MSc.

in TECHNOLOGY MANAGEMENT

DISSERTATION ON DEVELOPMENT OF A CONSUMER TECHNOLOGY ADOPTION MODEL FOR MOBILE DATA SERVICES WITH UTILITARIAN AND HEDONIC VALUE PROPOSITIONS IN SRI LANKA BY Ranga Perera

7TH SEPTEMBER 2009

DISSATATION

Student Name: Ranga Naresh Perera Email Address: ramga.n.perera@gmail.com Award Name: MSc in Technology Management 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

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.

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 Herbjrn 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

Table of Contents 1. Introduction..........................................................................................................................9 Problem overview.............................................................................................................11 Significance of the study...................................................................................................15 Scope and Limitations........................................................................................................18 2. Background.........................................................................................................................19 2.1 Mobile telecommunication industry overview............................................................19 2.2 Mobile Technology Evolution.....................................................................................22 2.3 Mobile Data Services...................................................................................................25 3. Literature review................................................................................................................30 3.1 Overview of the selected research area.......................................................................30 3.2 Review of literature on research subject......................................................................31 3.2.1 Motives Utility vs Hedonics......................................................................................31 3.2.2 Technology adoption models and Mobile Data Services adoption..............................33 3.3 Literature review on selected independent variables......................................................39 4. Solution...............................................................................................................................50 4.1Solution overview.........................................................................................................50 Proposed model for mobile services adoption in Sri Lanka (Sri Lanka Consumer Acceptance of Technology Model SLCAT)...................................................................51 4.2List of developed hypothesis .......................................................................................52 5. RESEARCH METHODOLOGY.......................................................................................58 5.1Research Philosophy.....................................................................................................59 5.3Research Strategy..........................................................................................................60 5.4 Pilot study.....................................................................................................................61 5.5 Time Horizon ..............................................................................................................61 5.6 Determining the Sample and Sample Size...................................................................62 5.7 Questionnaire design Likert scales used...................................................................64 5.8 Treatment of data.........................................................................................................65 6. Deliverable.........................................................................................................................66 6.1 Descriptive Analysis....................................................................................................66 6.2Statistical analysis of data.............................................................................................75 6.3Hypothesis Testing........................................................................................................78 6.4Simple liner model building .........................................................................................88 6.5Model building..............................................................................................................94 6.6Data Analysis Summary..............................................................................................103 7. Discussion.........................................................................................................................120 8. Recommendations............................................................................................................132 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......................................................................141 9 Future research..................................................................................................................142 References ..........................................................................................................................143

List of Tables Table 1: Mobile Telephony systems......................................................................................24 Table 2: Mobile Data Services classification........................................................................27 Table 3: Summary of Litreture review - Utilitarian motives.................................................48 Table 4: Summary of literature review - Hedonic motives...................................................49 Table 5: : Literature review summary - Attitude and intension............................................49 Table 6: Literature review summary - MDS with utilitarian and hedonic propositions.......49 Table 7: Hypothesis for utilitarian motives in SMS..............................................................52 Table 8: Hypothesis for hedonic motives in SMS.................................................................53 Table 9: Hypothesis of social influences in SMS..................................................................54 Table 10: Hypothesis attitude and intension in SMS.............................................................54 Table 11: Hypothesis for utilitarian motives - M-Ringtone..................................................55 Table 12: Hypothesis for hedonic motives in M-Ringtones..................................................56 Table 13: Hypothesis for Social influences - M-Ringtones..................................................56 Table 14: Hypothesis Attitude and intesion - M-Ringtone....................................................57 Table 15: Questionnaire distribution......................................................................................62 Table 16: Respondents by Gender.........................................................................................66 Table 17 : Respondents by Age.............................................................................................67 Table 18: Respondents by Province of residence..................................................................68 Table 19: Respondents by Education level............................................................................70 Table 20: Respondents by Employment status......................................................................71 Table 21: Respondents by monthly income level..................................................................72 Table 22: Colour Display vs Black/White display................................................................73 Table 23: Mobile Data Services Awareness..........................................................................74 Table 24: Test values for internal consistency SMS..........................................................76 Table 25: Test values for internal consistency - M-Ringtones..............................................77 Table 26: Correlation Matrix for SMS...................................................................................78 Table 27: Utilitarian model testing using SMS.....................................................................81 Table 28: List of accepted hypothesis (alternative) Utilitarian product.............................82 Table 29: List of Accepted Null Hypothesis.........................................................................82 Table 30: Correlation Matrix for hedonic motives................................................................83 Table 31: Hypothesis testing for Hedonic model..................................................................86 Table 32: List of accepted hypothesis Hedonic Product....................................................87 Table 33: Simple liner model building SMS......................................................................90 Table 34: Simple liner model building - Mobile Ringtones..................................................93 Table 35: Variable ranking based on correlation to Attitude towards adoption...................94

