adoption of mobile internet services: an exploratory study of mobile commerce early adopters

21
This article was downloaded by: [New York University] On: 06 October 2014, At: 00:56 Publisher: Taylor & Francis Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK Journal of Organizational Computing and Electronic Commerce Publication details, including instructions for authors and subscription information: http://www.tandfonline.com/loi/hoce20 Adoption of Mobile Internet Services: An Exploratory Study of Mobile Commerce Early Adopters Per E. Pedersen Published online: 18 Nov 2009. To cite this article: Per E. Pedersen (2005) Adoption of Mobile Internet Services: An Exploratory Study of Mobile Commerce Early Adopters, Journal of Organizational Computing and Electronic Commerce, 15:3, 203-222, DOI: 10.1207/ s15327744joce1503_2 To link to this article: http://dx.doi.org/10.1207/s15327744joce1503_2 PLEASE SCROLL DOWN FOR ARTICLE Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”) contained in the publications on our platform. However, Taylor & Francis, our agents, and our licensors make no representations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose of the Content. Any opinions and views expressed in this publication are the opinions and views of the authors, and are not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be relied upon and should be independently verified with primary sources of information. Taylor and Francis shall not be liable for any losses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilities whatsoever or howsoever caused arising directly or indirectly in connection with, in relation to or arising out of the use of the Content. This article may be used for research, teaching, and private study purposes. Any substantial or systematic reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any form to anyone is expressly forbidden. Terms & Conditions of access and use can be found at http:// www.tandfonline.com/page/terms-and-conditions

Upload: per-e

Post on 19-Feb-2017

213 views

Category:

Documents


1 download

TRANSCRIPT

Page 1: Adoption of Mobile Internet Services: An Exploratory Study of Mobile Commerce Early Adopters

This article was downloaded by: [New York University]On: 06 October 2014, At: 00:56Publisher: Taylor & FrancisInforma Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House,37-41 Mortimer Street, London W1T 3JH, UK

Journal of Organizational Computing and ElectronicCommercePublication details, including instructions for authors and subscription information:http://www.tandfonline.com/loi/hoce20

Adoption of Mobile Internet Services: An ExploratoryStudy of Mobile Commerce Early AdoptersPer E. PedersenPublished online: 18 Nov 2009.

To cite this article: Per E. Pedersen (2005) Adoption of Mobile Internet Services: An Exploratory Study of MobileCommerce Early Adopters, Journal of Organizational Computing and Electronic Commerce, 15:3, 203-222, DOI: 10.1207/s15327744joce1503_2

To link to this article: http://dx.doi.org/10.1207/s15327744joce1503_2

PLEASE SCROLL DOWN FOR ARTICLE

Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”) containedin the publications on our platform. However, Taylor & Francis, our agents, and our licensors make norepresentations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose of theContent. Any opinions and views expressed in this publication are the opinions and views of the authors, andare not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be relied upon andshould be independently verified with primary sources of information. Taylor and Francis shall not be liable forany losses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilities whatsoeveror howsoever caused arising directly or indirectly in connection with, in relation to or arising out of the use ofthe Content.

This article may be used for research, teaching, and private study purposes. Any substantial or systematicreproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in anyform to anyone is expressly forbidden. Terms & Conditions of access and use can be found at http://www.tandfonline.com/page/terms-and-conditions

Page 2: Adoption of Mobile Internet Services: An Exploratory Study of Mobile Commerce Early Adopters

Adoption of Mobile Internet Services:An Exploratory Study of Mobile

Commerce Early Adopters

Per E. PedersenDepartment of ICT

Agder University College

Even though the literature on the adoption and use of mobile services is quite exten-sive, surprisingly few studies have been found that have applied traditional models ofinformation and communication technology (ICT) adoption such as the technology ac-ceptance model. This suggests different perspectives have been applied in studies ofmobile ICT adoption and traditional ICT adoption. With the introduction of third gen-eration mobile network services, a convergence of mobile services and traditionalInternet services is expected. Thus, traditional models of ICT adoption may be applied,improving one’s understanding of the adoption of these services. However, studies ofthe use and adoption of mobile services have indicated that traditional adoption mod-els need to be extended and modified when applied to mobile services. In this study,we apply a modified version of the decomposed theory of planned behavior to theadoption behavior of early adopters of mobile commerce services. The study showsthat the extended and modified model has good fit to the early adopter data and that itexplains 49% of the early adopters’ intentions to use mobile commerce services. Themodel may be used as a basis for industry players’ evaluation of the adoption potentialof new mobile services.

mobile commerce, mobile Internet services, adoption,theory of planned behavior, technology acceptance

1. INTRODUCTION

A large scale deployment of third generation (3G) mobile networks is now takingplace. To pay back the investments made in network and service infrastructure,new end-user services should be developed, distributed, and adopted. To obtainwidespread adoption of these services, a set of requirement should be met. These re-quirements are technological, business strategic, and behavioral [1, 2]. First, com-plex services require an integration of network technologies, network, content, andsupplementary services. Second, adoption on the demand side requires wide-spread adoption of technology and service platforms among application develop-ers and service providers. Finally, end users implicitly specify a set of demand-side

JOURNAL OF ORGANIZATIONAL COMPUTINGAND ELECTRONIC COMMERCE 15(2), 203–222 (2005)

Correspondence and requests for reprints should be sent to Per E. Pedersen, Department of ICT,Agder University College, Grooseveien 36, 4876 Grimstad, Norway. Email: [email protected]

Dow

nloa

ded

by [

New

Yor

k U

nive

rsity

] at

00:

56 0

6 O

ctob

er 2

014

Page 3: Adoption of Mobile Internet Services: An Exploratory Study of Mobile Commerce Early Adopters

requirements that the services should meet. To understand these requirements,analyses of the context specific behavior of end-users should be conducted. Thesebehavioral, demand-side adoption requirements are our focus in this article.

