p value less than 0.05).
Prior to building the prediction models for the Utilitarian and Hedonic products, it was
decided to study the individual relationship and predictability through simple linear
regression models. The models were tested based on confidence level of 95% and
ANOVA tests (F value) were conducted. In addition to these tests the residuals were
analyzed through normal probability plots (based on residuals) and scatter diagrams to
analyze if the normality assumptions were violated by the models.
77
Summary table on internal consistency and validity testing
6.2.1 Utilitarian model testing using SMS
No Variable Cronbach’s alpha values
1. Perceived Usefulness 0.919 2. Perceived Ease of use 0.960 3. Comparative advantage 0.848 4. Social Influences 0.836 5. Pleasure 0.668 6. Arousal 0.71 7. Dominance 0.859 8. Attitude Towards Adoption 0.534 9. Adoption Intension 0.273
Table 24: Test values for internal consistency – SMS Detailed statistics are available in Annex 2.
6.2.2 Hedonic model testing using Mobile Ring tone
No Variable Cronbach’s alpha values
1. Perceived Usefulness 0.989 2. Perceived Ease of use 0.988 3. Comparative advantage 0.934 4. Social Influences 0.984 5. Pleasure 0.954 6. Arousal 0.934 7. Dominance 0.923 8. Attitude Towards Adoption 0.808 9. Adoption Intension 0.342
Table 25: Test values for internal consistency - M-Ringtones Detailed statistics are available in Annex 2.
78
6.3 Hypothesis Testing The hypothesis testing was done using Person correlation and hypothesis were accepted
or rejected based on 95% confidence level. The correlation matrix provides an overview
of the nature of correlations between the tested variables. It was decided to add the
statistical correlation matrix generated through SPSS in Annex 03 for further information.
The below correlation matrix does not contain information on significance levels. This
information is provided in the table on Utilitarian model testing using SMS and hedonic
model testing using Mobile Ringtone.
Correlation Matrix for Utilitarian motives in attitude towards adoption (SMS)
Construct PU EOU RA SI PL AR DO ATA AI
Perceived
Usefulness (PU) 1
Perceived Ease
of Use (EOU) 0.81 1
Relative
Advantage (RA) 0.82 0.83 1
Social
Influences(SI) 0.40 0.45 0.33 1
Pleasure(PL) 0.41 0.37 0.33 0.27 1
Arousal(AR) 0.35 0.30 0.28 0.21 0.68 1
Dominance(DO) 0.25 0.13 0.17 0.15 0.30 0.34 1
Attitude toward
adoption(ATA) 0.79 0.74 0.71 0.41 0.52 0.53 0.30 1
Adoption
Intension(AI) 0.45 0.49* 0.41 0.17 0.14 0.21 0.57 0.38 1
Table 26: Correlation Matrix for SMS
79
Utilitarian model testing using SMS
No Hypothesis Person correlation (two tailed test) result
Tested Significance level
Status of acceptance or rejection of hypothesis at 95% confidence level
1. H01:
There is weak influence of perceived usefulness in
the context of attitude towards adoption of utilitarian
Mobile Data Services in Sri Lanka
Ha1:
There is strong influence of perceived usefulness in
the context of attitude towards adoption of
utilitarian Mobile Data Services in Sri Lanka
0.785
p value less than 0.05
Reject null hypothesis ACCEPT ALTERNATIVE HYPOTHESIS
2. H02:
There is weak influence of perceived ease of use in
the context of attitude towards adoption of utilitarian
Mobile Data Services in Sri Lanka
Ha2:
There is strong influence of perceived ease of use in
the context of attitude towards adoption of utilitarian
Mobile Data Services in Sri Lanka
0.739
p value less than 0.05
Reject null hypothesis ACCEPT ALTERNATIVE HYPOTHESIS
3. H03:
There weak influence of relative advantage in the
context of attitude towards adoption of utilitarian
Mobile Data Services in Sri Lanka
Ha3: There is strong influence of relative advantage in the
context of attitude towards adoption of utilitarian
Mobile Data Services in Sri Lanka
0.711
p value less than 0.05
Reject null hypothesis ACCEPT ALTERNATIVE HYPOTHESIS
80
No Hypothesis Person
correlation (two tailed test) result
Tested Significance level
Status of acceptance or rejection of hypothesis at 95% confidence level
4. H04:
There is weak influence of social influences in the
context of attitude towards adoption of utilitarian
Mobile Data Services in Sri Lanka
Ha4:
There is strong influence of social influences in the
context of attitude towards adoption of utilitarian
Mobile Data Services in Sri Lanka
0.407
p value less than 0.05
ACCEPT NULL HYPOTHESIS Reject alternative hypothesis
5. H05:
There weak influence of pleasure in the context of
attitude towards adoption of utilitarian Mobile Data
Services in Sri Lanka
Ha5:
There is strong influence of pleasure in the context
of attitude towards adoption of utilitarian Mobile
Data Services in Sri Lanka
0.524
p value less than 0.05
Reject null hypothesis ACCEPT ALTERNATIVE HYPOTHESIS
6. H06:
There weak influence of arousal in the context of
attitude towards adoption of utilitarian Mobile Data
Services in Sri Lanka
Ha6:
There is strong influence of arousal in the context
of attitude towards adoption of utilitarian Mobile
Data Services in Sri Lanka
0.527
p value less than 0.05
Reject null hypothesis ACCEPT ALTERNATIVE HYPOTHESIS
81
No Hypothesis Person
correlation (two tailed test) result
Tested Significance level
Status of acceptance or rejection of hypothesis at 95% confidence level
7. H07:
There weak influence of dominance in the context
of attitude towards adoption of adopt utilitarian
Mobile Data Services in Sri Lanka
Ha7:
There is strong influence of dominance in the
context of attitude towards adoption of utilitarian
Mobile Data Services in Sri Lanka
0.303
p value less than 0.05
ACCEPT NULL HYPOTHESIS Reject alternative hypothesis
8. H08:
There is weak influence of attitude towards adoption
and adoption intension in the context utilitarian
Mobile Data Services in Sri Lanka
Ha8:
There is strong influence of attitude towards
adoption and adoption intension in the context
utilitarian Mobile Data Services in Sri Lanka
0.383
p value less than 0.05
ACCEPT NULL HYPOTHESIS Reject alternative hypothesis
Table 27: Utilitarian model testing using SMS
82
List of accepted hypothesis (alternative) – Utilitarian product
No Hypothesis
Hypothesis 1
There is strong influence of perceived usefulness in the context of attitude
towards adoption of utilitarian Mobile Data Services in Sri Lanka
Hypothesis 2
There is strong influence of perceived ease of use in the context of attitude
towards adoption of utilitarian Mobile Data Services in Sri Lanka
Hypothesis 3
There is strong influence of relative advantage in the context of attitude towards
adoption of utilitarian Mobile Data Services in Sri Lanka
Hypothesis 5
There is strong influence of pleasure in the context of attitude towards adoption
of utilitarian Mobile Data Services in Sri Lanka
Hypothesis 6
There is strong influence of arousal in the context of attitude towards adoption
of utilitarian Mobile Data Services in Sri Lanka Table 28: List of accepted hypothesis (alternative) – Utilitarian product
List of Accepted Null Hypothesis
No Hypothesis
Hypothesis 4
There is weak influence of social influences in the context of attitude towards
adoption of utilitarian Mobile Data Services in Sri Lanka
Hypothesis 7
There is weak influence of dominance in the context of attitude towards
adoption of utilitarian Mobile Data Services in Sri Lanka
Hypothesis 8
There is weak influence of attitude towards adoption and adoption intension in
the context utilitarian Mobile Data Services in Sri Lanka Table 29: List of Accepted Null Hypothesis
83
Hedonic model testing using Mobile Ringtone Two models were tested in the research. This section relates to the test data on the
Hedonic value proposition model testing. The statistical analysis table generated by SPSS
was included in Annex 3 for further reference.
Correlation Matrix for hedonic motives in attitude towards adoption (Mobile Ring tones)
Construct PU EOU RA SI PL AR DO ATA AI
Perceived
Usefulness (PU) 1.00
Perceived Ease
of Use (EOU) 0.96 1.00
Relative
Advantage (RA) 0.94 0.97 1.00
Social
Influences(SI) 0.96 0.98 0.97 1.00
Pleasure(PL) 0.90 0.92 0.94 0.92 1.00 Arousal(AR) 0.78 0.74 0.75 0.79 0.72 1.00 Dominance(DO) 0.51 0.51 0.51 0.51 0.48 0.49 1.00 Attitude toward
adoption(ATA) 0.92 0.93 0.97 0.92 0.94 0.74 0.54 1.00
Adoption
Intension(AI) 0.69 0.71 0.72 0.72 0.68 0.59 0.42 0.67 1.00
Table 30: Correlation Matrix for hedonic motives
84
Hypothesis testing for Hedonic model
No Hypothesis Person
correlation (two
tailed test)
result
Tested
Significance
level
Status of
acceptance or
rejection of
hypothesis at 95%
confidence level
1. H09:
There is weak influence of perceived usefulness in
the context of attitude towards adoption of hedonic
Mobile Data Services in Sri Lanka
Ha9:
There is strong influence of perceived usefulness in
the context of attitude towards adoption of hedonic
Mobile Data Services in Sri Lanka
0.922
p value less
than 0.05
Reject null
hypothesis
ACCEPT
ALTERNATIVE
HYPOTHESIS
2. H010:
There weak influence of perceived ease of use in
the context of attitude towards adoption of hedonic
Mobile Data Services in Sri Lanka
Ha10:
There is strong influence of perceived ease of use
in the context of attitude towards adoption of
hedonic Mobile Data Services in Sri Lanka
0.930
p value less
than 0.05
Reject null
hypothesis
ACCEPT
ALTERNATIVE
HYPOTHESIS
3. H011:
There weak influence of relative advantage in the
context of attitude towards adoption of hedonic
Mobile Data Services in Sri Lanka
Ha11:
There is strong influence of relative advantage in
the context of attitude towards adoption of hedonic
Mobile Data Services in Sri Lanka
0.965
p value less
than 0.05
Reject null
hypothesis
ACCEPT
ALTERNATIVE
HYPOTHESIS
85
No Hypothesis Person
correlation (two
tailed test)
result
Tested
Significance
level
Status of
acceptance or
rejection of
hypothesis at 95%
confidence level
4. H012:
There weak influence of social influence in the
context of attitude towards adoption of hedonic
Mobile Data Services in Sri Lanka
Ha12:
There is a strong influence of social influence in the
context of attitude towards adoption of hedonic
Mobile Data Services in Sri Lanka
0.917
p value less
than 0.05
Reject null
hypothesis
ACCEPT
ALTERNATIVE
HYPOTHESIS
5. H013:
There is weak influence of pleasure in the context
of attitude towards adoption of adopt hedonic
Mobile Data Services in Sri Lanka
Ha13:
There is strong influence of pleasure in the context
of attitude towards adoption of hedonic Mobile
Data Services in Sri Lanka
0.941
p value less
than 0.05
Reject null
hypothesis
ACCEPT
ALTERNATIVE
HYPOTHESIS
6. H014:
There weak influence of arousal in the context of
attitude towards adoption of hedonic Mobile Data
Services in Sri Lanka
Ha14:
There is strong influence of arousal in the context
of attitude towards adoption of hedonic Mobile
Data Services in Sri Lanka
0.742
p value less
than 0.05
Reject null
hypothesis
ACCEPT
ALTERNATIVE
HYPOTHESIS
86
No Hypothesis Person
correlation (two
tailed test)
result
Tested
Significance
level
Status of
acceptance or
rejection of
hypothesis at 95%
confidence level
7. H015:
There weak influence of dominance in the context
of attitude towards adoption of adopt hedonic
Mobile Data Services in Sri Lanka
Ha15:
There is strong influence of dominance in the
context of attitude towards adoption of hedonic
Mobile Data Services in Sri Lanka
0.538
p value less
than 0.05
Reject null
hypothesis
ACCEPT
ALTERNATIVE
HYPOTHESIS
8. H016:
There is weak influence of attitude towards
adoption and adoption intension in the context
hedonic Mobile Data Services in Sri Lanka
Ha16:
There is strong influence of attitude towards
adoption and adoption intension in the context
hedonic Mobile Data Services in Sri Lanka
0.672
p value less
than 0.05
Reject null
hypothesis
ACCEPT
ALTERNATIVE
HYPOTHESIS
Table 31: Hypothesis testing for Hedonic model
87
List of accepted hypothesis – Hedonic Product
No Hypothesis
Hypothesis 9 There is strong influence of perceived usefulness in the context of attitude
towards adoption of hedonic Mobile Data Services in Sri Lanka
Hypothesis 10 There is strong influence of perceived ease of use in the context of attitude
towards adoption of hedonic Mobile Data Services in Sri Lanka
Hypothesis 11 There is strong influence of perceived ease of use in the context of attitude
towards adoption of hedonic Mobile Data Services in Sri Lanka
Hypothesis 12 There is a strong influence of social influence in the context of attitude towards
adoption of hedonic Mobile Data Services in Sri Lanka
Hypothesis 13 There is strong influence of pleasure in the context of attitude towards adoption
of hedonic Mobile Data Services in Sri Lanka
Hypothesis 14 There is strong influence of arousal in the context of attitude towards adoption
of hedonic Mobile Data Services in Sri Lanka
Hypothesis 15 There is strong influence of dominance in the context of attitude towards
adoption of hedonic Mobile Data Services in Sri Lanka
Hypothesis 16 There is strong influence of attitude towards adoption and adoption intension
in the context hedonic Mobile Data Services in Sri Lanka Table 32: List of accepted hypothesis – Hedonic Product
88
6.4 Simple liner model building Variables in Model tested for Utilitarian value proposition (SMS)
No Variables Model
predictabilit
y percentage
(%) based
on the
Adjusted R
squared
value
ANOVA F
value test and
acceptability
at 95%
confidence
level – Accept
or Reject
status
Simple Liner
regression
equation
Comments on residual
value analysis
Reference in
Annex 6
1. Perceived
usefulness and
attitude
towards
adoption
61.5% Accept y = .092 + .332 X1 Model Acceptable.