List of Figures Figure 1: World mobile subscribers...................................Error: Reference source not found Figure 2: Cellular subscriber growth rate in Sri Lanka......Error: Reference source not found Figure 3: Evolution of GSM Technologies........................Error: Reference source not found Figure 4...............................................................................Error: Reference source not found Figure 5: Proposed classification of MDS.........................Error: Reference source not found Figure 6: Techno-centric MDS classification....................Error: Reference source not found Figure 7: Four tiered MDS classification...........................Error: Reference source not found Figure 8: Classification of consumer value........................Error: Reference source not found Figure 9: Proposed model for mobile services adoption in Sri Lanka...........Error: Reference source not found Figure 10: Research Onion (Saunders et al, 2007a)...........Error: Reference source not found Figure 11: Respondents by Gender....................................Error: Reference source not found Figure 12: Respondents by Age..........................................Error: Reference source not found Figure 13: Respondents by Province of residence.............Error: Reference source not found Figure 14: Respondents by Education level.......................Error: Reference source not found Figure 15: Respondents by Employment status.................Error: Reference source not found Figure 16: Respondents by monthly income level.............Error: Reference source not found Figure 17: Colour Display vs Black/White display Error: Reference source not found

Abbreviations

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; Mlarstig 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

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.

10

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 (Mlarstig 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

11

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 12

highlight the critical need for creating new value and new behavior patterns 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.

13

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.

14

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

15

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.

16

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.

17

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

18

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 observation in this analysis is that users of fixed phone lines are spending an increased number of minutes 19

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.

20

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. While Sri Lanka 21

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 Time Division Multiple

22

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.

23

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).

24

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.

25

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

26

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)

27

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) 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

28

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)

29

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

30

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. Fernndez & 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. Fernndez & 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 research. It is indeed surprising to

31

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 Fernndez & 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 Fernndez & Bonillo (2007) as multi-dimensional, identifies five key values that 32

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. Fernndez & 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 of use as the degree to which a person believes that using a particular system would be free of effort. The

33

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 behavioral control to

34

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

35

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. Bina et

36

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 Russells (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.

37

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.

38

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.

39

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 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.

40

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 subclassifications 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.

41

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 variable of arousal and its influence and interplay in the consumers buying decision, a more unique approach to

42

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.

43

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 individuals 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. Lpez-Nicols 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.

44

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 presence of

45

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 model. 46

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.

47

Summary of Literature Review 3.3.11 Utilitarian Motives

Variable Perceived Usefulness Literature 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) Perceived Ease of use Hong et al. (2006) 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) Relative Advantage Hong et al. (2006) Rogers (2005) Kulviwat et al. (2007) Plouffe et al. (2001) Moore & Benbasat (1991)
Table 3: Summary of Litreture review - Utilitarian motives

Finding Positive Relationship

Positive Relationship

Positive Relationship

3.3.12 Hedonic Motives

Variable Literature 48 Finding

Pleasure

Kulviwat et al. (2007) Lee et al. (2003) Mehrabian & Russell (1974) Wu et al. (2008) Wulf et al. (2006) Kulviwat et al. (2007) Wu et al. (2008) Lee et al. (2003) Wirtz et al. (2000) Nasco et al. (2008) Yani-de-Soriano & Foxall (2006)

Positive Relationship

Arousal

Positive Relationship

Dominance

Positive Relationship

Table 4: Summary of literature review - Hedonic motives

Attitude and Intension Variable Attitude and Intension

Literature Kulviwat et al. (2007) Bruner II & Kumar (2005)

Finding Positive relationship

Table 5: : Literature review summary - Attitude and intension

Mobile data services with Utilitarian and Hedonic value propositions Mobile Data Service Short Message Service Mobile Ringtone Literature Finding Cognitive utilitarian value proposition Hedonic proposition
Table 6: Literature review summary - MDS with utilitarian and hedonic propositions

value

49

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.

50

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

51

4.2

List of developed hypothesis

Model for analyzing Utilitarian value proposition Using SMS 4.2.1 No 1. Utilitarian motives in the adoption of Utilitarian Services (SMS)

Variable tested
the

Hypothesis Perceived usefulness in Hypothesis 1

context of Utilitarian Services

H01: There is weak influence of perceived usefulness in the context of attitude towards
adoption of utilitarian Mobile Data Services in Sri Lanka

adoption and attitude towards adoption

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

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

adoption and attitude towards adoption

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

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

adoption and attitude towards adoption

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

52

4.2.2 No 4.

Hedonic motives in the adoption of Utilitarian Services (SMS)

Variable tested

Hypothesis Pleasure in the context Hypothesis 4

of Utilitarian Services adoption and attitude towards adoption

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

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

context of Utilitarian Services adoption and attitude adoption towards

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

53

4.2.3 No 7.

Role of Social Influences in the context of Utilitarian value proposition

Variable tested

Hypothesis Social influence in the Hypothesis 7

context of Utilitarian Services adoption and attitude adoption towards

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 8.