Understanding the behavioral adoption requirements is important to both re-searchers and industry players. For researchers, an important issue is how mobileend-user services differ from traditional ICT services in ways that affect their adop-tion. For example, the personalization, location specificity, and ubiquity of theseservices are suggested as important characteristics making their adoption differentfrom other ICT services [3]. However, with the introduction of 3G services, a con-vergence is expected of mobile and traditional Internet-based services. Conse-quently, much of what has been learned from studies of the adoption of traditionalInternet services may be relevant to understanding the adoption of future mobileservices. From the perspective of industry players, understanding the process bywhich these services are adopted is also important. Three research questions are ofparticular interest: What services of this kind are likely to be adopted by end us-ers?, How does the end user decide to adopt these services?, and What influencesthat adoption decision? Of these questions, we focus on the last two in this article. Itis also likely that by investigating these questions, it may be possible to suggestwhat kinds of services are likely to be adopted under specific conditions.

To investigate these questions, we suggest combining theories and models oftraditional ICT adoption research with findings from behavioral studies of mobileservice adoption. Traditional Internet-based services are typically studied apply-ing diffusion and adoption research perspectives. These perspectives are not oftenapplied to studies of the adoption and use of mobile services. Instead, a domestica-tion research perspective is applied [4]. Even though domestication research stud-ies typically focus on the consequences of mobile service adoption and use, thesestudies have suggested relevant ways in which traditional adoption models maybe extended and modified to better explain the adoption of mobile services. In thisarticle, we suggest using domestication research findings to extend and modify thedecomposed theory of planned behavior (TPB; [5]) to provide a model explainingthe adoption of complex mobile services. We apply the model in an empirical studyof early adopters of mobile commerce services.

In the next section, we review some of the relevant adoption models used whenstudying the adoption of traditional Internet services, summarize findings in do-mestication research on the adoption and use of mobile services, and integratethese findings into a suggested adoption model. Next, we present the methodologyand results of a study of early adopters of mobile commerce. In the final section, wesummarize and discuss our main results, and we suggest directions for further re-search.

2. THEORY AND MODEL

Generally, studies of ICT adoption takes one of three possible approaches: a diffu-sion approach, an adoption approach, or a domestication approach. Diffusion re-searchers typically describe the aggregate adoption process a posteriori as an S-shaped function of time that may be used to categorize adopters of different kinds

204 PEDERSEN

Dow

nloa

ded

by [

New

Yor

k U

nive

rsity

] at

00:

56 0

6 O

ctob

er 2

014

Page 4: Adoption of Mobile Internet Services: An Exploratory Study of Mobile Commerce Early Adopters

[6]. Rogers [7] tried to explain the observed adoption behavior using characteristicsof the technology being introduced. Rogers [7] also described the diffusion processas consisting of four elements: an innovation or new technology, a social system, thecommunication channels of the social system, and time. Adoption researchers typi-cally describe and explain the adoption decision of individual end users applyingcognitive and social theories of decision making. Three models stand out as themost widely applied—the technology acceptance model (TAM), the theory of rea-soned action (TRA), and the TPB. Several hundred studies have been found apply-ing one of these models to explain end-users’ adoption and acceptance of differentkinds of ICT systems and applications [8]. Domestication research typically studiesthe adoption and use of technology in everyday life [9]. The perspective is domi-nated by sociologist researchers, and consequently, descriptive studies often char-acterize the adoption and use of technologies by demographic variables such as ageand gender. However, the main focus of domestication research is on the societalconsequences of the domestication of technology, that is, the process in which theuse of technology becomes integrated into a person’s everyday life. Because mypurpose for this article was to develop and test a model of individual end-useradoption, I focus on reviewing the contributions of adoption and domestication re-search relevant for the development of such a model.

2.1 ICT Adoption Research

In adoption research, the TAM [10, 11] focuses on the attitudinal explanations of in-tention to use a specific technology or service. It includes five concepts—perceiveduser friendliness, perceived usefulness, attitudes toward use, intention to use, andactual use. Although the model is mainly applied to explaining the adoption oftechnology within organizations, the constructs of the model are meant to be fairlygeneral [12]. Davis et al. [11] also originally described the variables of the model asuniversal to different types of computer systems and user populations. Thus, themodel may be applied to explain users’ intentions to use traditional Internet ser-vices [13, 14]. The TAM model has been both extended and modified. A typical ex-tension suggests antecedents and determinants of perceived user friendliness andperceived usefulness. Whereas the determinants of perceived user friendliness arebelieved to be rather general and have been given much attention, the determinantsof perceived usefulness are service dependent and have been given less attention[15, 16]. A second extension is suggested by introducing social determinants of useor intended use. Some have introduced these concepts as determinants of perceivedusefulness [8], whereas others have criticized the model for not incorporating suchissues at all [17]. A third extension has suggested including behavioral control oruser resources as explanatory concepts [18].

The TRA is a more general theory than the TAM and has been applied to explainbehavior beyond the adoption of technology. However, when applied to adoptionbehavior, the model includes four general concepts—behavioral attitudes, subjec-tive norms, intention to use, and actual use. The inclusion of subjective norm repre-sents an important addition when compared to the TAM. In TRA, subjective normis composed of users’ perception of how others think they should behave and their

AN EXPLORATORY STUDY OF MOBILE COMMERCE SERVICE ADOPTION 205

Dow

nloa

ded

by [

New

Yor

k U

nive

rsity

] at

00:

56 0

6 O

ctob

er 2

014

Page 5: Adoption of Mobile Internet Services: An Exploratory Study of Mobile Commerce Early Adopters

motivation to comply with the expectations of these referents [19]. TRA has beenapplied in its original form to explain the adoption of ICT applications [20], buttypically, TRA is used as a basis for modifying the TAM model with subjectivenorm as suggested previously.

The TPB was proposed as an extension of the TRA to account for conditions inwhich individuals do not have complete control over their behavior [21]. However,this theory also included determinants of behavioral attitudes and subjectivenorm. Models based on TPB have been applied to the explanation of different typesof behavior, but when applied to the adoption of ICT systems or services, themodel contains five concepts—behavioral attitudes, subjective norm, behavioralcontrol, intention to use, and actual use. The components of behavioral attitudeand subjective norm are the same in TPB as in TRA. In addition, the model includesbehavioral control as a perceived construct. Perceived behavioral control reflectsthe internal and external constraints on behavior and is directly related to both in-tention to use and actual use. Consequently, actual use is a weighted function of in-tention to use and perceived behavioral control.