45 degree upward
sloping plot observed and
Randomly scattered
residuals an even
horizontal band around
residual value of zero
observed
Tables 1-4
and Figures
1-2
2. Perceived ease
of use and
attitude
towards
adoption
54.5 % Accept y = .133 + .305 X1 Model Acceptable.
45 degree upward sloping
plot observed and
Randomly scattered
residuals an even
horizontal band around
residual value of zero
observed
Tables 5-8
and Figures
3-4
3. Comparative
advantage and
attitude
towards
adoption
50.4% Accept y = .266 + .343 X1 Model Acceptable.
45 degree upward sloping
plot observed and
Randomly scattered
residuals an even
horizontal band around
Tables 9-12
and Figures
5-6
89
No Variables Model
predictabilit
y percentage
(%) based
on the
Adjusted R
squared
value
ANOVA F
value test and
acceptability
at 95%
confidence
level – Accept
or Reject
status
Simple Liner
regression
equation
Comments on residual
value analysis
Reference in
Annex 6
residual value of zero
observed
4. Social
influences and
attitude
towards
adoption
16.3% Accept y = .749 + .201 X1 Reject Model.
Sparse and broken 45
degree upward sloping
plot observed. Low
random nature of the
scattered instances shown
in the plot
Tables 13-15
and Figures
7-8
5. Pleasure and
attitude
towards
adoption
27.3% Accept y = .933 + .653 X1 Reject Model.
Sparse and broken 45
degree upward sloping
plot observed. Low
random nature of the
scattered instances shown
in the plot
Tables 16-18
and Figures
9-10
6. Arousal and
attitude
towards
adoption
27.6% Accept y = 1.004 + .611 X1 Reject Model.
Sparse and broken 45
degree upward sloping
plot observed. Low
random nature of the
scattered instances shown
in the plot
Tables 19-21
and Figures
11-12
7. Dominance
and attitude
8.9% Accept y = 1.088 + .282 X1 Reject Model.
Sparse and broken 45
Tables 22-24
and Figures
90
No Variables Model
predictabilit
y percentage
(%) based
on the
Adjusted R
squared
value
ANOVA F
value test and
acceptability
at 95%
confidence
level – Accept
or Reject
status
Simple Liner
regression
equation
Comments on residual
value analysis
Reference in
Annex 6
towards
adoption
degree upward sloping
plot observed. Low
random nature of the
scattered instances shown
in the plot
13-14
8. Attitude
towards
adoption and
Adoption
Intension
14.4% Accept y = 3.923 + .569 X1 Reject Model.
Sparse and broken 45
degree upward sloping
plot observed. Low
random nature of the
scattered instances shown
in the plot
Tables 25-27
and Figures
15-16
Table 33: Simple liner model building – SMS The simple linear regression models were tested between the identified variables and
attitude towards adoption. Based on these linear regression models and residual analysis
it was identified that the variables of Social influence, pleasure arousal, dominance and
the relationship between attitude towards adoption and adoption intension had weak
model building capabilities. Even thought the models were acceptable within the 95%
confidence level of ANOVA test, their analysis of residuals indicate that these models
will not successfully provide the required prediction capability.
91
Variables in Model tested for Hedonic value proposition (Mobile Ringtone)
No Variables Model
predictabilit
y percentage
(%) based
on the
Adjusted R
squared
value
ANOVA F
value test and
acceptability
at 95%
confidence
level – Accept
or Reject
status
Simple Liner
regression
equation
Comments on residual
value analysis
Reference in
Annex 6
1. Perceived
usefulness and
attitude
towards
adoption
85%
Accept y = .039+ .516X1 Model Acceptable.
45 degree upward
sloping plot observed and
Randomly scattered
residuals an even
horizontal band around
residual value of zero
observed
Tables 28-30
and Figures
17-18
2. Perceived ease
of use and
attitude
towards
adoption
86.5%
Accept y = .029+ .494X1 Model Acceptable.
45 degree upward
sloping plot observed and
Randomly scattered
residuals an even
horizontal band around
residual value of zero
observed
Tables 31-33
and Figures
19-20
3. Comparative
advantage and
attitude
towards
adoption
93.1%
Accept y = .005+ .564X1 Model Acceptable.
45 degree upward
sloping plot observed and
Randomly scattered
residuals an even
horizontal band around
residual value of zero
observed
Tables 34-36
and Figures
21-22
92
No Variables Model
predictabilit
y percentage
(%) based
on the
Adjusted R
squared
value
ANOVA F
value test and
acceptability
at 95%
confidence
level – Accept
or Reject
status
Simple Liner
regression
equation
Comments on residual
value analysis
Reference in
Annex 6
4. Social
influences and
attitude
towards
adoption
84%
Accept y = .041+ .489X1 Model Acceptable.
45 degree upward
sloping plot observed and
Randomly scattered
residuals an even
horizontal band around
residual value of zero
observed
Tables 37-39
and Figures
23-24
5. Pleasure and
attitude
towards
adoption
88.5% Accept y = .044+ .684X1
Model Acceptable.
45 degree upward
sloping plot observed and
Randomly scattered
residuals an even
horizontal band around
residual value of zero
observed
Tables 40-42
and Figures
25-26
6. Arousal and
attitude
towards
adoption
54.9% Accept y = .331+ 1.226X1
Model Acceptable.
45 degree upward
sloping plot observed and
Randomly scattered
residuals an even
horizontal band around
residual value of zero
observed
Tables 43-45
and Figures
27-28
93
No Variables Model
predictabilit
y percentage
(%) based
on the
Adjusted R
squared
value
ANOVA F
value test and
acceptability
at 95%
confidence
level – Accept
or Reject
status
Simple Liner
regression
equation
Comments on residual
value analysis
Reference in
Annex 6
7. Dominance
and attitude
towards
adoption
28.8%
Accept y = .578+ 1.162X1
Model Acceptable.
45 degree upward
sloping plot observed and
Randomly scattered
residuals an even
horizontal band around
residual value of zero
observed
Tables 46-48
and Figures
29-30
8. Attitude
towards
adoption and
Adoption
Intension
45% Accept y = 2.953 + .726 X1 Model Acceptable.
45 degree upward
sloping plot observed and
Randomly scattered
residuals an even
horizontal band around
residual value of zero
observed
Tables 49-50
and Figures
31-32
Table 34: Simple liner model building - Mobile Ringtones The simple linear regression models were tested between the identified variables and
attitude towards adoption. Based on these linear regression models and residual analysis
it was identified that the model build based on all the selected variables will provide a
successful prediction model.
94
6.5 Model building
6.5.1 Utilitarian Product of SMS
6.5.2 Attitude towards adoption
Based on the statistical analysis below only five variables were identified to have strong
correlation of above 0.5 with the attitude towards adoption. These variable were
Perceived usefulness, perceived ease of use, comparative advantage, pleasure and
arousal. It was decided to analyzed the validity of the prediction model between the
incorporation of all selected variables and only utilitarian motives. The object of the
study was to evaluate if the prediction capability of the model was different between the
two selected options.
Variable ranking based on correlation to Attitude towards adoption
No Measured Variable Correlation (at 95% significance – two tailed test)
R squared value Ranking
1. Perceived Usefulness and attitude towards adoption in the context of an utilitarian product
0.79 61.50% 1
2. Perceived ease of use and attitude towards adoption in the context of an utilitarian product
0.74 54.50% 2
3. relative advantage and attitude towards adoption in the context of an utilitarian product
0.71 50.40% 3
4. Arousal and attitude towards adoption in the context of an utilitarian product
0.53 27.60% 4
5. Pleasure and attitude towards adoption in the context of an utilitarian product
0.52 27.30% 5
Table 35: Variable ranking based on correlation to Attitude towards adoption
95
Option 1: Testing of the prediction capability of attitude towards adoption with the
incorporation of Perceived Usefulness, Perceived Ease of Use, Comparative
advantage, Pleasure and Arousal.
Option 2: Testing of the prediction capability of attitude towards adoption with only
the incorporation of Perceived Usefulness, Perceived Ease of Use and Comparative
advantage.
Result
Test Model 1 : Option 1
Pearson Correlation
Significance (Two tailed) (N=409)
.849 0.000
Liner Model y = .031+ .169 PU+ .098 EOU+ .044 CA+ .109 PL+ .266 AR
where,
PU – Perceived Usefulness
EOU – Perceived Ease of Use
CA – Comparative Advantage
PL – Pleasure
AR – Arousal
The constant value, variable of comparative advantage and
Pleasure recorded p values greater than 0.05.
Model predictability (Adjusted R squared) 71.7%
ANOVA F value was noted at 207.720 indicating the significance is under
95% confidence. Model Accepted.
Acceptability of model based on Residual
analysis
ACCEPTED
Normal probability plot (Residual) 45 degree upward sloping plot observed
Scatter plot of the standardized residuals vs
standard fitted (Residual)
Randomly scattered in an even horizontal band around residual
value of zero
Reference for detailed test data ANNEX 06, 14-16 pages
96
Test Model 1 : Option 2
Model predictability (Adjusted R
squared)
64.4%
Liner Model y = .002+ .217 PU+ .107 EOU.035 CA
where,
PU – Perceived Usefulness
EOU – Perceived Ease of Use
CA – Comparative Advantage
The constant value and variable of comparative
advantage recorded p values greater than 0.05.
ANOVA F value was noted at 247.463 indicating the significance
is under 95% confidence. Model Accepted.
Pearson Correlation
Significance (Two tailed) (N=409)
.804
.000
Acceptability of model based on
Residual analysis
ACCEPTED
Normal probability plot (Residual) 45 degree upward sloping plot observed
Scatter plot of the standardized
residuals vs standard fitted
(Residual)
Randomly scattered in an even horizontal band
around residual value of zero
Reference for detailed test data ANNEX 06, 17-19 pages
97
6.5.3 Intension to adopt
The objective of this analysis using the multiple regression model building statistical
methodology is to identify a prediction model for intension to adopt SMS. Two models
will be developed for comparison purposes. The first model selected for evaluation will
consist of all the variables involved in the original model. The first model consists of the
six variables of perceived usefulness, perceived ease of use, comparative advantage,
pleasure, arousal and attitude towards adoption. The first five variables listed above
helped in developing the strong prediction model for attitude towards adoption, which
had a prediction capability of 71.7%.
Option 1: Testing of the prediction capability of the adoption intension with the
incorporation of perceived usefulness, perceived ease of use, comparative advantage,
social influence, pleasure, arousal, dominance and attitude towards adoption
Option 2: Testing of the prediction capability of the adoption intension with the
incorporation of perceived usefulness, perceived ease of use, comparative advantage,
pleasure, arousal and attitude towards adoption
98
Test Model 1 : Option 1
Model predictability (Adjusted R
squared)
40.0 %
Multiple regression Model Y = 3.139+ 0.165 EOU + 0.175 DO + 0.316 SO – 0.119 ATA +
0.85 PU – 0.032 AR + 0.158 PL
Where,
EOU – is Perceived Ease of Use
DO – Dominance
SO – Social Influences
ATA – Attitude towards adoption
PU- Perceived usefulness
AR- Arousal
PL- Pleasure
The variable of Perceived Usefulness, Perceived Ease of Use,
Comparative advantage, Pleasure and Arousal all recorded p
values greater than 0.05.