Variable tested
Relationship attitude

Hypothesis between Hypothesis 8

towards

adoption of Utilitarian service and Adoption Intension

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

54

4.2.5 No 9.

Model for analyzing Hedonic value proposition Using Mobile Ringtone

Variable tested

Hypothesis Perceived usefulness in Hypothesis 9

the context of Hedonic Services adoption and attitude adoption towards

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 . in

Perceived Ease of use the context of Hedonic Services

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

adoption and attitude towards adoption

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 adoption towards

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 No

Hedonic motives in the adoption of Hedonic Services (Mobile Ringtone)

Hypothesis 12 Pleasure in the context Hypothesis 12 . of Hedonic Services

adoption and attitude towards adoption

Variable tested

H012: There is weak influence of pleasure in the context of attitude towards adoption of
adopt hedonic Mobile Data Services in Sri Lanka

Ha12: There is strong influence of pleasure in the context of attitude towards adoption
of hedonic Mobile Data Services in Sri Lanka

55

No

Hypothesis 13 Arousal in the context Hypothesis 13 . of Hedonic Services

adoption and attitude towards adoption

Variable tested

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 . context of
attitude adoption

in

the

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

Hedonic towards

Services adoption and

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 No

Role of Social Influences in the context of Hedonic value proposition

Hypothesis 15 Social influence in the Hypothesis 15 . context of Hedonic

Services adoption and attitude adoption towards

Variable tested

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

56

4.2.8

Relationship between Attitude towards adoption and adoption intension in the context of hedonic value proposition

No

Variable tested 16 Relationship . attitude

adoption Intension of

Hypothesis between Hypothesis 16

towards hedonic

H016: There is weak influence of attitude towards adoption and adoption intension in
the context hedonic Mobile Data Services in Sri Lanka

service and Adoption

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

57

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 its 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.
Slide 4. 1

Figure 4.1

The research onion

Saunders, Lewis and Thornhill, Research Methods in Business , 4th Edition Pearson Education Limited 2007

Source : Mark Saunders, Philip Lewis and Adrian Thornhill 2006

Figure 10: Research Onion (Saunders et al, 2007a)

58

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.

5.1Research 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.

5.2Research 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.

59

5.3Research 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. 60

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, participants biasness, observer error and observers 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 due to the

61

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 7 3 7 3 3 3 3 29 Actual responses received by the deadline and used for the research 230 25 160 40 Actual responses received after the deadline and not used for the research 50 15 25 15 25 20 20 170

Western Southern Central North Western North Central Sabaragamuwa Uva

350 100 250 100 100 100 100 1100

455

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 was

62

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)

63

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); Lpez-Nicols & 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 questionnaire respondents that the states of emotion differ based on the circumstance. It was based on this input the word in general was included. 64

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.

65

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 Cumulative Frequency Female Male Total 225 184 409 Percent 55.0 45.0 100.0 Valid Percent 55.0 45.0 100.0 Percent 55.0 100.0

Table 16: Respondents by Gender

Figure 11: Respondents by Gender

66

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 Cumulative Frequency No comment 18-25 26-35 36-45 46-55 56-65 Total 12 180 163 45 7 2 409 Percent 2.9 44.0 39.9 11.0 1.7 .5 100.0 Valid Percent 2.9 44.0 39.9 11.0 1.7 .5 100.0 Percent 2.9 46.9 86.8 97.8 99.5 100.0

Table 17 : Respondents by Age

Figure 12: Respondents by Age

67

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 Cumulative Frequency Percent Valid Percent Percent

Western Province North Western Province Central Province Southern Province 0ther Total

207 30 142 18 12 409

50.6 7.3 34.7 4.4 2.9 100.0

50.6 7.3 34.7 4.4 2.9 100.0

53.5 60.9 95.6 100.0 2.9

Table 18: Respondents by Province of residence

Figure 13: Respondents by Province of residence

68

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.

69

6.1.5 Respondents by Education level

Education Level Cumulative Frequency Percent Valid Percent Percent

GCE Ordinary Level GCE Advance Level Diploma Level Degree and Above qualification Not responded Total

66 173 37 132 1 409

16.1 42.3 9.0 32.3 .2 100.0

16.1 42.3 9.0 32.3 .2 100.0

16.4 58.7 67.7 100.0 .2

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.

70

6.1.6 Respondents by Employment status

Current Employment Status Cumulative Frequency Not responded Employed Unemployed Total 1 277 131 409 Percent .2 67.7 32.0 100.0 Valid Percent .2 67.7 32.0 100.0 Percent .2 68.0 100.0

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.

71

6.1.7 Respondents by monthly income level

Monthly income Level Cumulative Frequency Percent Valid Percent Percent

Less than Rs.5000 More than Rs5000 and Less than Rs.10,000 More than Rs. 10,000 and Less than Rs.30,000 More than Rs.30,000 and Less than Rs. 50,000 More than Rs.50000 Not responded Total

62 73

15.2 17.8

15.2 17.8

19.8 37.7

211

51.6

51.6

89.2

35 9 19 409

8.6 2.2 4.6 100.0

8.6 2.2 4.6 100.0

97.8 100.0 4.6

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.

72

Colour Display vs Black/White display Cumulative Frequency Valid Black and White Display Color Display Not responded Total 283 108 18 409 69.2 26.4 4.4 100.0 69.2 26.4 4.4 100.0 73.6 100.0 4.4 Percent Valid Percent Percent

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.