The TPB has been applied to explain the adoption of such diverse systems asspreadsheets [22], computer resource centers [5], and recently, electronic com-merce services [17]. The role of subjective norm in TPB when compared to the TAMis somewhat unclear. Davis et al. [11] and Mathieson [22] found no support for a di-rect relation between subjective norm and intention to use. This has been attributedto little social pressure to use the systems studied by these authors. Later, a signifi-cant relation has been found both in studies in organizational [23] and electroniccommerce settings [17]. In a recent study, Venkatesh and Davis [16] also foundstrong support for a direct link between subjective norm and intention to use in astudy pooling results across four different studies and settings. The inclusion of be-havior control in the TPB model represents a valuable addition to the explanatorypower of TPB when compared to TAM. Both Mathieson and Taylor and Todd [5]found that the addition of behavioral control made their TPB model explain moreof the variance in intention to use than TAM. In TPB, behavioral control encom-passes two components. The first component is facilitating conditions representingthe resources required to use a specific system. Examples of such resources are fi-nancial resources or other ICT-related resources. The second component is self effi-cacy; that is, “an individual’s self-confidence in his/her ability to perform abehavior” ([5], p. 150).

TPB and TRA have both been criticized for not suggesting operational compo-nents or determinants of behavioral attitudes, subjective norm, and to some extent,behavioral control. To meet some of this criticism, specific components or determi-nants of the attitudinal concepts of the TPB model have been suggested. For exam-ple, Battacherjee [17] suggested incorporating the TAM model in TPB withperceived usefulness and user friendliness as the determinants of attitudes towarduse. Battacherjee also suggested subjective norm may be determined by externaland interpersonal influence and that the two components of perceived behavioralcontrol may also be treated as the determinants of behavioral control. Taylor andTodd [5] suggested what they termed a decomposed TPB, which also includes theTAM in the attitudinal part of TBP. However, Taylor and Todd also included com-patibility as a third determinant of attitude toward use, mainly inspired by the dif-

206 PEDERSEN

Dow

nloa

ded

by [

New

Yor

k U

nive

rsity

] at

00:

56 0

6 O

ctob

er 2

014

Page 6: Adoption of Mobile Internet Services: An Exploratory Study of Mobile Commerce Early Adopters

fusion theory of Rogers [7]. The determinants of subjective norm are believed to becontext dependent, and in the case of Taylor and Todd, peer influence and supe-rior’s influence are the suggested determinants. In mobile settings, Battacherjee’sdeterminants generally seem most relevant.

2.2 Behavioral Studies of Mobile Services

Recently, the perspective of domestication research has dominated behavioralstudies of mobile service use and adoption. The objects of study in research on theadoption of mobile end-user services may be the terminals, the services, or the us-ers. We found at least two types of terminal-oriented studies. One is those focusedon the design of the terminals [24], whereas the other focused on the terminal as anobject of expression [25]. Despite their lack of service orientation, these studies arerelevant to understanding the adoption and use of 3G end-user services. For exam-ple, the importance of terminals as objects of expression is also relevant to end-userservice adoption and should somehow be included in an adoption model of mobileservices. Service-oriented studies are often usability studies focusing on the inter-face between service and user. Most service-oriented studies are of users accessingsimple network services such as voice and messaging services [26]. Correspond-ingly few studies have focused on the kind of complex and integrating services thatare typical of 3G services such as, for example, mobile commerce services [1].Finally, the object of study may be the user of a mobile terminal and service. Greenet al. [27] described the studies focusing on the users of mobile services in four cate-gories: social science based studies that have treated the user as a social entity or as asocial actor and industry studies that have treated the user as an economic entity oran economic actor. Although studies of users as economic entities and actors are of-ten used by industry players, the findings in studies that have treated users as socialactors seem most relevant to the development of an adoption process model. Thus,we review some of the relevant findings of these studies organized by their contextsof study.

Much of the research that has studied the use of mobile services in work contextsis interesting because it studies the functional reasons for adoption. However, littleof it has been directed specifically at the adoption decision of end users. Instead,most mobile work studies have been usability studies applied to design user inter-faces and to develop work-related support applications. This differs from researchon technology acceptance in adoption research that has often focused on modelingthe adoption decision of end users in work-related contexts. Research on mobileservices in the leisure context has either focused directly on the functional use ofmobile services in leisure and everyday contexts [28] or on how the boundary be-tween the work and leisure contexts is blurred by the use of such services [29].These studies have indicated that there is a mix of work related and leisure relatedfunctional reasons for adoption, that the use of mobile services contributes to ablurring of work and leisure use, and that user friendliness is important at the earlystage of the adoption process but less important at later stages. This research hasalso indicated that services adopted for personal use in leisure context are oftenquickly adopted for professional use in work contexts. In addition, heavy users of

AN EXPLORATORY STUDY OF MOBILE COMMERCE SERVICE ADOPTION 207

Dow

nloa

ded

by [

New

Yor

k U

nive

rsity

] at

00:

56 0

6 O

ctob

er 2

014

Page 7: Adoption of Mobile Internet Services: An Exploratory Study of Mobile Commerce Early Adopters

mobile services are characterized by a large number of professional and personalroles, a high degree of integration between these roles, and the availability of com-munication media and services to support these roles [28]. Thus, it is likely to ex-pect that there will be a positive relation between the adoption of mobile commerceservices adopted for personal and professional reasons, especially among earlyadopters.

Domestication researchers often use demographic variables such as gender andage as proxies for identifying a distinction between end-user contexts. Age hasbeen the most widely applied demographic variable characterizing differences inadoption of mobile services. In these studies, the differences in adoption patternsbetween young people (teens, adolescents) and other users have been focused on.A recent conclusion is that service adoption and usage varies among young usersin a way that treating them as a homogeneous group with an adoption behaviorsystematically different from other age groups is not advisable [30]. Service usagefocuses on text and voice usage, with a slightly higher text service usage among fe-male than male users [31]. The use of mobile services is very well integrated in thedaily lives of teenagers. However, the impression that services are adopted fornonfunctional and social status reasons only [32] has been contradicted by recentdescriptive studies. For example, Karlsen et al. [33] found a remarkable orientationtoward usability and costs in their study of the potential adoption of new servicesamong Norwegian teenagers.