ANOVA F value was noted at 89.323 indicating the significance is under
95% confidence. Accepted
Pearson Correlation
Significance (Two tailed) (N=409)
.631
.000
Acceptability of model based on
Residual analysis
Accepted
DECSION ACCEPT MODEL
Reference for detailed test data ANNEX 06, 2-7 pages
99
Test Model 1 : Option 2
Model predictability (Adjusted R
squared)
24.2%
Multiple regression Model When building the model it was noted that the variables of
Attitude towards adoption, Comparative advantage, pleasure
and arousal all failed in their t test values recording
significance above 95% confidence. Therefore it was decided
not to proceed with building the model.
ANOVA The F value is 22.680. This is significant under 95 % confidence (p
value is less than 0.05) Pearson Correlation
Significance (Two tailed) (N=409)
.503
.000
Acceptability of model based on
Residual analysis
Rejected
DECSION REJECT MODEL
Reference for detailed test data ANNEX 06, 7-10 pages
Based on the above analysis it was decided to accept model 1 as the prediction model
for Adoption intension for SMS
100
6.5.4 Hedonic Product of Mobile Ringtone The statistical analysis identified that both utilitarian and hedonic motives were
influencing the adoption intension of the survey respondents. The below analysis is
aimed at identifying key variables and their influence on the prediction capability of the
Hedonic product adoption model. In order to test the models two alternative options
were developed. The first option consists variables from the original model. The second
option consists of variables representing hedonic motives. Through this analysis it would
be possible to identify if hedonic motives have a better prediction capability than
utilitarian and hedonic motives combined.
Option 1: Testing of the prediction capability of the adoption intension with the
incorporation of perceived usefulness, perceived ease of use, comparative advantage,
social influence, pleasure, arousal and attitude towards adoption
Option 2: Testing of the prediction capability of the adoption intension with the
incorporation of only Hedonic motives.
101
Test Model 2 : Option 1
Model predictability (Adjusted R
squared)
54.2 %
Liner Model y = 2.872 + .014 HePU + .200 HeEOU + .720 HeCA -.273 HeSO
+ .221 HePL + .286 HeAR + .241 HeDO -.841 HeATA
where,
HePU – Perceived Usefulness HeEOU – Perceived Ease of Use
HeCA – Comparative Advantage
HeSO – Social influences
HePL – Pleasure
HeAR - Arousal
HeDO - Dominance
HeATA – Attitude towards adoption
When building the model it was noted that the variables of
Perceived Usefulness, perceived ease of use and social
influences all failed in their t test values recording significance
above 95% confidence.
ANOVA The F value is 61.471. This is significant under 95 % confidence (p
value is less than 0.05) Pearson Correlation
Significance (Two tailed) (N=409)
0.743 0.000
Acceptability of model based on
Residual analysis
ACCEPTED
Normal probability plot (Residual) 45 degree upward sloping plot observed
Scatter plot of the standardized
residuals vs standard fitted (Residual)
Randomly scattered in an even horizontal band around residual
value of zero
Reference for detailed test data ANNEX 06, 8-11 pages
102
Test Model 2 : Option 2
Model predictability (Adjusted R
squared)
49.1 %
Liner Model y = 2.913 +.345 HePL + .311HeAR + .181HeDO+.095 HeATA
where,
HePL – Pleasure HeAR - Arousal
HeDO - Dominance
HeATA – Attitude towards adoption
When building the model it was noted that the variables of
Dominance and Attitude towards adoption failed in their t test
values recording significance above 95% confidence.
ANOVA The F value is 99.335. This is significant under 95 % confidence (p
value is less than 0.05)
Pearson Correlation
Significance (Two tailed) (N=409)
.743 0.000
Acceptability of model based on
Residual analysis
ACCEPTED
Normal probability plot (Residual) 45 degree upward sloping plot observed
Scatter plot of the standardized
residuals vs standard fitted
(Residual)
Randomly scattered in an even horizontal band around
residual value of zero
Reference for detailed test data ANNEX 06, 12-14 pages
Based on this analysis test model 2: Option 1 was selected as its prediction capability was
above that of its alternative.
103
6.6 Data Analysis Summary
6.6.1 Utilitarian product – SMS adoption model testing
Perceived Usefulness
The questions to measure perceived usefulness were developed on a five point likert scale
with points between ‘Strongly Agree’, ‘agree’, ‘no comment’, ‘disagree’, ‘strongly
disagree’. Coding of the responses were done from left to right starting with 5 marks for
‘Strongly agree’ and 1 mark for ‘Strongly disagree’. 3 marks were awarded for the
response ‘no comment’. The mean value of the responses was measured at 3.4 with a
standard deviation of 1.28 (annex 2, table 2). This response mean indicates that the
majority of respondents agreed with the variables assigned to measure perceived
usefulness in the context of SMS usage. The internal consistency and reliability of the
variable were tested using Cronbach’s alpha, which was measured at 0.919 (table 24)
indicating a very good fit as noted by Hair et al. (1998)(as cited by Kulviwat et al. (2007).
The two tailed Pearson Correlation test between the variable and Attitude towards
adoption indicated a correlation of 0.785 with p value less than 0.05(annex 3). This
indicated that there is significant correlation between the two variables at 95% confidence
level. Therefore the Null hypothesis was rejected and alternative hypothesis of “There is
strong influence of perceived usefulness in the context of attitude towards adoption of
utilitarian Mobile Data Services in Sri Lanka” was accepted. To further analyse the
nature the relationship between these two variables regression analysis was conducted.
The adjusted R squared value was 0.615 indicating that the predictability of the linear
model at 95% confidence level was 61.5% . Based on the regression analysis an
individual linear prediction model was built with the equation of y = .092 + .332 X1. (y
being the Attitude towards adoption and X being the Perceived usefulness). Two tests
were conducted prior to the acceptance of the simple linear regression model. First,
ANOVA testing of the model was conducted at 95% confidence level. The model was
accepted based on this test. Residual analysis of the model was conducted as the second
104
test. The Normal probability plot indicated a 45 degree sloping plot while the scatter plot
indicated even horizontal band around residual value of zero. In addition to these
statistical relationships the correlation between Perceived Usefulness and Intension to
adopt was observed at 0.451. This strength of correlation was above that between the
Attitude towards adoption and Intension to adopt.
Perceived Ease of Use
The questions to measure perceived ease of use were developed on a five point likert
scale with points between ‘Strongly Agree’, ‘agree’, ‘no comment’, ‘disagree’, ‘strongly
disagree’. Coding of the responses were done from left to right starting with 5 marks for
‘Strongly agree’ and 1 mark for ‘Strongly disagree’. 3 marks were awarded for the
response ‘no comment’. The mean value of the responses was measured at 3.5 with a
standard deviation of 1.3(annex 2, table 5). This mean value indicates that the majority of
respondents agreed with the variables assigned to measure perceived ease of use in the
context of SMS service usage. The internal consistency and reliability of the variable
were tested using Cronbach’s alpha, which was measured at 0.96 (table 24) indicating a
very good fit. The two tailed Pearson Correlation test between the variable and Attitude
towards adoption indicated a correlation of 0.739 with p value less than 0.05(annex 3).
This indicated that there is significant correlation between the two variables at 95%
confidence level. Therefore the Null hypothesis was rejected and alternative hypothesis
of “There is strong influence of perceived ease of use in the context of attitude towards
adoption of utilitarian Mobile Data Services in Sri Lanka” was accepted. To further
analyse the nature of the relationship between these two variables, regression analysis
was conducted. The adjusted R squared value was 0.545 indicating that the predictability
of the linear model at 95% confidence level was 54.5%. Based on the regression analysis
an individual linear prediction model was built based on the equation of y = .133 + .305
X1. (y being the Attitude towards adoption and X being the Perceived Ease of use). Two
tests were conducted prior to the acceptance of the simple linear regression model. First,
ANOVA testing of the model was conducted at 95% confidence level. The model was
accepted based on this test. Residual analysis of the model was conducted as the second
105
test. The Normal probability plot indicated a 45 degree sloping plot while the scatter plot
indicated even horizontal band around residual value of zero. In addition to these
statistical relationships the correlation between Ease of use and Intension to adopt was
observed at 0.493 (annex 3). This is the highest noted strength between the identified
independent variables and the Intension to use variable.
Comparative Advantage
The questions to measure comparative advantage were developed on a five point likert
scale with points between ‘Strongly Agree’, ‘agree’, ‘no comment’, ‘disagree’, ‘strongly
disagree’. Coding of the responses were done from left to right starting with 5 marks for
‘Strongly agree’ and 1 mark for ‘Strongly disagree’. 3 marks were awarded for the
response ‘no comment’. The mean value of the responses was measured at 2.9 with a
standard deviation of 1.1(annex 2, table 8). This response indicates that the majority of
respondents did not agree with the proposition. However a detailed analysis of the
frequency charts and histogram indicates that over 60% of the sample respondents agreed
with the variables associated with comparative advantage in using SMS. The internal
consistency and reliability of the variables were tested using Cronbach’s alpha which was
measured at 0.85 (table 24) indicating a very good fit. The two tailed Pearson Correlation
test between the variable and Attitude towards adoption indicated a correlation of 0.711
with p value less than 0.05(annex 3). This indicated that there is significant correlation
between the two variables at 95% confidence level. Therefore the Null hypothesis was
rejected and alternative hypothesis of “There is strong influence of relative advantage in
the context of attitude towards adoption of utilitarian Mobile Data Services in Sri Lanka”
was accepted. To further analyse the nature of the nature of the relationship between
these two variables regression analysis was conducted. The adjusted R squared value was
0.504 indicating that the predictability of the linear model at 95% confidence level was
50.4%. Based on the regression analysis an individual linear prediction model was built
based on the equation of y = .266 + .343 X1. (y being the Attitude towards adoption and
X being the comparative advantage). Two tests were conducted prior to the acceptance of
the simple linear regression model. First, ANOVA testing of the model was conducted at
106
95% confidence level. The model was accepted based on this test. Residual analysis of
the model was conducted as the second test. The Normal probability plot indicated a 45
degree sloping plot while the scatter plot indicated even horizontal band around residual
value of zero. In addition to these statistical relationships the correlation between
comparative advantage and Intension to adopt was observed at 0.408(annex 3). This is
the third highest strength identified between the independent variables and the Intension
to use variable in utilitarian product adoption model. This strength of correlation was
above that between the Attitude towards adoption and Intension to adopt.
Social Influences
The questions to measure social influences were developed on a five point likert scale
with points between ‘Strongly Agree’, ‘agree’, ‘no comment’, ‘disagree’, ‘strongly
disagree’. Coding of the responses were done from left to right starting with 5 marks for
‘Strongly agree’ and 1 mark for ‘Strongly disagree’. 3 marks were awarded for the
response ‘no comment’. The mean value of the responses was measured at 2.39 with a
standard deviation of 1.1(annex 2, table 11). This response indicates that the majority of
respondents did not agree with the proposition. This notion is confirmed as 65 % of the
respondents disagreed with the influence of social factors as a reason for using SMS. The
internal consistency and reliability of the variables were tested using Cronbach’s alpha
which was measured at 0.836 (table 24) indicating a very good fit. The two tailed Pearson
Correlation test between the variable and Attitude towards adoption indicated a
correlation of 0.407 with p value less than 0.05(annex 3). This indicated that there is
significant correlation between the two variables at 95% confidence level. Therefore the
Null hypothesis was accepted. The accepted hypothesis was “There is a weak influence
of social influences in the context of attitude towards adoption of utilitarian Mobile Data
Services in Sri Lanka”. To further analyse the nature of the nature of the relationship
between these two variables regression analysis was conducted. The adjusted R squared
value was 0.163 indicating that the predictability of the linear model at 95% confidence
level was 16.3%. Based on the regression analysis an individual linear prediction model
was built based on the equation of y = .749 + .201 X1. (y being the Attitude towards
107
adoption and X being the Social influences). Two tests were conducted prior to the
acceptance of the simple linear regression model. First, ANOVA testing of the model was
conducted at 95% confidence level. The model was accepted based on this test. Residual
analysis of the model was conducted as the second test. Sparse and broken 45 degree
upward sloping plot observed. Low random instances were observed in the scatter plot.