73

6.1.8 Mobile Data Services Awareness

Frequency Not No aware comment Percentage Not No aware comment

No 1 2 3 4 5 6 7 8 9 10 11 12

Mobile Data Service

Aware

Aware

SMS MMS Local language messaging Mobile E-mail Ringtones Download Icons, Logos and wallpapers Mobile Games Listen to music/radio Ask/ Send credit Book movie tickets through mobile Mobile Banking Mobile internet
Table 23: Mobile Data Services Awareness

395 315 353 345 365 336 361 343 312 226 316 335

54 18 24 4 29 10 12 47 135 48 29

14 40 38 40 40 44 38 54 50 48 45 45

96.6 77 86.3 84.4 89.2 82.2 88.3 83.9 76.3 55.3 77.3 81.9

13.2 4.4 5.9 1 7.1 2.4 2.9 11.5 33 11.7 7.1

3.4 9.8 9.3 9.8 9.8 10.8 9.3 13.2 12.2 11.7 11 11

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.

74

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 Cronbachs 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 Cronbachs 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 Cronbachs 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.

75

Summary table on internal consistency and validity testing

6.2.1 Utilitarian model testing using SMS

No 1 . 2 . 3 . 4 . 5 . 6 . 7 . 8 . 9 . Variable Perceived Usefulness Perceived Ease of use Comparative advantage Social Influences Pleasure Arousal Dominance Attitude Towards Adoption Adoption Intension Cronbachs alpha values 0.919 0.960 0.848 0.836 0.668 0.71

0.859 0.534 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 1 Variable Perceived Usefulness 76 Cronbachs alpha values 0.989

2 . 3 . 4 . 5 . 6 . 7 . 8 . 9 .

Perceived Ease of use Comparative advantage Social Influences Pleasure Arousal Dominance Attitude Towards Adoption Adoption Intension

0.988 0.934 0.984 0.954 0.934 0.923 0.808 0.342

Table 25: Test values for internal consistency - M-Ringtones

Detailed statistics are available in Annex 2.

77

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 Perceived Usefulness (PU) Perceived Ease of Use (EOU) Relative Advantage (RA) Social Influences(SI) Pleasure(PL) Arousal(AR) Dominance(DO) Attitude toward adoption(ATA) Adoption Intension(AI) PU 1 0.81 0.82 0.40 0.41 0.35 0.25 0.79 0.45 1 0.83 0.45 0.37 0.30 0.13 0.74 0.49* 1 0.33 0.33 0.28 0.17 0.71 0.41 1 0.27 0.21 0.15 0.41 0.17 1 0.68 0.30 0.52 0.14 1 0.34 0.53 0.21 EOU RA SI PL AR DO ATA AI

1 0.30 0.57 1 0.38 1

Table 26: Correlation Matrix for SMS

78

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 Reject null 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 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 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 ACCEPT ALTERNATIVE HYPOTHESIS p value less than 0.05 Reject null hypothesis 0.739 ACCEPT ALTERNATIVE HYPOTHESIS p value less than 0.05 Reject null hypothesis 0.785 p value less than 0.05

2.

3.

ACCEPT ALTERNATIVE HYPOTHESIS

79

No

Hypothesis

Person correlation (two tailed test) result 0.407

Tested Significance level

Status of acceptance or rejection of hypothesis at 95% confidence level ACCEPT NULL HYPOTHESIS

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 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 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 ACCEPT ALTERNATIVE HYPOTHESIS 0.527 p value less than 0.05 Reject null hypothesis 0.524 p value less than 0.05 Reject null hypothesis Reject alternative hypothesis p value less than 0.05

5.

ACCEPT ALTERNATIVE HYPOTHESIS

6.

80

No

Hypothesis

Person correlation (two tailed test) result 0.303

Tested Significance level

Status of acceptance or rejection of hypothesis at 95% confidence level ACCEPT NULL HYPOTHESIS

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 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 27: Utilitarian model testing using SMS

p value less than 0.05

Reject alternative hypothesis

8.

0.383

p value less than 0.05

ACCEPT NULL HYPOTHESIS

Reject alternative hypothesis

81

List of accepted hypothesis (alternative) Utilitarian product No Hypothesis 1 Hypothesis 2 Hypothesis 3 Hypothesis 5 Hypothesis 6 Hypothesis There is strong influence of perceived usefulness in the context of attitude towards adoption of utilitarian Mobile Data Services in Sri Lanka There is strong influence of perceived ease of use in the context of attitude towards adoption of utilitarian Mobile Data Services in Sri Lanka There is strong influence of relative advantage in the context of attitude towards adoption of utilitarian Mobile Data Services in Sri Lanka There is strong influence of pleasure in the context of attitude towards adoption of utilitarian Mobile Data Services in Sri Lanka 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 4 Hypothesis 7 Hypothesis 8 Hypothesis There is weak influence of social influences in the context of attitude towards adoption of utilitarian Mobile Data Services in Sri Lanka There is weak influence of dominance in the context of attitude towards adoption of utilitarian Mobile Data Services in Sri Lanka 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