A number of explanations are suggested of young people’s adoption of mobileservices. Among these explanations are the suggestion that the adoption behaviorcan be illuminated by a “theory of fashion” [34], by the use of services as “ritual giftgiving” [25], by treating the mobile phone as “symbolic capital,” [32] or as an in-strument in “family differentiation and a symbol of individuality” [35, 36] and theuse of services as a “group marker or social identifier” [37] or as a “self-identifier”[38]. Currently, these explanations should all be treated as tentative because noneof them has undergone formal hypothesis development and confirmatory testing.Still, they suggest important explanations that eventually will have to be integratedas parts of a more formal theory of adoption. For example, the importance of inter-personal and media influence inspired by a theory of fashion should be a part ofsuch a theory. Similarly, the relation between social reasons for use and social rea-sons for adoption should somehow be integrated. A general conclusion that maybe drawn from these studies is that social networks and the position of the adopterin social networks are important determinants of adoption. These mechanismsshould somehow be integrated into a theory of mobile service adoption, but it isalso necessary that this theory integrates mechanisms in which services areadopted for functional reasons as well.

Because researchers from many different traditions and areas of research havebeen involved in behavioral end-user adoption studies, a multitude of methodolo-gies has been applied in these studies. A far greater number of studies applyingqualitative than quantitative social science methodologies have been found. For ex-ample, scenario analysis [39], focus-group interviews [34], traditional observa-tional methods [37], and recently, diary and log-based methodologies [28] havebeen applied. Quantitative methodologies have also been applied, for example, us-ing traditional survey [40] or quasi-experimental methodology [41], but the num-

208 PEDERSEN

Dow

nloa

ded

by [

New

Yor

k U

nive

rsity

] at

00:

56 0

6 O

ctob

er 2

014

Page 8: Adoption of Mobile Internet Services: An Exploratory Study of Mobile Commerce Early Adopters

ber of studies applying these methodologies to study the adoption of complexservices has been limited by the slow introduction and adoption of these services.Because both the models and methodologies of domestication research are explor-atory, the findings summarized here are better suited as a basis for extending adop-tion research models than as a basis for integrating of adoption research intodomestication research models.

2.3 Model

Because a convergence is expected of traditional Internet services and mobile ser-vices, we suggest that extending and modifying an adoption research model basedon findings in domestication research may be useful. However, in a choice of mod-els, parsimony is also important. From the perspective of explaining mobile serviceadoption, the inclusion of subjective norm and behavioral control from the decom-posed TPB seems necessary because these services are used in social contexts, andtechnology facilitating conditions certainly limit their potential adoption [1]. Thus,we suggest a modified version of the decomposed TPB model of Taylor and Todd[5]. As a basis for further modification, behavioral attitudes are determined by per-ceived usefulness and user friendliness as in the TAM. Subjective norm is deter-mined by external and interpersonal influence, and behavioral control is deter-mined by self-efficacy and facilitating conditions as in Battacherjee’s [17] version ofthe decomposed TPB. This model includes many of the most important social deter-minants of end-user adoption suggested by domestication research studies we pre-sented previously. To comply more fully with domestication research findings,some modifications of this basic model are suggested. First, we introduce directlinks from the determinants of subjective norm to the determinants of behavioral at-titudes because some of the most important determinants of usefulness and userfriendliness are user expectations communicated through media and social net-works. Thus, we include the relevance of a theory of fashion suggested by Ling [34]in the model. We also include links from subjective norm to behavioral attitudes be-cause studies of mobile services have shown how both attitudes toward use and be-havioral intentions to use these services are developed in social networks and are af-fected by social norms.

Studies in domestication research have also focused on the importance of indi-viduality and the relation between individuality and social pressure as both a de-terminant and consequence of mobile service use [32, 42]. Thus, determinants ofindividuality and resistance to social pressure should be included as determinantsor moderators of subjective norms. We suggest including the concept of self-control as an extension of the self-efficacy concept of TPB and as an additionalmoderator of subjective norm. Whereas self-efficacy (related to adoption) is an in-dividual’s self-confidence in that adoption will lead to the desired behavior [43],self-control is often believed to include self-efficacy but also go beyond it [44]. Forexample, self-control is related to time dependence when an individual choosesnot to consume something today because the utility is believed to be higher fromconsuming the good at a later point in time.

When using the decomposed TBP as a basis, the behavioral control part of themodel already includes many of the relevant issues treated in domestication re-

AN EXPLORATORY STUDY OF MOBILE COMMERCE SERVICE ADOPTION 209

Dow

nloa

ded

by [

New

Yor

k U

nive

rsity

] at

00:

56 0

6 O

ctob

er 2

014

Page 9: Adoption of Mobile Internet Services: An Exploratory Study of Mobile Commerce Early Adopters

search. For example, facilitating conditions may include the importance of supplyside actions taken by service providers and operators to stimulate quick adoptionof services. Examples of such actions are the offerings of low-price and prepaidschemes. In addition, the self-efficacy we discussed previously includes many ofthe issues found relevant in studies of the work/leisure context and studies of howa user very quickly adopts to the idea of being a mobile service user [28]. The sug-gested model is shown in Figure 1.

Until more research has been done on integrating domestication research find-ings into adoption research models, our model should be treated as indicative.Thus, formal hypotheses should so far not be developed based on the model. How-ever, some rather general propositions may be put forward. We suggest a generalproposition that our model concepts and measures are reliable and valid (P1). Sec-ond, we propose that based on the findings of domestication research we reviewedpreviously, the simple TAM should be extended with subjective norm and behav-ioral control as in the TPB (P2). Finally, we propose that based on our review of do-mestication research findings, the traditional TPB model should be modified (P3)with the addition of expectancy relations, social influence of attitudes toward use,and self-control. We gave the theoretical arguments for the extensions and modifi-cations previously.

210 PEDERSEN

Figure 1. Decomposed theory of planned behavior modified by domestication researchfindings.

Dow

nloa

ded

by [

New

Yor

k U

nive

rsity

] at

00:

56 0

6 O

ctob

er 2

014

Page 10: Adoption of Mobile Internet Services: An Exploratory Study of Mobile Commerce Early Adopters

3. METHODOLOGY

To analyze our propositions, we conducted an empirical study of early adopters ofmobile commerce services. We applied a simple one-group posttest design in whichrespondents were recruited from discussion forums, bulletin boards, or communitygroups on the Web or newsgroups on news servers. Altogether, we identified 85 fo-rums containing active users discussing mobile phones or personal digital assis-tants. In addition, we posted an invitation to the “Keitai-L” and “tokyopc-mobile”e-mail lists to better recruit respondents from the Asian region. All material was col-lected during the first 2 weeks of December 2001. The total sample is shown in Table1 with the corresponding forums from which the respondents were recruited.