Therefore it is recommended that this liner model be rejected. In addition to these
statistical relationships the correlation between social influences and Intension to adopt
was observed at 0.567. This is the strongest identified correlation other than motives of
Utility and Hedonics, between the independent variables and the Intension to use variable
in utilitarian product adoption model. This strength of correlation was above that between
the Attitude towards adoption and Intension to adopt.
Pleasure
The questions to measure pleasure was developed on a five point likert scale. Two states
of positive pleasure and two states of negative pleasure were identified with the option of
‘no comment’ as the scales. Marks were given as 2 and 1 for both positive and negative
emotional states, where ‘highly pleased’ or ‘highly displeased’ emotions received 2
marks and the ‘no comment’ response was allocated 0 marks. The mean value of the
responses was measured at 0.44 with a standard deviation of 0.44 (annex 2, table 14).
This response mean indicates that the majority of respondents did not feel any emotion
associated with pleasure when using SMS. However a detailed analysis of the frequency
charts and histogram indicates that over 62% of the sample respondents registered a
response associated with the variables pleasure in using SMS, while 152 (37%) of the
respondents indicate that they did not have any emotions identified with pleasure when
using SMS. The internal consistency and reliability of the variables were tested using
Cronbach’s alpha which was measured at 0.66 (table 24). This measure indicates a less
than optimum fit (Hair et al. (1998)(as cited by Kulviwat et al. (2007)). Three additional
rounds of cross matching the responses were conducted to asses if the Cronbach’s alpha
could be improved if certain responses were removed. But this effort was a failure as
none of these cross matching improve the level beyond 0.66. The two tailed Pearson
108
Correlation test between the variable and Attitude towards adoption indicated a
correlation of 0.524 with p value less than 0.05(annex 3). This indicated that there is
significant correlation between the two variables at 95% confidence level. Therefore the
Null hypothesis was rejected and alternative hypothesis of “There is strong influence of
pleasure in the context of attitude towards adoption of utilitarian Mobile Data Services in
Sri Lanka” was accepted. To further analyse the nature of the relationship between these
two variables, regression analysis was conducted. The adjusted R squared value was
0.273 indicating that the predictability of the linear model at 95% confidence level was
27.3%. Based on the regression analysis an individual linear prediction model was built
based on the equation of y = .933 + .653 X1. (y being the Attitude towards adoption and X
being the pleasure). Two tests were conducted prior to the acceptance of the simple linear
regression model. First, ANOVA testing of the model was conducted at 95% confidence
level. The model was accepted based on this test. Residual analysis of the model was
conducted as the second test. Sparse and broken 45 degree upward sloping plot observed.
Low random instances were observed in the scatter plot. Therefore it is recommended
that this liner model be rejected. In addition to these statistical relationships the
correlation between Pleasure and Intension to adopt was observed at 0.170. This was the
weakest relationship identified between the independent variables and the Intension to
use variable in utilitarian product adoption model.
Arousal
The arousal variable was developed based on a five point likert scale and marks were
allocated based on the same technique used with the Pleasure variable. The mean value of
the responses was measured at 0.36 with a standard deviation of 0.46 (annex 2, table 17).
This response indicates that the majority of respondents did not feel any emotions
associated with arousal when using SMS. In a detailed analysis of the frequency charts
and histogram indicates that only 47% of the sample respondents registered a response
associated with the variable of arousal in using SMS, while 215 (52%) of the respondents
indicate that they did not have any emotions identified with pleasure when using SMS.
The internal consistency and reliability of the variables were tested using Cronbach’s
109
alpha which was measured at 0.71(table 24). Nunnally & Bernstein (1994) (as cited by
Kulviwat et al. (2007) notes that internal consistency reliability levels between 0.7-0.8
are considered acceptable levels. The two tailed Pearson Correlation test between the
variable and Attitude towards adoption indicated a correlation of 0.527 with p value less
than 0.05(annex 3). This indicated that there is significant correlation between the two
variables at 95% confidence level. Therefore the Null hypothesis was rejected and
alternative hypothesis of “There is strong influence of arousal in the context of attitude
towards adoption of utilitarian Mobile Data Services in Sri Lanka” was accepted. To
further analyse the nature of the relationship between these two variables regression
analysis was conducted.
The adjusted R squared value was 0.276 indicating that the predictability of the linear
model at 95% confidence level was 27.6%. Based on the regression analysis an
individual linear prediction model was built based on the equation of y = 1.004 + .611 X1.
(y being the Attitude towards adoption and X being the arousal). Two tests were
conducted prior to the acceptance of the simple linear regression model. First, ANOVA
testing of the model was conducted at 95% confidence level. The model was accepted
based on this test. Residual analysis of the model was conducted as the second test.
Sparse and broken 45 degree upward sloping plot observed. Low random instances were
observed in the scatter plot. Therefore it is recommended that this liner model be rejected.
In addition to these statistical relationships the correlation between arousal and Intension
to adopt was observed at 0.143. This was the weakest relationship identified between the
hedonic independent variables and the Intension to use variable in utilitarian product
adoption model.
Dominance
The arousal variable was developed based on a five point likert scale and marks were
allocated based on the same technique used with the Pleasure variable. The mean value of
the responses was measured at 0.48 with a standard deviation of 0.58. This response
mean indicates that the majority of respondents did feel any emotions associated with
110
arousal when using SMS. In a detailed analysis of the frequency charts and histogram
indicates that 51% of the sample respondents registered a response associated with the
variable of dominance in using SMS, while 48% of the respondents indicate that they did
not have any emotions identified with arousal when using SMS.
The internal consistency and reliability of the variables were tested using Cronbach’s
alpha which was measured at 0.859(table 24). Nunnally & Bernstein (1994) (as cited by
Kulviwat et al. (2007)) notes that internal consistency reliability levels above 0.8 are
considered good. The two tailed Pearson Correlation test between the variable and
Attitude towards adoption indicated a correlation of 0.303 with p value less than
0.05(annex 3). This indicated that there is no significant correlation between the two
variables at 95% confidence level. Therefore the Null hypothesis was accepted. The
accepted null hypothesis is “There is weak influence of dominance in the context of
attitude towards adoption of utilitarian Mobile Data Services in Sri Lanka”.
To further analyse the nature of the nature of the relationship between these two variables
regression analysis was conducted. The adjusted R squared value was 0.089 indicating
that the predictability of the linear model at 95% confidence level was 8.9%. Based on
the regression analysis an individual linear prediction model was built based on the
equation of y = 1.004 + .611 X1. (y being the Attitude towards adoption and X being the
dominance). Two tests were conducted prior to the acceptance of the simple linear
regression model. First, ANOVA testing of the model was conducted at 95% confidence
level. The model was accepted based on this test. Residual analysis of the model was
conducted as the second test. Sparse and broken 45 degree upward sloping plot observed.
Low random instances were observed in the scatter plot. Therefore it is recommended
that this liner model be rejected. In addition to these statistical relationships the
correlation between arousal and Intension to adopt was observed at 0.205.
Attitude towards adoption and intension The five point likert scale used in this instance registered responses based on “bad/good”,
“negative/positive”, “favorable/ unfavorable”, “pleasant/unpleasant” scales. Marks were
111
allocated from 5 to 1, where the positive responses earned 5-4 marks while the negative
responses earned 2-1 marks. 3 marks were awarded for the response ‘no comment’. The
mean value of the responses for attitude towards adoption was measured at 1.22 with a
standard deviation of 0.54. (annex 2, table 24). This indicates that the attitude towards
adoption was low.
For intension to adopt the likert scales used were “unlikely/likely”,
“improbable/probable”, “impossible/possible”. With marks awarded using the same
technique as attitude. The mean value for intension to adopt was measured at 4.62 with a
standard deviation of 0.81 (annex 2, table 25). The internal consistency and reliability of
the variables were tested using Cronbach’s alpha which was measured at 0.534 for
Attitude towards adoption and 0.273 for intension. Both of these tests indicate that the
internal consistency of these variables are poor.
The two tailed Pearson Correlation test between the Attitude towards adoption and
intension to adopt is 0.383 with p value less than 0.05 (annex 3). This indicated that there
is no significant correlation between the two variables at 95% confidence level. Therefore
the Null hypothesis was accepted. The accepted null hypothesis is “There is weak
influence of attitude towards adoption and adoption intension in the context utilitarian
Mobile Data Services in Sri Lanka”. To further analyse the nature of the nature of the
relationship between these two variables regression analysis was conducted. The adjusted
R squared value was 0.144 indicating that the predictability of the linear model at 95%
confidence level was 14.4%. Based on the regression analysis an individual linear
prediction model was built based on the equation of y = 3.923 + .569 X1. (y being the
Intension to adopt and X being the Attitude towards adoption). Two tests were conducted
prior to the acceptance of the simple linear regression model. First, ANOVA testing of
the model was conducted at 95% confidence level. The model was accepted based on this
test. Residual analysis of the model was conducted as the second test. Sparse and broken
45 degree upward sloping plot observed. Low random instances were observed in the
scatter plot. Therefore it is recommended that this liner model be rejected.
112
113
6.6.2 Hedonic product – Mobile Ringtone adoption model testing
The general construct of the questions and scales used in testing the variables are similar
to those described in the earlier section. Therefore comments on the likert scaling will not
be included in the following analysis of response data. Detailed statistical information
relating to mean, standard deviation and frequencies are available in Annex 2.
Perceived Usefulness
The mean value of the responses was measured at 1.56 with a standard deviation of 1.8
(annex 2, table 28). This response mean indicates that the majority of respondents
strongly disagreed with the variables assigned to measure perceived usefulness in the
context of Mobile Ringtone usage. The internal consistency and reliability of the
variables were tested using Cronbach’s alpha which measured at 0.989 (table 25). The
two tailed Pearson Correlation test between the variable and Attitude towards adoption
indicated a correlation of 0.922 with p value less than 0.05(annex 4). This indicated that
there is significant correlation between the two variables at 95% confidence level.
Therefore the Null hypothesis was rejected and alternative hypothesis of “There is strong
influence of perceived usefulness in the context of attitude towards adoption of hedonic
Mobile Data Services in Sri Lanka” was accepted. To further analyse the nature of the
nature of the relationship between these two variables regression analysis was conducted.
The adjusted R squared value was 0.855 indicating that the predictability of the linear
model at 95% confidence level was 85.5%. Based on the regression analysis an
individual linear prediction model was built based on the equation of y = .092 + .332 X1.
(y being the Attitude towards adoption and X being the Perceived usefulness). Two tests
were conducted prior to the acceptance of the simple linear regression model. First,
ANOVA testing of the model was conducted at 95% confidence level. The model was
accepted based on this test. Residual analysis of the model was conducted as the second
test. The Normal probability plot indicated a 45 degree sloping plot while the scatter plot
indicated even horizontal band around residual value of zero. Therefore it is
recommended that this liner model be accepted. In addition to these statistical
114
relationships the correlation between Perceived Usefulness and Intension to adopt was
observed at 0.69. This strength of correlation was above that between the Attitude
towards adoption and Intension to adopt.
Perceived Ease of Use
The mean value of the responses was measured at 1.6 with a standard deviation of 1.9.
This response indicates that the majority of respondents disagreed with the variables
assigned to measure perceived ease of use in the context of Mobile Ringtone usage. The
internal consistency and reliability of the variables were tested using Cronbach’s alpha
which was measured at 0.988 (table 25) indicating a very good fit. The two tailed Pearson
Correlation test between the variable and Attitude towards adoption indicated a
correlation of 0.93 with p value less than 0.05(annex 4). This indicated that there is
significant correlation between the two variables at 95% confidence level. Therefore the
Null hypothesis was rejected and alternative hypothesis of “There is strong influence of
perceived ease of use in the context of attitude towards adoption of hedonic Mobile Data Services
in Sri Lanka” was accepted. To further analyse the nature of the relationship between these
two variables regression analysis was conducted. The adjusted R squared value was 0.865
indicating that the predictability of the linear model at 95% confidence level was 86.5%.
Based on the regression analysis an individual linear prediction model was built based on
the equation of y = .029+ .494X1 (y being the Attitude towards adoption and X being the
Perceived Ease of use). Two tests were conducted prior to the acceptance of the simple
linear regression model. First, ANOVA testing of the model was conducted at 95%
confidence level. The model was accepted based on this test. Residual analysis of the
model was conducted as the second test. The Normal probability plot indicated a 45
degree sloping plot while the scatter plot indicated even horizontal band around residual
value of zero. Therefore it is recommended that this liner model be accepted. In addition
to these statistical relationships the correlation between Ease of use and Intension to
adopt was observed at 0.714. This is the second highest noted strength between the
identified independent variables of utilitarian motive and the Intension to use variable.