82

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 Perceived Usefulness (PU) Perceived Ease of Use (EOU) Relative Advantage (RA) Social Influences(SI) Pleasure(PL) Arousal(AR) Dominance(DO) Attitude toward adoption(ATA) Adoption Intension(AI) PU 1.00 0.96 0.94 0.96 0.90 0.78 0.51 0.92 0.69 1.00 0.97 0.98 0.92 0.74 0.51 0.93 0.71 1.00 0.97 0.94 0.75 0.51 0.97 0.72 1.00 0.92 0.79 0.51 0.92 0.72 1.00 0.72 0.48 0.94 0.68 1.00 0.49 0.74 0.59 EOU RA SI PL AR DO ATA AI

1.00 0.54 0.42 1.00 0.67 1.00

Table 30: Correlation Matrix for hedonic motives

83

Hypothesis testing for Hedonic model

No Hypothesis Person correlation (two tailed test) result 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 2. Mobile Data Services in Sri Lanka 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 3. hedonic Mobile Data Services in Sri Lanka 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 No Mobile Data Services in Sri Lanka Hypothesis Person correlation (two Tested Significance ACCEPT ALTERNATIVE HYPOTHESIS Status of acceptance or 0.965 p value less than 0.05 ACCEPT ALTERNATIVE HYPOTHESIS Reject null hypothesis 0.930 p value less than 0.05 ACCEPT ALTERNATIVE HYPOTHESIS Reject null hypothesis 0.922 p value less than 0.05 Reject null hypothesis Tested Significance level Status of acceptance or rejection of hypothesis at 95% confidence level

84

No

Hypothesis

Person correlation (two tailed test) result tailed test) result

Tested Significance level

Status of acceptance or rejection of hypothesis at 95% confidence level rejection of hypothesis at 95% confidence level

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 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 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 ACCEPT ALTERNATIVE HYPOTHESIS ACCEPT ALTERNATIVE HYPOTHESIS p value less than 0.05 Reject null hypothesis 0.941 p value less than 0.05 ACCEPT ALTERNATIVE HYPOTHESIS Reject null hypothesis 0.917 p value less than 0.05 Reject null hypothesis

5.

6.

0.742

85

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 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 31: Hypothesis testing for Hedonic model

0.538

p value less than 0.05

Reject null hypothesis

ACCEPT ALTERNATIVE HYPOTHESIS 0.672 p value less than 0.05 Reject null hypothesis

8.

ACCEPT ALTERNATIVE HYPOTHESIS

86

List of accepted hypothesis Hedonic Product No Hypothesis 9 Hypothesis 10 Hypothesis 11 Hypothesis 12 Hypothesis 13 Hypothesis 14 Hypothesis 15 Hypothesis 16 Hypothesis There is strong influence of perceived usefulness in the context of attitude towards adoption of hedonic Mobile Data Services in Sri Lanka There is strong influence of perceived ease of use in the context of attitude towards adoption of hedonic Mobile Data Services in Sri Lanka There is strong influence of perceived ease of use in the context of attitude towards adoption of hedonic Mobile Data Services in Sri Lanka There is a strong influence of social influence in the context of attitude towards adoption of hedonic Mobile Data Services in Sri Lanka There is strong influence of pleasure in the context of attitude towards adoption of hedonic Mobile Data Services in Sri Lanka There is strong influence of arousal in the context of attitude towards adoption of hedonic Mobile Data Services in Sri Lanka There is strong influence of dominance in the context of attitude towards adoption of hedonic Mobile Data Services in Sri Lanka 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

87

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 1. Perceived usefulness and attitude towards adoption value 61.5% ANOVA F value test and acceptability at 95% confidence level Accept or Reject status 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 2. Perceived ease of use and attitude towards adoption 54.5 % Accept y = .133 + .305 X1 observed Model Acceptable. 45 degree upward sloping plot observed and Randomly scattered residuals an even horizontal band around residual value of zero 3. Comparative advantage and attitude towards adoption 50.4% Accept y = .266 + .343 X1 observed Model Acceptable. 45 degree upward sloping plot observed and Randomly scattered residuals an even horizontal band around residual value of zero Tables 9-12 and Figures 5-6 Tables 5-8 and Figures 3-4 Tables 1-4 and Figures 1-2 Simple Liner regression equation Comments on residual value analysis Reference in Annex 6

88

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 Accept

Simple Liner regression equation

Comments on residual value analysis

Reference in Annex 6

4.

Social influences and attitude towards adoption

16.3%

y = .749 + .201 X1

observed Reject Model. Sparse and broken 45 degree upward sloping plot observed. Low random nature of the scattered instances shown in the plot Reject Model. Sparse and broken 45 degree upward sloping plot observed. Low random nature of the scattered instances shown in the plot Reject Model. Sparse and broken 45 degree upward sloping plot observed. Low random nature of the scattered instances shown in the plot Reject Model. Sparse and broken 45 degree upward sloping plot observed. Low random nature of the scattered instances shown

Tables 13-15 and Figures 7-8

5.

Pleasure and attitude towards adoption

27.3%

Accept

y = .933 + .653 X1

Tables 16-18 and Figures 9-10

6.

Arousal and attitude towards adoption

27.6%

Accept

y = 1.004 + .611 X1

Tables 19-21 and Figures 11-12

7.