The final sample consisted of 232 respondents answering the major parts of thesurvey. A minimum completion time was set to 3 min, and all participants whocompleted the whole survey in less than 3 min were considered careless respon-dents and excluded from further analysis (4 participants). Average completiontime for the rest of the participants was 11 min, 49 sec.

We primarily recruited respondents located in North America and Europe de-spite efforts to reach respondents in Asia as well. When compared to the generalInternet population, our participants were somewhat younger than those reportedin the Georgia Tech University 10th GVU User Survey.1 When compared to theGVU User Survey, female users were underrepresented in our sample. For the edu-cation variable, our participants were very similar to the participants in the GVUUser Study. When considering the issue of representative samples, one shouldkeep in mind that our participants were meant to be representative of early adopt-ers, not necessarily of Internet users in general. Thus, the low number of femaleparticipants and the higher general education of our participants are generally inaccordance with other studies of early adopters [45].

We designed a questionnaire containing multiple measures of each of the 11concepts of our model. The concepts were measured by the participants indicatingtheir agreement with a set of statements using a 7-point scale ranging from 1(strongly disagree) to 7 (strongly agree). We measured user friendliness using itemsfrom the original items of Davis et al. [11] adapted to our setting. Similar operationsare found also in Taylor and Todd [5] and in Battacherjee [17]. Usefulness was mea-sured using five items indicating the original dimensions of time saving, improve-ment, efficiency, usefulness, and quality of Davis [10]. Because the setting of mobilecommerce services is an everyday life situation, we had to convert the originalitems of Davis into everyday life terms. The items were then discussed and pre-tested with native English speaking participants. We measured attitude towarduse using five bipolar adjectives indicating different aspects of the attitude towarduse. The items were very similar to those that have been used by Davis, Taylor andTodd, and Battacherjee. The measure of external influence was based on threesources of influence—media, society, and profession. Thus, it both includes and ex-tends the measures used by Battacherjee and Taylor and Todd. We based the mea-sure of interpersonal influence on Battacherjee’s extension of the measures used by

AN EXPLORATORY STUDY OF MOBILE COMMERCE SERVICE ADOPTION 211

1See http://www.gvu.gatech.edu/user_surveys/

Dow

nloa

ded

by [

New

Yor

k U

nive

rsity

] at

00:

56 0

6 O

ctob

er 2

014

Page 11: Adoption of Mobile Internet Services: An Exploratory Study of Mobile Commerce Early Adopters

Taylor and Todd and adapted to our setting. We measured self-control by items re-flecting indicators of self-control such as resisting group pressure, superior influ-ence, and group conformity. The items were based on a subsection of the self-control measure suggested by Rosenbaum [44]. We measured subjective norm us-ing three items almost identical to the items that have been used by Mathieson [22],Battacherjee, and Venkatesh and Davis [16]. We based the self-efficacy, facilitatingconditions, and behavioral control measures on Battacherjee and Taylor and Todd.We measured actual use and intention to use by presenting 30 mobile commerceservices organized along the consumer life cycle and included both personal andprofessional mobile commerce services. We asked the participants to indicate ifthey had used any of these services and if they intended to use any of these serviceswithin the next 6 months. In Table 2, the five most widely adopted services areshown along with the five services with the highest usage intentions, illustratingthat the most frequently indicated services were mainly personal mobile commerceservices. Thus, the stimulus context of the study was not focusing particularly onservices for personal or professional use, but most participants seemed to self-interpret the context as a setting of personal mobile commerce service use.

We applied the sums of the number of services indicated as measures of actualuse and intentions to use. We also measured intention to use with a three-item scaleadapted from Battacherjee [17] and Mathieson [22]. Consequently, we based all ourmeasures on previously validated measures (see Venkatesch & Davis [8]). Mea-

212 PEDERSEN

Table 1Recruitment Forums and Number of Responses

Forums N Newsgroups N

Unidentified 9 alt.cellulara 6www.forum.nokia.com 7 uk.telecoma 5www.esato.com 18 de.compa 1www.microsoft.com 10 dk.marked 1www.reviewcentre.com 1 dk.teknika 4www.wirelessinanutshell.com 9 fr.reseaux.telecomsa 1www.brighthand.coma 15 no.it.telekom.mobil 11www.pdastreet.coma 27 tw.bbs.rec.mobilecomm 4www.howardforums.coma 19 uk.adverts.telecom.mobile 2www.pocketpcpassion.com 2 aus.comms.mobile 8www.syllas.com 3 pl.misc.telefonia.gsm 1clubs.yahoo.coma 10 hr.ponuda.gsm 1www.mobildebat.dk 2 hr.alt.cellular.gsm 1debat.passagen.se/mobil 1 es.technica.redes.telefonia.movi 2www.forum.siemens.com 2 it.tlca 1forum.hardware.no 1 comp.sys.palmtops.pilot 3www.handyfragen.de 5 se.sator.sys.handdator 2cell.exchange.ph 1 swnet teknikk.telefoni 3forums.internet.com 1 sfnet.viestinta.matkapuhelimet 5www.mobilecomputing.com 1www.popularwireless.com 1 Keitai-L 18www.itavisen.no 6 Tokyopc-mobile 1

aPostings on several subforums within the indicated forum were made.

Dow

nloa

ded

by [

New

Yor

k U

nive

rsity

] at

00:

56 0

6 O

ctob

er 2

014

Page 12: Adoption of Mobile Internet Services: An Exploratory Study of Mobile Commerce Early Adopters

sures and corresponding reliabilities are shown in Table 3. All measurement itemsare shown in the Appendix.

In addition to the concepts of our adoption model, we also measuredinnovativeness, adapting Goldsmith and Hofacker’s [46] product innovativenessmeasure to our setting of mobile services. Goldsmith [47] also applied thisinnovativeness measure to Internet users to identify innovators, majority users,and laggards. The measure had a reliability of α = 0.95. Goldsmith [47] used a 5-point scale instead of our 7-point scale and observed a mean of 18.0 and a standarddeviation of 5.4. Goldsmith [47] identified innovators as the 20% of the participantswith the highest innovativeness score. When converted to Goldsmith’s [47] scale,our participants had a mean innovativeness of 23.5 and a standard deviation of 6.This indicates that the majority of our participants were in the innovator category,and our assumption that our participants were early adopters seems appropriate.