115
This strength of correlation was above that between the Attitude towards adoption and
Intension to adopt.
Comparative Advantage
The mean value of the responses was measured at 1.47 with a standard deviation of 1.75.
This response mean indicates that the majority of respondents did not agree with the
proposition. The internal consistency and reliability of the variables were tested using
Cronbach’s alpha which was measured at 0.93 (table 25) indicating a very good fit. The
two tailed Pearson Correlation test between the variable and Attitude towards adoption
indicated a correlation of 0.965 with p value less than 0.05(annex 4). This indicated that
there is significant correlation between the two variables at 95% confidence level.
Therefore the Null hypothesis was rejected and alternative hypothesis of “There is strong
influence of relative advantage in the context of attitude towards adoption of hedonic Mobile
Data Services in Sri Lanka” was accepted. To further analyse the nature of the nature of the
relationship between these two variables regression analysis was conducted. The adjusted
R squared value was 0.931 indicating that the predictability of the linear model at 95%
confidence level was 93.1%. Based on the regression analysis an individual linear
prediction model was built based on the equation of y = .005+ .564X1 (y being the
Attitude towards adoption and X being the comparative advantage). Two tests were
conducted prior to the acceptance of the simple linear regression model. First, ANOVA
testing of the model was conducted at 95% confidence level. The model was accepted
based on this test. Residual analysis of the model was conducted as the second test. The
Normal probability plot indicated a 45 degree sloping plot while the scatter plot indicated
even horizontal band around residual value of zero. Therefore it is recommended that this
liner model be accepted. In addition to these statistical relationships the correlation
between comparative advantage and Intension to adopt was observed at 0.721. This
strength of correlation was above that between the Attitude towards adoption and
Intension to adopt.
116
Social Influences
The mean value of the responses was measured at 1.63 with a standard deviation of 1.9.
This response mean indicates that the majority of respondents did not agree with the
proposition. The internal consistency and reliability of the variables were tested using
Cronbach’s alpha which was measured at 0.984 (table 25) indicating a very good fit. The
two tailed Pearson Correlation test between the variable and Attitude towards adoption
indicated a correlation of 0.725 with p value less than 0.05(annex 4). This indicated that
there is significant correlation between the two variables at 95% confidence level.
Therefore the Null hypothesis was rejected. The accepted alternative hypothesis was
“There is strong influence of social influence in the context of attitude towards adoption of
hedonic Mobile Data Services in Sri Lanka”. To further analyse the nature of the nature of
the relationship between these two variables regression analysis was conducted. The
adjusted R squared value was 0.840 indicating that the predictability of the linear model
at 95% confidence level was 84.%. Based on the regression analysis an individual linear
prediction model was built based on the equation of y = .041+ .489X1. (y being the
Attitude towards adoption and X being the Social influences). Two tests were conducted
prior to the acceptance of the simple linear regression model. First, ANOVA testing of
the model was conducted at 95% confidence level. The model was accepted based on this
test. Residual analysis of the model was conducted as the second test. The Normal
probability plot indicated a 45 degree sloping plot while the scatter plot indicated even
horizontal band around residual value of zero. Therefore it is recommended that this liner
model be accepted. In addition to these statistical relationships the correlation between
social influences and Intension to adopt was observed at 0.725.
This strength of correlation was above that between the Attitude towards adoption and
Intension to adopt.
117
Pleasure
The mean value of the responses was measured at 1.16 with a standard deviation of 1.41
indicating that most respondents felt a pleasurable emotion associated with using mobile
ringtones. The internal consistency and reliability of the variables were tested using
Cronbach’s alpha which was measured at 0.954 (table 25) for the measures indicating
good fit. The two tailed Pearson Correlation test between the variable and Attitude
towards adoption indicated a correlation of 0.941 with p value less than 0.05(annex 4).
This indicated that there is significant correlation between the two variables at 95%
confidence level. Therefore the Null hypothesis was rejected and alternative hypothesis
of “There is strong influence of pleasure in the context of attitude towards adoption of hedonic
Mobile Data Services in Sri Lanka” was accepted. To further analyse the nature of the
relationship between these two variables regression analysis was conducted. The adjusted
R squared value was 0.885 indicating that the predictability of the linear model at 95%
confidence level was 88.5%. Based on the regression analysis an individual linear
prediction model was built based on the equation of y = .044+ .684X1. (y being the
Attitude towards adoption and X being the pleasure). Two tests were conducted prior to
the acceptance of the simple linear regression model. First, ANOVA testing of the model
was conducted at 95% confidence level. The model was accepted based on this test.
Residual analysis of the model was conducted as the second test. The Normal probability
plot indicated a 45 degree sloping plot while the scatter plot indicated even horizontal
band around residual value of zero. Therefore it is recommended that this liner model be
accepted. In addition to these statistical relationships the correlation between Pleasure
and Intension to adopt was observed at 0.685.
Arousal
The mean value of the responses was measured at 0.411 with a standard deviation of 0.62
indicating that most respondents felt emotions of arousal associated with using mobile
ringtones. The internal consistency and reliability of the variables were tested using
Cronbach’s alpha which was measured at 0.93. This was noted to be a good fit. The two
118
tailed Pearson Correlation test between the variable and Attitude towards adoption
indicated a correlation of 0.742 with p value less than 0.05(annex 4). This indicated that
there is significant correlation between the two variables at 95% confidence level.
Therefore the Null hypothesis was rejected and alternative hypothesis of “There is strong
influence of arousal in the context of attitude towards adoption of hedonic Mobile Data Services
in Sri Lanka” was accepted. To further analyse the nature of the nature of the relationship
between these two variables regression analysis was conducted. The adjusted R squared
value was 0.549 indicating that the predictability of the linear model at 95% confidence
level was 54.9%. Based on the regression analysis an individual linear prediction model
was built based on the equation of y = .331+ 1.226X1. (y being the Attitude towards
adoption and X being the arousal). Two tests were conducted prior to the acceptance of
the simple linear regression model. First, ANOVA testing of the model was conducted at
95% confidence level. The model was accepted based on this test. Residual analysis of
the model was conducted as the second test. The Normal probability plot indicated a 45
degree sloping plot while the scatter plot indicated even horizontal band around residual
value of zero. Therefore it is recommended that this liner model be accepted. In addition
to these statistical relationships the correlation between arousal and Intension to adopt
was observed at 0.595.
Dominance
The mean value of the responses was measured at 0.22 with a standard deviation of 0.47
indicating that most respondents felt emotions associated with dominance while using
mobile ringtones.. The internal consistency and reliability of the variables were tested
using Cronbach’s alpha which was measured at 0.923. The two tailed Pearson Correlation
test between the variable and Attitude towards adoption indicated a correlation of 0.538
with p value less than 0.05(annex 4). This indicated that there is no significant correlation
between the two variables at 95% confidence level. Therefore the Null hypothesis was
rejected and the accepted alternative hypothesis is “There is strong influence of dominance
in the context of attitude towards adoption of hedonic Mobile Data Services in Sri Lanka”.
119
To further analyse the nature of the nature of the relationship between these two variables
regression analysis was conducted. The adjusted R squared value was 0.288 indicating
that the predictability of the linear model at 95% confidence level was 28.8%. Based on
the regression analysis an individual linear prediction model was built based on the
equation of y = .578+ 1.162X1. (y being the Attitude towards adoption and X being the
dominance Two tests were conducted prior to the acceptance of the simple linear
regression model. First, ANOVA testing of the model was conducted at 95% confidence
level. The model was accepted based on this test. Residual analysis of the model was
conducted as the second test. The Normal probability plot indicated a 45 degree sloping
plot while the scatter plot indicated even horizontal band around residual value of zero.
Therefore it is recommended that this liner model be accepted. In addition to these
statistical relationships the correlation between arousal and Intension to adopt was
observed at 0.421.
Attitude towards adoption and intension The mean value of the responses for attitude towards adoption was measured at 0.837
with a standard deviation of 1.03. The mean value for intension to adopt was measured at
3.56 with a standard deviation of 1.12. The internal consistency and reliability of the
variables were tested using Cronbach’s alpha which was measured at 0.808 for Attitude
towards adoption and 0.342 for intension. While the goodness of fit of the attitude
towards adoption was within acceptable range acceptability and consistency of intension
to adopt failed. The two tailed Pearson Correlation test between the Attitude towards
adoption and intension to adopt is 0.672 with p value less than 0.05(annex 4). This
indicated that there is significant correlation between the two variables at 95% confidence
level. Therefore the Null hypothesis was rejected. The accepted alternative hypothesis is
“There is strong influence of attitude towards adoption and adoption intension in the
context hedonic Mobile Data Services in Sri Lanka”. To further analyse the nature of the
nature of the relationship between these two variables regression analysis was conducted.
The adjusted R squared value was 0.450 indicating that the predictability of the linear
model at 95% confidence level was 45.0%. Based on the regression analysis an
individual linear prediction model was built based on the equation of y = 3.923 + .569 X1.
120
(y being the Intension to adopt and X being the Attitude towards adoption). Two tests
were conducted prior to the acceptance of the simple linear regression model. First,
ANOVA testing of the model was conducted at 95% confidence level. The model was
accepted based on this test. Residual analysis of the model was conducted as the second
test. The Normal probability plot indicated a 45 degree sloping plot while the scatter plot
indicated even horizontal band around residual value of zero. Therefore it is
recommended that this liner model be accepted.
121
7. Discussion Based on the accepted hypothesis, the following tables were produced to identify existing
literature that support these findings.
Utilitarian value proposition and variables accepted
Hypothesis
No
Measured Variable Correlation
(at 95%
significance
– two tailed
test)
R squared
value
Relevant Literature
Hypothesis
1
Perceived Usefulness and attitude
towards adoption in the context of an
utilitarian product
0.79 61.5% Pedersen et al. (2002)
Nysveen et al. (2005)
Kulviwat et al. (2007)
Bruner II & Kumar (2005)
Kim et al. (2009)
Hypothesis
2
Perceived ease of use and attitude
towards adoption in the context of an
utilitarian product
0.74 54.5 % Pedersen et al. (2002)
Nysveen et al. (2005)
Kulviwat et al. (2007)
Bruner II & Kumar (2005)
Kim et al. (2009)
Hypothesis
3
Relative advantage and attitude towards
adoption in the context of an utilitarian
product
0.71 50.4% Kulviwat et al. (2007)
Rogers (2005)
Hypothesis
5
Pleasure and attitude towards adoption
in the context of an utilitarian product
0.52 27.3% Kulviwat et al. (2007)
Wu et al. (2008) Hypothesis
6
Arousal and attitude towards adoption in
the context of an utilitarian product
0.53 27.6% Kulviwat et al. (2007)
Wu et al. (2008)
122
Hedonic value proposition and variables accepted
No Measured Variable Correlation
(at 95%
significance
– two tailed
test)
R squared
value
Relevant Literature
Hypothesis 9
Perceived Usefulness and
attitude towards adoption in
the context of an hedonic
product
0.92 85% Pedersen et al. (2002)
Nysveen et al. (2005a)
Kulviwat et al. (2007)
Hypothesis 10
Perceived ease of use and
attitude towards adoption in
the context of an hedonic
product
0.93 86.5% Pedersen et al. (2002)
Nysveen et al. (2005a)
Kulviwat et al. (2007)
Hypothesis 11
relative advantage and attitude
towards adoption in the
context of hedonic product
0.97 93.1% Kulviwat et al. (2007)
Hypothesis 12
social influences and attitude
towards adoption in the
context of an hedonic product
0.92 84% Kulviwat et al. (2007)
Kulviwat et al. (2008)
Hypothesis 13
Pleasure and attitude towards
adoption in the context of an
hedonic product
0.94 88.5% Kulviwat et al. (2007)
Hypothesis 14
Arousal and attitude towards
adoption in the context of an
hedonic product
0.74 54.9% Kulviwat et al. (2007)
Wu et al. (2008)
Hypothesis 15
Dominance and attitude
towards adoption in the
context of an hedonic product
0.54 28.8% Nasco et al. (2008)
123
Discussion
This research into the development of models to analyze the relationship between the
value propositions of mobile data services and motives towards adoption in Sri Lanka has
national and international significance. Within the national context, there are no
published research or statistics on Mobile Data Services adoption rates, important and
emerging mobile data services trends, important mobile data services infrastructure or
demographic trends on mobile data services popularity. Commenting on this point
Carlsson et al. 2005) notes that the mobile telecommunication industry is still focused on
selling handset instead of on mobile data services diffusion. The focus on mobile data
services in Sri Lanka also remains at an embryonic stage. This can be observed by the
publications of the national telecommunications regulator of Sri Lanka only focusing on
mobile handset penetration rates (TRC-SL 2008). The area of mobile data services truly
remains a blind spot within the regulatory and industry context in the island. This
weakness in industry focus is indeed concerning considering that the future survival of
the telecommunication industry will depend on mobile data services as indicated by the
steadily declining average revenue per user on voice charges across the global
telecommunication industry (ABI Research 2009).