Dominance and attitude towards adoption

8.9%

Accept

y = 1.088 + .282 X1

Tables 22-24 and Figures 13-14

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 Accept

Simple Liner regression equation

Comments on residual value analysis

Reference in Annex 6

8.

Attitude towards adoption and Adoption Intension

14.4%

y = 3.923 + .569 X1

in the plot 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.

90

Variables in Model tested for Hedonic value proposition (Mobile Ringtone)

No Variables Model predictabilit y percentage (%) based on the Adjusted R squared 1. Perceived usefulness and attitude towards adoption value 85% ANOVA F value test and acceptability at 95% confidence level Accept or Reject status 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 2. Perceived ease of use and attitude towards adoption 86.5% Accept y = .029+ .494X1 observed Model Acceptable. 45 degree upward sloping plot observed and Randomly scattered residuals an even horizontal band around residual value of zero 3. Comparative advantage and attitude towards adoption 93.1% Accept y = .005+ .564X1 observed Model Acceptable. 45 degree upward sloping plot observed and Randomly scattered residuals an even horizontal band around residual value of zero 4. Social influences and 84% Accept y = .041+ .489X1 observed Model Acceptable. 45 degree upward Tables 37-39 and Figures Tables 34-36 and Figures 21-22 Tables 31-33 and Figures 19-20 Tables 28-30 and Figures 17-18 Simple Liner regression equation Comments on residual value analysis Reference in Annex 6

91

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

attitude towards adoption

sloping plot observed and Randomly scattered residuals an even horizontal band around residual value of zero observed Model Acceptable. 45 degree upward sloping plot observed and Randomly scattered residuals an even horizontal band around residual value of zero observed Model Acceptable. 45 degree upward sloping plot observed and Randomly scattered residuals an even horizontal band around residual value of zero observed

23-24

5.

Pleasure and attitude towards adoption

88.5%

Accept

y = .044+ .684X1

Tables 40-42 and Figures 25-26

6.

Arousal and attitude towards adoption

54.9%

Accept

y = .331+ 1.226X1

Tables 43-45 and Figures 27-28

7.

Dominance and attitude towards adoption

28.8%

Accept

y = .578+ 1.162X1

Model Acceptable. 45 degree upward sloping plot observed and Randomly scattered

Tables 46-48 and Figures 29-30

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

residuals an even horizontal band around residual value of zero 8. Attitude towards adoption and Adoption Intension 45% Accept
y = 2.953 + .726 X1

observed 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.

93

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) 0.79 R squared value Ranking 61.50% 1 0.74 54.50% 2 0.71 50.40% 3 0.53 27.60% 4 0.52 27.30% 5

1.

Perceived Usefulness and attitude towards adoption in the context of an utilitarian product Perceived ease of use and attitude towards adoption in the context of an utilitarian product relative advantage and attitude towards adoption in the context of an utilitarian product Arousal and attitude towards adoption in the context of an utilitarian product Pleasure and attitude towards adoption in the context of an utilitarian product

2.

3.

4.

5.

Table 35: Variable ranking based on correlation to Attitude towards adoption

94

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) Liner Model .849 0.000 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 Model predictability (Adjusted R squared) ANOVA Acceptability of model based on Residual analysis Normal probability plot (Residual) Scatter plot of the standardized residuals vs standard fitted (Residual) Reference for detailed test data Pleasure recorded p values greater than 0.05. 71.7% F value was noted at 207.720 indicating the significance is under 95% confidence. Model Accepted. ACCEPTED 45 degree upward sloping plot observed Randomly scattered in an even horizontal band around residual value of zero ANNEX 06, 14-16 pages

Test Model 1 : Option 2

95

Model predictability (Adjusted R squared) Liner Model

64.4%
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

ANOVA Pearson Correlation Significance (Two tailed) (N=409) Acceptability of model based on Residual analysis Normal probability plot (Residual) Scatter plot of the standardized residuals vs standard fitted (Residual)
Reference for detailed test data

advantage recorded p values greater than 0.05. F value was noted at 247.463 indicating the significance is under 95% confidence. Model Accepted.
.804 .000

ACCEPTED 45 degree upward sloping plot observed Randomly scattered in an even horizontal band around residual value of zero
ANNEX 06, 17-19 pages

96

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

97

Test Model 1 : Option 1 Model predictability (Adjusted R squared) Multiple regression Model

40.0 % 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. F value was noted at 89.323 indicating the significance is under 95% confidence. Accepted .631 .000 Accepted ACCEPT MODEL ANNEX 06, 2-7 pages

ANOVA Pearson Correlation Significance (Two tailed) (N=409) Acceptability of model based on Residual analysis DECSION Reference for detailed test data

98

Test Model 1 : Option 2 Model predictability (Adjusted R squared) Multiple regression Model

24.2%
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 Pearson Correlation Significance (Two tailed) (N=409) Acceptability of model based on Residual analysis
DECSION Reference for detailed test data

The F value is 22.680. This is significant under 95 % confidence (p

value is less than 0.05) .503 .000

Rejected
REJECT MODEL 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

99

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.