4. RESULTS

From Table 2, we find that the mobile commerce services most frequently adopted aresimple services such as direct download of services to the mobile terminal and searchand alert services. Most likely, these services have been implemented using shortmessaging service (SMS)-based platforms. However, the users expected quite differ-ent services to be adopted in the near future. In particular, payment services and loca-tion based services were indicated as interesting for near future use. The expectationsof users corresponded well to what has been communicated in the mobile service in-dustry as services likely to be introduced in the near future. In general, these servicesare mainly adopted for personal use. Thus, the following results should mainly be in-terpreted within the context of personal mobile commerce services.

AN EXPLORATORY STUDY OF MOBILE COMMERCE SERVICE ADOPTION 213

Table 2Most Widely Adopted Mobile Commerce Services

Services Adopted Frequency (%)Services Intended to Be Used

in the Next 6 Months Frequency (%)

Direct product/servicedownload/buying to mobiledevice (ring tones, logos, musicdownload)

54.8 Mobile payment service at the“point-of-sale” (credit cardreplacement)

39.9

Searching “yellow pages” orother commercial directoryusing a mobile device

42.5 Mobile payment service onvending machine (andparking lot, toll roadpayment)

38.2

Alert service related tocommercial matters such asoffers, warnings, and so forth

40.4 Mobile payment service onthe Internet

37.3

Consumer related gaming andentertainment services onmobile device

39.5 Mobile location-dependentsearch or alert service

36.8

Mobile access to entertainmentrelated reservation service(cinema, concert tickets)

37.3 Searching “yellow pages” orother commercial directoryusing a mobile device

34.6

Dow

nloa

ded

by [

New

Yor

k U

nive

rsity

] at

00:

56 0

6 O

ctob

er 2

014

Page 13: Adoption of Mobile Internet Services: An Exploratory Study of Mobile Commerce Early Adopters

To further examine the validity of the measurement model, we conducted a con-firmatory factor analysis. In the analysis, the measurement model showed accept-able fit.2 The observed fit indexes were χ2/df = 2.13, normed fit index (NFI) = 0.95,comparative fit index (CFI) = 0.97, incremental fit index (IFI) = 0.97, and root meansquare error of approximation (RMSEA) = 0.075. Table 3 shows all items used, cor-responding reliabilities, squared multiple correlations, and standardized regres-sion weights.

Based on the discussion of reliabilities in the methodology section and the re-sults of Table 3, we conclude that the measurement model applied in our study was

214 PEDERSEN

Table 3Results of Measurement Model Analysis

Concept Item R2 β Coefficients Cronbach’s α

User friendliness 1 .52 .72* .932 .67 .82*3 .76 .87*4 .85 .92*5 .80 .89*

Usefulness 1 .68 .82* .912 .57 .75*3 .85 .92*4 .65 .80*5 .69 .83*

External influence 1 .59 .77* .732 .72 .85*3 .26 .51*

Interpersonal influence 1 .46 .68* .862 .69 .83*3 .83 .91*4 .55 .73*

Self-control 1 .69 .83* .832 .73 .86*3 .49 .70*

Self efficacy 1 .70 .83* .872 .31 .56*3 .80 .90*4 .87 .93*

Facilitating conditions 1 .31 .56* .822 .34 .58*3 .44 .66*4 .61 .78*5 .57 .76*6 .43 .65*

*p < .01.

2We generally employ parsimony adjusted measures of fit only. According to Browne and Cudeckcited in Arbuckle and Wothke [49], an RMSEA less then 0.08 is acceptable. According to Bentler cited inBattacherjee [17], χ2/df should be less than 5, preferably less than 2, and all other indexes should be closeto 1 [5]. In general, we apply the rules that χ2/df ≈ 2 or better, RMSEA < 0.08, and all other indexes ≈ 1indicate acceptable fit.

Dow

nloa

ded

by [

New

Yor

k U

nive

rsity

] at

00:

56 0

6 O

ctob

er 2

014

Page 14: Adoption of Mobile Internet Services: An Exploratory Study of Mobile Commerce Early Adopters

reliable and valid. Thus, we found no need for any major redesign of the measuresused in the TAM and the TRA and TPB models when applied to study the adoptionof mobile commerce services except from adapting measures to the relevant con-text of the service being studied. These findings support P1 and indicate that atleast when viewed from a measurement perspective, adoption research modelsmay successfully be applied to the study of mobile service adoption.

Turning to model estimation, we started with estimating the simple TAM, thenadded the subjective norm part of our model, and finally added the behavioral con-trol part. In all additions, the relations were as indicated in Figure 1. The simpleTAM showed good fit with χ2/df = 1.90, NFI = 0.97, CFI = 0.99, IFI = 0.99, andRMSEA = 0.066. However, the model only explained 30% of the variance in inten-tion to use and 17% of the variance in actual use. Extending the model with subjec-tive norm and expectancy relations produced a model with similar fit and noincrease in explanatory power. The exact fit indexes were χ2/df = 1.83, NFI = 0.96,CFI = 0.98, IFI = 0.98, RMSEA = 0.064, and the explained variance exactly the sameas in the simple TAM. Thus, a more complex model including subjective normshowed no improvement in explanatory power and data fit.

When we added the behavioral control part to the model, it showed an accept-able fit, and the exact fit indexes were χ2/df = 2.12, NFI = 0.92, CFI = 0.96, IFI = 0.96,and RMSEA = 0.074. The model showed a considerable increase in explanatorypower. In Figure 2, all fit indexes, squared multiple correlations, and regression co-efficients are shown. In general, squared multiple correlations are shown withinthe squares, indicating each latent variable and standardized regression coeffi-cients, and corresponding levels of significance are shown along each relation ar-row. Only significant relations are shown. The first conclusion that may be drawnfrom the comparisons of the three estimated models is that although adding sub-jective norm did not individually contribute to increased explanatory power, add-ing behavioral control certainly did.