Within a global context it is noted by Gao & Rafiq (2009) that there is general lack of
primary research on the mobile telecommunication industry in developing countries as
oppose that of developed countries. While the underlying technology remains the same,
the cultural and social influences need to be better understood within developing
countries. This is very important from mobile data services perspective because these
services go beyond the homogeneous nature of voice and propel the notions of
convenience and personalization (Clarke & Flaherty 2003)(as cited by (Heinonen & Pura
2006). Therefore this research attempt to use a soundly tested and accepted technology
adoption model and observe its behavior within the Sri Lankan user context. It is hoped
that this research would lay the foundation towards building of a localized adoption
model for the country.
124
The research findings indicate that the Sri Lanka Consumer Acceptance of Technology
model (SL-CAT) presents a valid basis for analysis and prediction of technology
adoption intension. The prediction capability of SL-CAT was derived based on multiple
regression modeling. Here the prediction capability of the SL-CAT for SMS was
recorded at 40% and for mobile ringtones at 54%.This prediction capability of SL-CAT is
better than Technology Adoption Model prediction rates of 17%-33% (Davis 1989; Davis
et al. 1989; Chau & Hu 2001)(as cited by (Kulviwat et al. 2007). Further in comparison
to adoption model presented by Kulviwat et al. (2007) where the prediction capability
was listed at 53%, SL-CAT results for hedonic product of Mobile ringtones was on par.
However, in comparison to the models presented by Pedersen et al. (2002) and Nysveen
et al. (2005a) these results of SL-CAT are relatively weak. In the two comparable studies
for SMS (62%) and mobile gaming (67%) the recorded capabilities of Pedersen et al.
(2002) models are better than SL-CAT. Nysveen et al. (2005a) models noted an average
prediction capability of 72% for SMS and Mobile Games.
When analyzing the reasons for deviation between the SL-CAT and those of Pedersen et
al. (2002) and Nysveen et al. (2005), the method in which hedonic motives are analyzed
are a key factor. The SL-CAT uses the complex dimensions of Mehrabian & Russell
(1974) model of pleasure, arousal and dominance motives to record hedonic motives. In
comparison the models of Pedersen et al. (2002) and Nysveen et al. (2005) use a single
variable of “perceived enjoyment” to capture hedonic motives. Therefore, instead of
understanding the dimensionalities of the hedonic motives, these theoretical models can
be observed as “rounding-off” all the hedonic motives into one basket. However, if
researchers and mobile telecommunication industry are to better understand the
dimensionality of hedonics it is crucial that researchers go beyond the all encompassing
basket of “fun and enjoyment”. While accepting that the pleasure, arousal and dominance
variable may not be comprehensive, they do present an equitable starting point. Therefore
it could be argued that SL-CAT has the ability to better understand the motive of
hedonics in comparison to the existing research models.
125
The significant deviation between the prediction results for SMS (predictability 40%) and
Ringtones (predictability 54%) was considered as a point of concern. One of the key
reasons for the low predictability rate of the SMS model was the low internal consistency
and reliability of the variables as tested using Cronbach’s alpha. The variables of pleasure
(0.66), arousal (0.71), attitude towards adoption (0.534) and adoption intension (0.273)
recorded the lowest internal consistency rates. The lowest rates of internal consistence
acceptable was above 0.7 as noted by Hair et al. (1998)(as cited by Kulviwat et al.
(2007). The unacceptably low rates of Cronbach’s alpha for attitude towards adoption
and adoption intension had directly resulted in diminishing the integrity of the SL-CAT
model. While efforts were undertaken to optimize the internal consistency rate, these
were unsuccessful in significantly inducing improvement. One possible reasons for these
low alpha scores were considered to be related to the translation of the questionnaire from
English to Sinhala and Tamil. However in comparison the internal consistency rates in
the SMS model, the variables of attitude towards adoption and intension to adopt motives
were recorded alpha rates of at 0.808 and 0.342 in the mobile ringtone. Unlike in the SL-
CAT SMS model, all other independent variables recorded healthy rates above 0.8 in the
SL-CAT mobile ringtone study. Therefore these identified anomalies in the design of the
questionnaire and its relation to the testing variable needs to be improved.
When presenting the Consumer Acceptance of Technology model, Kulviwat et al. (2007)
notes that all variables had internal consistency rates between 0.76 to 0.93. The internal
consistency rate noted by Pedersen et al. (2002) and Nysveen et al. (2005a) were above
0.75. Based on this analysis it is important to note that due to the poor design of the SL-
CAT questionnaire the overall validity of the model has got effected adversely. This is
especially acute in the context of the variables of attitude towards adoption and adoption
intension. Therefore an improved and refined questionnaire design may provide better
insights into the prediction capabilities of the SL-CAT model.
Noting the failure of the SMS model in significantly predicting adoption intension, it was
decided to further analyze the relationship between the identified independent variables
and attitude towards adoption. The objective of this process was to present a scale-down
126
model of prediction for discussion. Based on the analysis, two models were built and the
accepted final model incorporated the five variables of Perceived Usefulness, Ease of use,
comparative advantage, pleasure and arousal. The multiple linear regression model
presented a prediction capability of 71.7%. Therefore in considering these results, it
could be stated with reliability that the low internal consistency rates and low correlation
between the design constructs of attitude towards adoption and adoption intension
resulted in the SMS model having an overall predictability of 40%.
Model for SMS and individual variables
The proposed adoption model for SMS product identified five key variables have a direct
effect on influencing the adoption of this mobile data service. The strongest influencer of
attitude is the variable of perceived usefulness. Revalidating the theoretical propositions
of (Davis 1989; Davis et al. 1989) the perceived usefulness emerged as the strongest
variable with a direct relationship to attitude towards adoption. While Kulviwat et al.
(2007) observed with surprise the strong relationship between usefulness and adoption
intension, this same nature of relationship was identified in this research where the
correlation between usefulness and intension was 0.451. Usefulness also showed strong
relationship between the ease of use (0.815) and comparative advantage (0.821)
variables. This indicates that more appreciation the customer has about the usefulness of
this product, greater will be their motivation to use and consume the service.
Comparative advantage is a variable with limited research beyond those conducted by
Rogers (2005). Until the variable was adopted into the Consumer Acceptance of
Technology model, there was no available research into the nature of relationship
between comparative advantage and perceived usefulness. While Kulviwat et al. (2007)
research confirmed the strong correlation between these two variables, this SL-CAT
research in Sri Lanka also re-discovers this strong relationship. However the research
conducted by Kulviwat et al. (2007) did not identify any significant interrelationship
between relative advantage and perceived ease of use. However the SL-CAT research
discovers that there is indeed a strong relationship between comparative advantage and
127
ease of use (0.832). This would mean that improving the ease of use of the product
improves its comparative advantage in the perception of the customer. This was noted by
Jenson (2006), although without market research that, the consumers attraction towards
SMS instead of MMS due to its simple design and few steps involved in operating the
service. This relationship needs to be further explored in the context of other mobile data
services, especially during the product designing stages and product localizing stages.
This research indicates that the adoption of mobile services, with cognitive utilitarian
motives such as SMS could be influenced by focusing on this relationship.
While the relationship between perceived ease of use and attitude towards adoption was
not established by research conducted by Kulviwat et al. (2007), the SL-CAT research
model firmly establishes this relationship. Extending beyond the observed correlation
between ease of use and attitude towards adoption, the research indicates that ease of use
may also influence intension, at a moderate level (correlation of 0.49 was observed). It
should also be noted that from the group of utilitarian motives that were tested in this
research, perceived ease of use had the strongest relationship with adoption intension.
These relationships are new discoveries in the available literature and may be unique in
the Sri Lankan consumer environment. It was also noted with surprise that ease of use
had a moderate relationship (0.45) with social influences. This is indeed unique
relationship which could point to the general perceptions or societal attitude towards the
ease of using a given mobile data service positively or negatively influencing the final
adoption decision. This research finding was also observed by Lu et al. (2005) in a recent
research into social influences and adoption of technology. Further this research
discovery also has a very positive potential towards the diffusion of mobile data services
in Sri Lanka., indicating that if the societies perceptions could be changed through
sustained education and information such as by advertising of the ease of use of mobile
data services, this may have a positive impact on adoption.
The role of societal influence has been a key discussion issue in information systems
research. While the initial Consumer acceptance of technology model (Kulviwat et al.
2007) did not incorporate societal influences subsequent research by (Kulviwat et al.
128
(2008) identified the importance of this variable in adoption. This research into adoption
of SMS shows, surprisingly that social influences have a strong influence on adoption
intension (0567) than on attitude toward adoption. This finding is different from those of
Pedersen et al. (2002) where the social influences were considered very weak in
influencing adoption intension. The researcher points to the fact that in Scandinavian
markets where mobile penetration rates have exceeded 80%, mobile data service products
such as SMS are considered mature technologies and is used by a majority of the
population. In comparison to this setting, in Sri Lanka mobile data services may be
considered a relatively new technology that is rapidly gaining ground. Therefore the
influence of society on the adoption intension is significant. Indeed in this research,
societal influences pointed to a stronger relationship to adoption intension than the
attitude towards adoption. This finding suggests that an individual may adopt the
technology primarily due to societal influences rather than a specific utilitarian or
hedonic motive. Here again, this research is important for future strategy building and
marketing budget allocations of mobile telecommunications providers. The greater effort
that is invested to influence peers and opinion leaders to adopt the technology, would
result in greater overall adoption.
In selecting the Consumer Acceptance of technology model, one of the main interests of
the research was to identify the behavior of motives of cognition and hedonics on
products categorized as having predominantly utilitarian and hedonic value propositions.
Therefore, it was not expected to have significant readings of the pleasure, arousal and
dominance motives in SMS product. The research finding also confirmed these
assumptions to a greater extent due to only pleasure (0.524) and arousal (0.527) having
significant correlation with attitude towards adoption. However, unlike the research
findings of Kulviwat et al. (2007) arousal indicated to have the same level of correlation
as pleasure with attitude towards adoption. In further analyzing this relationship it was
noted that there is a significantly strong correlation between pleasure and arousal of
0.675. this relationship was well above the levels identified by Kulviwat et al. (2007) of
0.54. It is indeed interesting as to why the researchers have not commented on this
relationship. While the measures to detect pleasure and arousal are closely aligned, this
129
research finding requires further study and analysis to identify how increasing pleasure
would also result in an increase in arousal. Based on the presented research data it could
be noted that these two variable behave similarly, even in their influence of adoption
intension (pleasure – 0.17, Arousal .143). It is also noteworthy that the motives of
pleasure and arousal were present in users of a predominantly utilitarian product.
Therefore it could be concluded that in order to successfully launch a product with
utilitarian dimensionalities, the presence of hedonic motivators are also important.
It was not surprising to observe that the dominance motive was not present in the strong
influencers of attitude or intension. This variable had failed to be present in the original
Consumer Adoption of Technology model presented by Kulviwat et al. (2007). However
the importance or lack of it should not be discounted in the overall analysis. While
dominance is in essence relating to being in control or being under the control of another,
this research finds that dominance motive was not present among the surveyed
respondents. It is however noteworthy of Nasco et al. (2008) research finding of
dominance being a hidden covert motive than an overt motive and being task specific.
Therefore, it is suggested that further research into understanding the role of dominance
be undertaken.
130
Model for Ringtones and individual variables
The selection of mobile ringtones to test the hedonic motive was not the original plan of
the research. It was originally planned to use mobile games, as this was the product of
choice to study hedonics in global markets. However, to the surprise of the researcher
only ten of the forty respondents of the pilot study group who were aged between 18-21
played mobile games regularly. The balance ten respondents of aged between 22-35,
while acknowledging their awareness of mobile games did not use mobile games
frequently. Therefore it was necessary to find a viable alternative which displayed the
characteristics of a product used with hedonic intension and which was being used by a
majority of the test population. The selection of the mobile ringtone was undertaken
based on these criteria. It was also noted by the researcher (subjectively) that the only
mobile data service aggressively being position to the Sri Lankan consumers are mobile
ringtones. While this observation needs to be proven or disproved based on market
research, mobile ringtones were selected for the research as the next best alternative to
mobile games.