100

Test Model 2 : Option 1 Model predictability (Adjusted R squared) Liner Model

54.2 % 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 Pearson Correlation Significance (Two tailed) (N=409) Acceptability of model based on Residual analysis Normal probability plot (Residual) Scatter plot of the standardized residuals vs standard fitted (Residual) Reference for detailed test data

The F value is 61.471. This is significant under 95 % confidence (p

value is less than 0.05)

0.743 0.000 ACCEPTED 45 degree upward sloping plot observed Randomly scattered in an even horizontal band around residual value of zero ANNEX 06, 8-11 pages

101

Test Model 2 : Option 2 Model predictability (Adjusted R squared) Liner Model

49.1 %
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 Pearson Correlation Significance (Two tailed) (N=409) Acceptability of model based on Residual analysis Normal probability plot (Residual) Scatter plot of the standardized residuals vs standard fitted (Residual)
Reference for detailed test data

The F value is 99.335. This is significant under 95 % confidence (p

value is less than 0.05) .743 0.000

ACCEPTED 45 degree upward sloping plot observed Randomly scattered in an even horizontal band around residual value of zero
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.

102

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 Cronbachs 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 test. The Normal probability plot indicated a 45 degree

103

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 Cronbachs 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 test. The Normal probability plot indicated a 45 degree sloping plot while the scatter plot indicated

104

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 Cronbachs 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 95% confidence level. The model was accepted based on this test. Residual analysis of the

105

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 Cronbachs 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 adoption and X being the Social influences). Two tests were conducted prior to the acceptance of the simple linear

106

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 Cronbachs 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 Cronbachs 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 Correlation test between the variable and Attitude towards adoption indicated a correlation of 0.524 with p value

107

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 Cronbachs alpha which was measured at 0.71(table 24). Nunnally & Bernstein (1994) (as cited by Kulviwat

108

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 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

109

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 Cronbachs 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 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

110

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 Cronbachs 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.

111

0.527

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 Cronbachs 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 relationships the correlation 112

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 Cronbachs 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. This strength of correlation was above that between the Attitude towards adoption and Intension to adopt.

113

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 Cronbachs 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.

114

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 Cronbachs 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.

115

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 Cronbachs 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 Cronbachs alpha which was measured at 0.93. This was noted to be a good fit. The two

116

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 Cronbachs 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.

117

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 Cronbachs 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. (y being the Intension to adopt and X being the Attitude towards adoption). Two tests were conducted prior to the acceptance

118

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.

119

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 Hypothesis 1 Perceived Usefulness and attitude test) 0.79 61.5% R squared value Relevant Literature

Pedersen et al. (2002) Nysveen et al. (2005) Kulviwat et al. (2007) Bruner II & Kumar (2005) Kim et al. (2009) Pedersen et al. (2002) Nysveen et al. (2005) Kulviwat et al. (2007) Bruner II & Kumar (2005) Kim et al. (2009) Kulviwat et al. (2007) Rogers (2005)

towards adoption in the context of an utilitarian product

Hypothesis 2

Perceived ease of use and attitude towards adoption in the context of an utilitarian product

0.74

54.5 %

Hypothesis 3 Hypothesis 5 Hypothesis 6

Relative advantage and attitude towards adoption in the context of an utilitarian product Pleasure and attitude towards adoption in the context of an utilitarian product Arousal and attitude towards adoption in the context of an utilitarian product

0.71

50.4%

0.52 0.53

27.3% 27.6%

Kulviwat et al. (2007) Wu et al. (2008) Kulviwat et al. (2007) Wu et al. (2008)

120

Hedonic value proposition and variables accepted

No Measured Variable Correlation (at 95% significance two tailed Hypothesis 9 Perceived Usefulness and test) 0.92 85% R squared value Relevant Literature

Pedersen et al. (2002) Nysveen et al. (2005a) Kulviwat et al. (2007)

attitude towards adoption in the context of an hedonic product Perceived ease of use and attitude towards adoption in the context of an hedonic product relative advantage and attitude towards adoption in the context of hedonic product social influences and attitude towards adoption in the context of an hedonic product Pleasure and attitude towards adoption in the context of an hedonic product Arousal and attitude towards adoption in the context of an hedonic product Dominance and attitude towards adoption in the context of an hedonic product 0.54 28.8% 0.74 54.9% 0.94 88.5% 0.92 84%

Hypothesis 10

0.93

86.5%

Pedersen et al. (2002) Nysveen et al. (2005a) Kulviwat et al. (2007)

Hypothesis 11 Hypothesis 12 Hypothesis 13 Hypothesis 14 Hypothesis 15

0.97

93.1%

Kulviwat et al. (2007)

Kulviwat et al. (2007) Kulviwat et al. (2008) Kulviwat et al. (2007)

Kulviwat et al. (2007) Wu et al. (2008) Nasco et al. (2008)

121

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. 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

122

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. 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

123

and reliability of the variables as tested using Cronbachs 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 Cronbachs 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 SLCAT 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 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

124

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 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

125

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. (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

126

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 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,

127

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.

128

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.

129

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 leaders perception towards the adoption of this technology may

130

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.