Thus, one is led to believe that the subjective norm part of the model may be ex-cluded. However, when trying to remove the subjective norm part of the modelfrom the final model, model fit was unacceptable. The fit indexes were χ2/df = 2.57,NFI = 0.94, CFI = 0.96, IFI = 0.96, and RMSEA = 0.088, indicating that the simplermodel without subjective norm showed a considerably worse fit than the complexmodel. Furthermore, the explanatory power of this model was not as good as thatof the complex model. Whereas the complex model of Figure 1 explained 49% ofthe variance in intention to use mobile commerce services, the reduced model ex-plained 46% of the variance in intention to use. Thus, our second conclusion is thatsubjective norm both improves model fit and adds to the explanatory power of themodel when being combined with behavioral control. Even though our tests arequalitative and exploratory, we conclude that the simple TAM should be extendedwith both subjective norm and behavioral control when explaining the adoption ofmobile commerce services. These findings generally support P2.

When comparing the suggested model of Figure 1 and the final model of Figure2, a set of observations were made. First, we found a significant relation betweenexternal influence and perceived usefulness. Second, we found no direct effect ofuser friendliness on attitudes toward use. Thus, the effect of user friendliness isonly indirect through perceived usefulness. Third, no direct effects were found be-

AN EXPLORATORY STUDY OF MOBILE COMMERCE SERVICE ADOPTION 215

Dow

nloa

ded

by [

New

Yor

k U

nive

rsity

] at

00:

56 0

6 O

ctob

er 2

014

Page 15: Adoption of Mobile Internet Services: An Exploratory Study of Mobile Commerce Early Adopters

tween usefulness and intention to use. Fourth, we found no direct effects of behav-ioral control on actual use. However, we found the suggested effect of subjectivenorm on attitudes toward use, indicating that attitudes toward use are socially in-fluenced in the model both directly through subjective norm and indirectlythrough external influence. We also found that social control was a very importantmoderator of subjective norm. It contributed significantly to our model and ex-plained 45% of the variance in subjective norm. Without the social control variable,38% of the variance in subjective norm was explained. Thus, we conclude that thedecomposed TPB should be further modified when explaining the adoption of mo-bile commerce services. This finding generally supports P3. Even though all ourpropositions were only tested qualitatively, we conclude that they are generallysupported by our analyses.

216 PEDERSEN

Figure 2. Final model; *p < .05, **p < .01.Dow

nloa

ded

by [

New

Yor

k U

nive

rsity

] at

00:

56 0

6 O

ctob

er 2

014

Page 16: Adoption of Mobile Internet Services: An Exploratory Study of Mobile Commerce Early Adopters

5. CONCLUSIONS AND DISCUSSION

For the theoretical part of this study, we concluded that there seems to be a lack ofstudies that have applied traditional ICT-adoption theory to the adoption of mobileservices. Instead, domestication research represents the dominating theoreticalperspective in studies of mobile service adoption. However, findings from domesti-cation research may be used to extend and modify current adoption models tobetter explain the adoption of mobile services. We suggested extending the TAMwith subjective norm and behavioral control into a decomposed TPB. The modifica-tions consisted of adding relations in the model from subjective norm to attitude to-ward use, adding expectancy relations between external and interpersonal influ-ence and usefulness and user friendliness, and of introducing the concept of self-control as a moderator of subjective norm.

In an empirical study of early adopters of mobile commerce services, we foundthat the traditional measures of adoption research models were reliable and thatthey also could be defended theoretically when being applied to explain the adop-tion of mobile commerce services. We found support for a need to extend the TAMwith behavioral control. We found less support for extending the model with sub-jective norm, but when extending the model with both subjective norm and behav-ioral control, the subjective norm part also contributed to good fit with the data,and the effects of subjective norm on intentions to use were significant. Our find-ings also generally supported the proposition that the decomposed TPB should bemodified when explaining the adoption of mobile commerce services.

There are some threats to the internal and external validity of our study. Thenumber of participants could have been larger, but very few of the relevant modelcoefficients were found in the category of “close to significant” or “just significant.”In fact, most of the relevant coefficients were significant at the 1% level, indicatingthat the power of our tests is acceptable. One of the other issues that may havethreatened the internal validity of our study is the self-selection of participants. Forthis to be a threat to validity though, the selection procedure should systematicallyfavor recruitment of participants that make a complex adoption decision not easilycaptured by simple models such as the TAM that do not focus on user friendliness,that are not easily influenced by social norms, and that are very self-confident intheir ability to utilize mobile services. Most of these characteristics are relevant toinnovators and early adopters in general, and we see no reason that our innovatorswere systematically different from other users in the innovator category. This isalso supported by an analysis of the demographics of the users. Another issue is thecontextual situation of our study. Participants were asked to report their opinionsrelating to their own experience with mobile commerce services, but history andmaturation effects may have occurred. However, these issues are only relevant ifhistory and maturation effects contributed systematically to our findings. We seeno reason why these effects should not contribute to increased unsystematic errorrather than systematic bias in the direction of our results. Still, there are severalthreats to the external validity of the study. First, the participants of the study wererecruited as representative of early adopters of mobile services. Less innovativeparticipants may rely more on user friendliness of services, be more influenced bytheir peers, and put more weight on the self-identifying or social identifying role of

AN EXPLORATORY STUDY OF MOBILE COMMERCE SERVICE ADOPTION 217

Dow

nloa

ded

by [

New

Yor

k U

nive

rsity

] at

00:

56 0

6 O

ctob

er 2

014

Page 17: Adoption of Mobile Internet Services: An Exploratory Study of Mobile Commerce Early Adopters

using mobile services. Second, the setting of the study was related to the serviceswe investigated and the research setting we introduced. We have discussed the in-ternal validity of the research setting as a setting in which participants are re-minded of their use of mobile commerce services. As such, the setting isrepresentative of a potential adoption setting of a particular mobile commerce ser-vice. However, the importance of functional and instrumental factors in the deci-sion to adopt mobile commerce services may indicate that our findings are lessrelevant to the potential adoption decision of less functional services, such as, forexample, gaming services, personalization services, or multimedia services on mo-bile terminals. Still, our conclusions are relevant to other services in which func-tional issues are relevant such as network mediating services, personal informationmanagement (PIM) services, or mobile access to company intranet services. Thatsaid, the study should be treated as exploratory. Because priority was given to in-ternal validity rather than external validity, our conclusions are mainly relevant tounderstand the adoption decisions of innovators and early adopters using instru-mental mobile services.