The proposed model for predicting Mobile Ringtone adoption was considered more
successful in its overall capability than that of the SMS model. Unlike the SMS model
this model uses the utilitarian variables, Hedonic and social influences. The strength of
the correlation between perceived usefulness, attitude towards adoption and adoption
intension was noted at 0.922 and 0.690 respectively. These relationships well above those
identified by Kulviwat et al. (2007). Further unlike in the context of SMS the usefulness
variable indicates to have strong correlations with utilitarian and hedonic variables alike.
This is indeed an important relational discovery. Further usefulness is not the variable
with the strongest relationship to attitude towards adoption. This is an interesting
development because all research findings of Pedersen et al. (2002), Nysveen et al.
(2005a) and Kim et al. (2009) found that perceived usefulness as the most important
indicator of relationship between attitude and adoption intension even in the context of
products with predominantly hedonic values propositions. Therefore this finding may be
unique in the context of Sri Lanka.
131
The variable of comparative advantage also recorded similar relationships with
usefulness (0.941), ease of use (0.965), social influences (0.974), pleasure (0.935),
arousal (0.754) and dominance (0.513). In Kulviwat et al. (2007) research there was no
significant relationship between the utilitarian motives and hedonics in general. However
in research done by Pedersen et al. (2002) there is a relatively weak correlation identified
between the move of perceived enjoyment and usefulness of 0.37, comparably, these
finding point to a stronger relationship in the Sri Lankan market context. The ease of use
variable also recorded strong relationships between the utilitarian and hedonic motives on
par with those of comparative advantage. These finding would indicate that while
consumer appreciates the hedonic motives of fun associated with Mobile ringtones, they
also have selected the product based on logical cognitive reasons.
Based on these observations of the behavior of utilitarian motives on predominantly
hedonically motivated products, can an explanation to minimal usage of mobile games be
propositioned?. These findings do suggest that in order for even hedonically valued
products to propagate into mass circulation, there needs to be a logical reasoning. Further
it was noted that the utilitarian motives of usefulness, ease of use and comparative
advantage were more strongly present than hedonic motives even in mobile ringtone
users. The utilitarian motives had strong correlations with the attitude towards adoption
and adoption intension than the relationship of hedonic motives with these same
dependant variables.
Unlike in the context of SMS, the correlation between social influences and adoption
intension was relatively weak (0.421). The relation between attitude towards adoption
and social influences were marginally stronger (0.538). While this would indicate that
society has a relatively moderate to weak influence on the adoption of the technology,
this conclusion requires further study. It should be noted that the overall promotion,
branding and market positioning of mobile data services in Sri Lanka remains very low.
Therefore the society and its opinion leader’s perception towards the adoption of this
132
technology may be weak to indifferent. Thus, this weakness could be leveraged by the
mobile telecommunication industry to their advantage to induce diffusion.
The variables of pleasure and arousal in the context of Mobile ringtones also behaved
with the same level of correlation between them, as in SMS (0.722). However, unlike in
the context of SMS, their correlations to adoption intension was much stronger
(pleasure:0.725; Arousal:0.685). Therefore the observation that the variables of pleasure
and arousal have the same behavior patterns, as in the case of SMS is reconfirmed. These
observations required further analysis and decision making during future research. Do we
maintain the same model of Pleasure and Arousal? or in order to improve the Consumer
Acceptance of Technology, do we substitute one variable with a better predictor variable?
In order for the general acceptance of the Consumer Acceptance of Technology model, it
needs to be able to provide greater and more robust prediction capability. Therefore it is
suggested that the models proposed by Pedersen et al. (2002) and Nysveen et al. (2005)
be evaluated with the performance of CAT and the model improved accordingly.
Once again the dominance variable behaved differently than that of pleasure and arousal.
The correlation of dominance to attitude towards adoption (0.742) and intension (0.595)
was significantly higher than in the context of SMS. It has also maintained significantly
strong relations with utilitarian and hedonic motives. While this study has identified the
behavior of dominance in the context of utilitarian and hedonic products, it is indeed
difficult to explain the implications of this motive on the overall research model. The
researcher agrees with the notion of Nasco et al. (2008), that this variable has a hidden
nature to it. Rather than discounting its importance, it is suggested that future research
focus on better understanding dominance.
133
8. Recommendations
8.1 Key determinants of Utilitarian value proposition based mobile data services
adoption variables
Utilitarian value proposition and variables accepted
Hypothesis
No
Measured Variable Correlation
(at 95%
significance
– two tailed
test)
R squared
value
Relevant Literature
Hypothesis
1
Perceived Usefulness and attitude
towards adoption in the context of an
utilitarian product
0.79 61.5% Pedersen et al. (2002)
Nysveen et al. (2005)
Kulviwat et al. (2007)
Bruner II & Kumar (2005)
Kim et al. (2009)
Hypothesis
2
Perceived ease of use and attitude
towards adoption in the context of an
utilitarian product
0.74 54.5 % Pedersen et al. (2002)
Nysveen et al. (2005)
Kulviwat et al. (2007)
Bruner II & Kumar (2005)
Kim et al. (2009)
Hypothesis
3
Relative advantage and attitude towards
adoption in the context of an utilitarian
product
0.71 50.4% Kulviwat et al. (2007)
Rogers (2005)
Hypothesis
5
Pleasure and attitude towards adoption
in the context of an utilitarian product
0.52 27.3% Kulviwat et al. (2007)
Wu et al. (2008) Hypothesis
6
Arousal and attitude towards adoption in
the context of an utilitarian product
0.53 27.6% Kulviwat et al. (2007)
Wu et al. (2008)
134
8.2 Key determinants of hedonic value proposition based mobile data services
adoption variables
Hedonic value proposition and variables accepted
No Measured Variable Correlation
(at 95%
significance
– two tailed
test)
R squared
value
Relevant Literature
Hypothesis 9
Perceived Usefulness and
attitude towards adoption in
the context of an hedonic
product
0.92 85% Pedersen et al. (2002)
Nysveen et al. (2005a)
Kulviwat et al. (2007)
Hypothesis 10
Perceived ease of use and
attitude towards adoption in
the context of an hedonic
product
0.93 86.5% Pedersen et al. (2002)
Nysveen et al. (2005a)
Kulviwat et al. (2007)
Hypothesis 11
relative advantage and attitude
towards adoption in the
context of hedonic product
0.97 93.1% Kulviwat et al. (2007)
Hypothesis 12
social influences and attitude
towards adoption in the
context of an hedonic product
0.92 84% Kulviwat et al. (2007)
Kulviwat et al. (2008)
Hypothesis 13
Pleasure and attitude towards
adoption in the context of an
hedonic product
0.94 88.5% Kulviwat et al. (2007)
Hypothesis 14
Arousal and attitude towards
adoption in the context of an
hedonic product
0.74 54.9% Kulviwat et al. (2007)
Wu et al. (2008)
Hypothesis 15
Dominance and attitude
towards adoption in the
context of an hedonic product
0.54 28.8% Nasco et al. (2008)
135
8.1 SL-CAT Relational Model 1 – Relationship between Utilitarian, Hedonic and
Social influences on Utilitarian value propositioned Mobile Data Services
Product - SMS
Based on the model building exercise to identify an optimum solution between the
identified variables of perceived usefulness, ease of use, comparative advantage,
pleasure, arousal, dominance and social influences, the SL-CAT relational model for
SMS adoption is recommended. The diagram below notes the identifies correlations.
8.2 SL-CAT Relational Model 2 – Relationship between Utilitarian, Hedonic and
Social influences on Hedonic value propositioned Mobile Data Services Product
– Mobile Ringtones
Based on the model building exercise to identify an optimum solution between the
identified variables of perceived usefulness, ease of use, comparative advantage,
pleasure, arousal, dominance and social influences, the following model is recommended.
136
0.785
0.711
0.524
0.527
0.303
0.383
0.407 0.451
0.408
0.567
0.567
0.815
0.832 0.821
SL-CAT model 1-correlation diagram of Utilitarian Product of SMS adoption model
137
0.421
0.538 0.922
0.714 0.930
0.965 0.721
0.917
0.941
0.742 0.595
0.685
0.725
0.627
SL-CAT Model 2 - Correlation diagram of Hedonic Product of Mobile Ringtone adoption model
0.941
0.722
0.491
138
Proposed staircase model for adoption of Mobile Data Services
139
8.3 Prediction Model 1 – Utilitarian Product of SMS adoption intension
Proposed Model Two models were developed in relation to the attitude towards adoption and adoption
intension. This was due to the failure of the adoption intension model to develop strong
prediction capability. The attitude towards adoption model was prepared as a secondary,
scaled-down version of the original model. The adoption intension model had a
prediction capability of 40% while incorporating the independent variables of perceived
usefulness, perceived ease of use, comparative advantage, social influences, pleasure and
arousal. While the overall significance of the model in terms of ANOVA analysis was
acceptable at 95% confidence level, individual variable of of Perceived Usefulness,
Perceived Ease of Use, Comparative advantage, Pleasure and Arousal all recorded p
values greater than 0.05. The multiple liner regression equation for the model as noted
below. The second scale down model on attitude towards adoption presented a prediction
capability of 71.4% . This model incorporated the variables of Perceived Usefulness,
Perceived Ease of Use, Comparative Advantage, Pleasure and Arousal. These selected
variables had a correlation of above 0.5 between them and attitude towards adoption.
while the overall significance of the model in terms of ANOVA analysis was acceptable
PREDICTION CAPABILITY = 40%
Y = 3.139+ 0.165 EOU + 0.175 DO + 0.316 SO – 0.119 ATA + 0.85 PU – 0.032 AR + 0.158 PL
Where,
EOU – is Perceived Ease of Use
DO – Dominance
SO – Social Influences
ATA – Attitude towards adoption
PU- Perceived usefulness
AR- Arousal
PL- Pleasure
140
at 95% confidence level, individual variable of comparative advantage and pleasure
recorded individual p> 0.05.
These models can be used to predict the adoption of mobile data services.
8.4 Prediction Model 2 – Utilitarian Product of SMS attitude towards
adoption
PREDICTION CAPABILITY = 71.7%
y = .031+ .169 PU+ .098 EOU+ .044 CA+ .109 PL+ .266 AR
where,
PU – Perceived Usefulness
EOU – Perceived Ease of Use
CA – Comparative Advantage
PL – Pleasure
AR – Arousal
141
8.5 Prediction Model 2 – Prediction model for SMS based on the selected variables
of
PREDICTION CAPABILITY = 54%
y = 2.872 + .014 HePU + .200 HeEOU + .720 HeCA -.273 HeSO + .221 HePL + .286
HeAR + .241 HeDO -.841 HeATA
where,
HePU – Perceived Usefulness HeEOU – Perceived Ease of Use
HeCA – Comparative Advantage
HeSO – Social influences
HePL – Pleasure
HeAR - Arousal
HeDO - Dominance
HeATA – Attitude towards adoption
142
8.6 Proposed staircase model
The staircase model presented above represents the interaction of significant variables
involved in the adoption process. For adoption to take place each of the minimum
required variables identified through the model building process must be present.
Therefore as the customer overcomes and interacts with each of the variables, their
potential to increase using mobile data services will also increase.
143
9 Future research Refine the SL-CAT model by identifying key weakness in its construct
Research into the correlation of demographic variables of age, income, gender and the
potential to adopt new mobile data services.
144
References ABI Research, 2009. Economic conditions reducing mobile end-user spending up to
15%. IntoMobile. Available at: http://www.intomobile.com/2009/05/15/abi-research-economic-conditions-reducing-mobile-end-user-spending-up-to-15.html [Accessed May 25, 2009].
Adams, D., Nelson, R. & Todd, P., 1992. Perceived usefulness, ease of use,
and usage of information technology: A replication. MIS Quarterly, 16(2), 227–247.
Anderson, J.C. & Gerbing, D.W., 1988. Structural Equation Modeling in Practice: A
Review and Recommended Two-Step Approach. Psychological Bulletin, 103(3), 411-423.
Ayyagari, R., 2006. Examination of hedonism in TAM research. In Southern Association
for Information Systems Conferance. Florida, USA. Babin, B., Darden, W. & Griffin, M., 1994. Work and/or Fun: Measuring
Hedonic and Utilitarian Shopping. Journal of Consumer Research, 20(4), 644–656.
Bina, M., Karaiskos, D. & Giaglis, G., 2007. Motives and Barriers Affecting the Use of
Mobile Data Services. In Toronto, Ontario, Canada. Bruner II, G.C. & Kumar, A., 2005. Applying T.A.M. to consumer usage of
handheld Internet devices. Journal of Business Research, 58, 553–558. Carlsson, C., Carlsson, J. & Walden, P., 2006. Nice Mobile Services do not Fly.