131

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 Hypothesis 1 Perceived Usefulness and attitude test) 0.79 61.5% R squared value Relevant Literature

Pedersen et al. (2002) Nysveen et al. (2005) Kulviwat et al. (2007) Bruner II & Kumar (2005) Kim et al. (2009) Pedersen et al. (2002) Nysveen et al. (2005) Kulviwat et al. (2007) Bruner II & Kumar (2005) Kim et al. (2009) Kulviwat et al. (2007) Rogers (2005)

towards adoption in the context of an utilitarian product

Hypothesis 2

Perceived ease of use and attitude towards adoption in the context of an utilitarian product

0.74

54.5 %

Hypothesis 3 Hypothesis 5 Hypothesis 6

Relative advantage and attitude towards adoption in the context of an utilitarian product Pleasure and attitude towards adoption in the context of an utilitarian product Arousal and attitude towards adoption in the context of an utilitarian product

0.71

50.4%

0.52 0.53

27.3% 27.6%

Kulviwat et al. (2007) Wu et al. (2008) Kulviwat et al. (2007) Wu et al. (2008)

132

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 Hypothesis 9 Perceived Usefulness and test) 0.92 85% R squared value Relevant Literature

Pedersen et al. (2002) Nysveen et al. (2005a) Kulviwat et al. (2007)

attitude towards adoption in the context of an hedonic product Perceived ease of use and attitude towards adoption in the context of an hedonic product relative advantage and attitude towards adoption in the context of hedonic product social influences and attitude towards adoption in the context of an hedonic product Pleasure and attitude towards adoption in the context of an hedonic product Arousal and attitude towards adoption in the context of an hedonic product Dominance and attitude towards adoption in the context of an hedonic product 0.54 28.8% 0.74 54.9% 0.94 88.5% 0.92 84%

Hypothesis 10

0.93

86.5%

Pedersen et al. (2002) Nysveen et al. (2005a) Kulviwat et al. (2007)

Hypothesis 11 Hypothesis 12 Hypothesis 13 Hypothesis 14 Hypothesis 15

0.97

93.1%

Kulviwat et al. (2007)

Kulviwat et al. (2007) Kulviwat et al. (2008) Kulviwat et al. (2007)

Kulviwat et al. (2007) Wu et al. (2008) Nasco et al. (2008)

133

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.

134

SL-CAT model 1-correlation diagram of Utilitarian Product of SMS adoption model

0.815 0.821 0.832

0.785

0.451 0.407
0.567

0.567

0.408 0.711

0.524 0.527 0.303

0.383

135

SL-CAT Model 2 - Correlation diagram of Hedonic Product of Mobile Ringtone adoption model

0.538 0.941 0.922 0.714 0.930 0.965 0.627 0.917 0.722 0.941 0.742 0.721 0.421

0.725 0.685 0.595

0.491

136

Proposed staircase model for adoption of Mobile Data Services

137

8.3 Prediction Model 1 Utilitarian Product of SMS adoption intension

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

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 at 95% confidence level, individual variable of comparative advantage and pleasure recorded individual p> 0.05. 138

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

139

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

140

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.

141

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.

142

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-researcheconomic-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), 644656. 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, 553558. 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, 699719. Childers, T.L. et al., 2001. Hedonic and utilitarian motivations for online retail shopping behavior. Journal of Retailing, 77(4), 511-535. 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.

143

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. Fernndez, R.S. & Bonillo, M..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, 309323. 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. 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), 18191834.

144

Hu, P. et al., 1999. Examining the technology acceptance model using physician acceptance of telemedicine technology. Journal of Management Information Systems, 16, 91113. 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/ITUD/ict/newslog/ITU+Expects+Global+Mobile+Penetration+Rate+To+Reach+50+Pe rcent+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), 706712. Kulviwat, S. et al., 2007. Toward a Unified Theory of Consumer Acceptance Technology. Psychology & Marketing, 24(12), 10591084. 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. 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), 254265.

145

Lpez-Nicols, 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. Mlarstig, Y., Holmstrm, O. & Davidsson, P., 2007. The Swedish Telecommunications Market 2007, Stockholm, Sweden: National Post and Telecom Agency. Marketing Sciences Institute, 2006. Research Priorities (20062008) 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, 192222. 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: McGrawHill. Nysveen, H., Pedersen, P.E. & Thorbjrnsen, 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. & Thorbjrnsen, H., 2005b. Intentions to Use Mobile Services: Antecedents and Cross-Service Comparisons. Journal of the Academy of Marketing Science, 33(3), 330-346. 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.

146

Pedersen, P.E., Nysveen, H. & Thorbjrnsen, 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, 208222. 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. Wirtz, J., Mattila, A.S. & Tan, R.L.P., 2000. The Moderating Role of Target-Arousal on the Impact of Affect on SatisfactionAn Examination in the Context of Service Experiences. Journal of Retailing, 76(3), 347-365.

147

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), 493498. 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, 11351145. 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 Dont 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. Sekaran. U, 2006, Research Methods For Business: A Skill Building Approach, 4th ed,New Delhi:John Wiley & Sons

148

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.c w/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.

149