Observations of Mobile Services and the Finnish Consumers. In eValues. Slovenia.
Carlsson, C. et al., 2005a. Adoption of Mobile Services across Different Technologies. In
eIntegration in Action. Bled, Slovenia. Carlsson, C. et al., 2005b. Asynchronous Adoption Patterns of Mobile Services. In
Hawaii, USA . Chau, P. & Hu, P., 2001. Information technology acceptance by individual professionals:
A model comparison approach. Decision Sciences, 32, 699–719. Childers, T.L. et al., 2001. Hedonic and utilitarian motivations for online retail shopping
behavior. Journal of Retailing, 77(4), 511-535.
145
Clarke, I. & Flaherty, T., 2003. Mobile Portals: The Development of M-Commerce Gateways. In Mobile Commerce: Technology, Theory, and Applications. Idea Group Publishing, pp. 185-201.
Cohen, J. & Areni, C., 1991. Affect and consumer behavior. In Handbook of consumer
behavior. Englewood Cliffs, NJ: Prentice-Hall. Davis, F., 1989. Perceived usefulness, perceived ease of use, and user acceptance of
information technology. MIS Quarterly, September, 318-340. Davis, F., Bagozzi, R. & Warshaw, P., 1989. User acceptance of computer technology: A
comparison of two theoretical models. Management Science, 35, 982-1003. Fernández, R.S. & Bonillo, M.Á.I., 2007. The concept of perceived value: a systematic
review of the research. Marketing Theory, 7(4), 427-451. Fishbein, M. & Ajzen, I., 1975. Belief, attitude, intention and behavior: An introduction
to theory and research, Boston: Addison-Wesley. G15, 2008. THE ICT REVOLUTION AND G-15 COUNTRIES
Suggestions for bridging the Digital Divide, G-15 COUNTRIES. Available at: www.g15.org/workingpaper4.pdf [Accessed June 7, 2009].
Gao, P. & Rafiq, A., 2009. The transformationofthemobiletelecommunicationsindustryin
Pakistan:Adevelopingcountryperspective. TelecommunicationsPolicy, 33, 309–323.
Gilbert, L. & Han, H., 2004. Understanding mobile data services adoption: Demography,
attitudes or needs ? Technological Forecasting and Social Changes, 72, 327-337. Gilbert, L. & Kendall, J.D., 2003. A Marketing Model for Mobile Wireless Services. In
Hawaii, USA . Glotz, P., Bertschi, S. & Locke, C., 2005. Thumb Culture: The Meaning of Mobile
Phones for Society, Bielefeld: transcript Verlag. Hair, J.F. et al., 1998. Multivariate
Data Analysis with Readings 5th ed., Englewood Cliffs, NJ: Prentice Hall. Heijden, V., 2004. User acceptance of hedonic information systems. MIS Quarterly,
28(4), 695-704. Heinonen, K. & Pura, M., 2006. Classifying Mobile Services. In Helsinki.
146
Hong, S., Thong, J.Y. & Tam, J.Y., 2006. Understanding continued information technology usage behavior: A comparison of three models in the context of mobile internet. Decision Support Systems, 42(3), 1819–1834.
Hu, P. et al., 1999. Examining the technology
acceptance model using physician acceptance of telemedicine technology. Journal of Management Information Systems, 16, 91–113.
ITU, 2009. Executive Summary. OECD Communications Outlook 2009, 10. Available at:
http://www.oecd.org/dataoecd/24/32/43472431.pdf [Accessed August 1, 2009]. ITU, 2007. ITU expects global mobile penetration rate to reach 50 percent in early 2008.
ITU, Press Release. Available at: http://www.itu.int/ITU-D/ict/newslog/ITU+Expects+Global+Mobile+Penetration+Rate+To+Reach+50+Percent+In+Early+2008.aspx [Accessed May 6, 2009].
ITU, 2008. Worldwide mobile cellular subscribers to reach 4 billion mark late 2008. ITU,
Press Release. Available at: http://www.itu.int/newsroom/press_releases/2008/29.html [Accessed August 1, 2009].
Jenson, S., 2006. Default Thinking: Why Mobile Services are setup to fail, "The Inside
Text: Social, Cultural and Design Perspectives on SMS". Available at: http://www.jensondesign.com/DefaultThinking.pdf [Accessed June 7, 2009].
Kim, B., Choi, M. & Han, I., 2009. User behaviors toward mobile data services: The role
of perceived fee and prior experience. Expert Systems with Applications, 36(4), 8528-8536.
Kim, H., Chan, H.C. & Gupta, S., 2007. Value-based Adoption of Mobile Internet:
An empirical investigation. Decision Support Systems, 43(1), 111-126. Kulviwat, S., Bruner II, G.C. & Al-Shuridah, O., 2008. The role of social influence on
adoption of high tech innovations: The moderating effect of public/private consumption. Journal of Business Research, 62(7), 706-712.
Kulviwat, S. et al., 2007. Toward a Unified Theory
of Consumer Acceptance Technology. Psychology & Marketing, 24(12), 1059–1084.
Kunin, C., Vickery, G. & Wunsch-Vincent, S., 2005. DIGITAL BROADBAND
CONTENT: MOBILE CONTENT NEW CONTENT FOR NEW PLATFORMS, France: Organisation for Economic Co-operation and Development.
147
Lee, M., Suh, K. & Whang, J., 2003. The impact of situation awareness information on consumer attitudes in the Internet shopping mall. Electronic Commerce Research and Applications, 2(3), 254–265.
López-Nicolás, C. & Molina-Castilio, F., 2008. An assessment of advanced mobile
services acceptance: Contributions from TAM and diffusion theory models. Information and Management, 45, 359-364.
Lu, J., Yao, J. & Yu, C., 2005. Personal innovativeness, social influences and adoption of
wireless internet services via mobile technology. Journal of Strategic Information Systems, 14, 245-268.
Mälarstig, Y., Holmström, O. & Davidsson, P., 2007. The Swedish Telecommunications
Market 2007, Stockholm, Sweden: National Post and Telecom Agency. Marketing Sciences Institute, 2006. Research Priorities (2006–2008) A guide to MSI
research programs and procedures, Marketing Sciences Institute. Available at: http://www.msi.org/pdf/MSI_RP06-08.pdf [Accessed June 1, 2006].
Mehrabian, A. & Russell, J.A., 1974. An Approach to Environmental Psychology.
In MIT Press. Moore, G.C. & Benbasat, I., 1991. Development of an instrument to measure the
perceptions of adopting an information technology innovation. Information Systems Research, 2, 192–222.
Nasco, S.A. et al., 2008. The CAT model: Extensions and moderators of dominance in
technology acceptance. Psychology and Marketing, 25(10), 987-1005. Nielsen Research, 2008. Realizing Potential: Overcoming Barriers to the U.S. Mobile
Advertising Market, USA: The Nielsen Company. Nunnally, J.C. & Bernstein, I., 1994. Psychometric theory 3rd ed., New York: McGraw-
Hill. Nysveen, H., Pedersen, P.E. & Thorbjørnsen, H., 2005a. Explaining intension to use
mobile chat services: Moderating effects of gender. Journal of Consumer Marketing, 25(5), 247-256.
Nysveen, H., Pedersen, P.E. & Thorbjørnsen, H., 2005b. Intentions to Use Mobile
Services: Antecedents and Cross-Service Comparisons. Journal of the Academy of Marketing Science, 33(3), 330-346.
148
Pedersen, P.E., Methlie, L. & Nysveen, H., 2008. An exploratory study of business model design and costumer value in heterogeneous network services, Bergen, Norway: Foundation for Research in Economics and Business Administration.
Pedersen, P.E., Nysveen, H. & Thorbjørnsen, H., 2002. The adoption of mobile services:
A cross service study., Norway: Foundation for Research in Economics and Business Administration.
Plouffe, C.R., Hulland, J.S. & Vandenbosch, M., 2001. Research report: Richness
versus parsimony in modeling technology adoption decisions: Understanding merchant adoption of a smart card-based payment system. Information Systems Research, 12, 208–222.
Ratan, S.R.A., Alam, M.S. & Sohel-Uz-Zaman, A.S.M., 2007. Innovation Strategy: A
Realistic Example in Bangladesh Mobile Telecommunication Service Marketing. The Cost and Management, 35(1), 49-58.
Rogers, E.M., 2005. Diffusion of Innovations, New York: Free Press. Sekaran, U., 2006. Research menthods for business - A skills building approach 4th ed.,
Wileys Publication. Sheth, J., Newman, B. & Gross, B., 1991. Why We Buy What We Buy: A
Theory of Consumption Values. Journal of Business Research, 22(2), 159-170. Taylor, S. & Todd, P., 1995. Understanding information technology usage: A test of
competing models. Information Systems Research, 6, 144-176. TRC-SL, 2008. Statistical Overview - 2008 December, Sri Lanka: Telecommunication
Regulatory Commission of Sri Lanka. Available at: http://www.trc.gov.lk/images/pdf/stodec1.pdf [Accessed January 8, 2009].
Umino, A., 2004. DEVELOPMENT OF THIRD-GENERATION MOBILE SERVICES IN
THE OECD, Organisation for Economic Co-operation and Development. Venkatesh, V., 2000. Determinants of Perceived Ease of Use: Integrating Perceived
Behavioral Control, Computer Anxiety and Enjoyment into the Technology Acceptance Model. Information Systems Research, 11, 342-365.
Venkatesh, V. & Davis, F., 2000. A Theoretical Extension of the Technology Acceptance
Model: Four Longitudinal Field Studies. Management Science, 46(2), 186-204. Verkasalo, H., 2006. Mobile Data Service Evolution - Empirical Observations on Packet
Data Service Adoption. In Denmark.
149
Wirtz, J., Mattila, A.S. & Tan, R.L.P., 2000. The Moderating Role of Target-Arousal on the Impact of Affect on Satisfaction—An Examination in the Context of Service Experiences. Journal of Retailing, 76(3), 347-365.
Wu, C., Cheng, F. & Yen, D.C., 2008. The atmospheric factors of online storefront
environment design: An empirical experiment in Taiwan. Information & Management, 45(7), 493–498.
Wulf, K.D. et al., 2006. The role of pleasure in web site success. Information &
Management, 43, 434-446. Yani-de-Soriano, M. & Foxall, R., 2006. The emotional power of place: The fall and rise
of dominance in retail research . Journal of Retailing and Consumer Services, 13(6), 403-416.
Bandura, A., 2001. Social cognitive theory of mass communication. Media Psychology, 3, 265-298.
Heijden, V., 2004. User acceptance of hedonic information systems. MIS Quarterly,
28(4), 695-704. Leung, L. & Wei, 2000. More than just talk on the move: Uses and gratifications of the
cellular phone. J&MC Quarterly, 77, 308-320. Lucas, H. & Spitler, V., 1999. Technology use and performance: A field study of broker
workstations. Decision Sciences, 30, 291-311. Mathieson, K., 1991. Predicting user intentions: Comparing the technology acceptance
model with the theory of planned behavior. Information Systems Research, 2(3), 173-191.
Riemenschneider, C.K., Hardgrave, B.C. & Davis, F., 2002. Explaining software
developer acceptance of methodologies: A comparison of five theoretical models. IEEE Transactions on Software Engineering, 28, 1135–1145.
Taylor, S. & Todd, P., 1995. Understanding information technology usage: A test of
competing models. Information Systems Research, 6, 144-176. Tjan, A., 2001. Finally, a Way to Put Your Internet Portfolio in Order. Harvard Business
Review, February, 76-85. Venkatesh, V. & Morris, M.G., 2000. Why Don’t Men Ever Stop to Ask For Directions?
Gender, Social Influence, and Their Role in Technology Acceptance and Usage Behavior. MIS Quarterly, 24, 115-139.
Wei, R., 2008. Motivations for using the mobile phone for mass communications and
entertainment. Telematics and Informatics, 25, 36-46.
150
Sekaran. U, 2006, Research Methods For Business: A Skill Building Approach, 4th ed,New Delhi:John Wiley & Sons Kothari. C.R., 2002. Research Methodology: methods & Techniques. 2nd ed. New Delhi: New Age International Saunders, M., Lewis, P. & Thornhill, A., 2007a,"Research Onion", [Online], Available from:http://wps.pearsoned.co.uk/ema_uk_he_saunders_resmethbus_4/65/16695/4274089.cw/content/index.html, [Accessed 4th July 2009] Saunders, M., Lewis, P. & Thornhill, A., 2007b, Research Methods for Business Students, 4th ed, England:Pearson Education Limited.
151