Some of our findings require further discussion. We only found expectancy rela-tions between external influence and usefulness. However, this relation is particu-larly important because it indicates that users’ expectations of usefulness areimportant in their adoption process. Thus, if these expectations are not fulfilled,initial adoption may turn into what Carroll et al. [48] termed disappropriation.Disappropriation means that after an initial expectancy driven adoption, actual ex-perience with a service may reveal hidden costs or lack of usefulness that makesthem discontinue their use. Because it is likely that early adopters’ initial adoptionis expectancy driven, fulfilling these expectations may be particularly important toensure continued use among these users. The rather weak support for includingsubjective norm in the model contrasts the importance of interpersonal influencethat has been documented in domestication research. The role of subjective normhas been unclear in adoption research [22], and this has mainly been attributed tothe type of ICT applications and services that have been studied in adoption re-search. One would expect that for mobile commerce services, subjective normshould play a more important role. However, there seems to be a relation betweenthe instrumentality of a service and the importance of subjective norm that is unex-plored. For example, both the lack of a direct relation between usefulness and in-tentions often found in studies of ICT adoption in professional contexts and theweak support for subjective norm may be explained by the participants who pri-marily focused on services for personal use. For professional services, managerialincentives stimulating use are often reflected in observed direct effects of useful-ness and high influence of professional norms. We also found little support for a di-rect effect of user friendliness on attitudes toward use. Ling [31] suggested thatuser friendliness may not be an issue in mobile services because most of these ser-vices are almost self explanatory and very easy to use. A more plausible explana-tion is that user friendliness only has an effect on attitude to use in the early phaseof an adoption process but quickly looses relevance. This is supported by a longitu-dinal study by Venkatesh and Davis [16] that showed that user friendliness lostmuch of its explanatory power in the TAM when studying users over a 3-monthperiod. Our finding of a relation between external influence and perceived useful-

218 PEDERSEN

Dow

nloa

ded

by [

New

Yor

k U

nive

rsity

] at

00:

56 0

6 O

ctob

er 2

014

Page 18: Adoption of Mobile Internet Services: An Exploratory Study of Mobile Commerce Early Adopters

ness indicates an expectancy relation between the two concepts. To confirm this re-lation, an expectancy model should be applied to study this relation separately. Afinal finding was the lack of support for a direct effect of behavioral control on ac-tual use. We have attributed this lack of finding to our measure of actual use beingbased on the number of services used, not the frequency or amount of use. A simi-lar observation of lack of explanatory power when applying an unweighted fre-quency measure was made by Lederer et al. [13]. Thus, our measure of actual useneeds refinement in future studies and will have to be both weighted with amountof use and combined with attitudinal measures.

Even though the developed model seems promising in providing explanations,and not just descriptions, of adoption behavior, our research should be extended inseveral ways. First, our model explains the adoption decisions of individuals usingperceived concepts such as usefulness and self-efficacy. However, the determinants ofthese perceived concepts have not been investigated. For example, service propertiesmay determine perceived usefulness, whereas individual traits may determine self-efficacy. Furthermore, operator characteristics may be an important determinant ofthe perceived facilitating conditions. Suggesting and testing such determinants are im-portant issues in future research. Second, the importance of both determinants andperceived concepts may vary across mobile services. For example, the adoption of mo-bile commerce services may be less influenced by the determinants of subjective norm,whereas these determinants, and the subjective norm concept itself, may be more im-portant for modeling the adoption of services for the management of social relation-ships. Before service providers and developers can use our model as a basis fordeveloping an adoption evaluation framework, more research is needed on how theimportance of the determinants and perceived concepts vary across services. We cur-rently apply our adoption model to text messaging, chat services, payment services,and gaming services to compare the model across services and get a better under-standing of the adoption process of these complementary services. Third, the samemay be true for different categories of users. For innovators or early adopters with ex-perience and knowledge of mobile services, user friendliness may not be an issue indeveloping attitudes toward use. In a group of late adopters or laggards, user friendli-ness may be much more important in explaining the decision to adopt a specific mo-bile service. Thus, we should also extend our research to investigating the adoptiondecisions of users of various categories. Still, we suggest that the theoretical and em-pirical work done in this study provides a solid basis for extending our research in thesuggested directions. We also suggest that the model provides a good basis for indus-try players developing a service evaluation framework to determine the adoption po-tential of new services. In particular, the model seems well suited for developing suchframeworks for services that are adopted for functional reasons and services that aredirected specifically at innovative user categories.

REFERENCES

[1] P. E. Pedersen, “An adoption framework for mobile commerce,” presented at the I3E—First IFIPConference of E-Commerce, Zürich, Switzerland, 2001.

[2] P. E. Pedersen, L. B. Methlie, and H. Thorbjørnsen, “Understanding mobile commerce end-useradoption: A triangulation perspective and suggestions for an exploratory service evaluationframework,” presented at the HICSS–35, Big Island, Hawaii, 2002.

AN EXPLORATORY STUDY OF MOBILE COMMERCE SERVICE ADOPTION 219

Dow

nloa

ded

by [

New

Yor

k U

nive

rsity

] at

00:

56 0

6 O

ctob

er 2

014

Page 19: Adoption of Mobile Internet Services: An Exploratory Study of Mobile Commerce Early Adopters

[3] M. Rask and N. Dholakia, “Next to the customer’s heart and wallet: Frameworks for exploring theemerging m-commerce arena,” Proc. AMA Winter Marketing Educator’s Conf., vol. 12, 2001, pp.372–378.

[4] L. Haddon, “An agenda for mobility in everyday life for ICT researchers,” Cost 269 Working Pa-per, London School of Economics, London, 2001.

[5] S. Taylor and P. A. Todd, “Understanding information technology usage: A test of competingmodels,” Information Systems Research, vol. 6, no. 2, pp. 144–176, 1995.

[6] V. Mahajan, E. Muller, and F. M. Bass, “New product diffusion models in marketing: A review anddirections for research,” Journal of Marketing, vol. 54, no. 1, pp. 1–27, 1990.

[7] E. M. Rogers, Diffusion of Innovations, 4th ed. New York: Free Press, 1995.[8] V. Venkatesch and F. D. Davis, “A model of the antecedents of perceived ease of use: Development

and test,” Decision Sciences, vol. 27, no. 3, pp. 451–481, 1996.[9] R. Silverstone and E. Hirsch, Consuming Technologies. London: Routledge, 1992.

[10] F. D. Davis, “Perceived usefulness, perceived ease of use, and user acceptance of information tech-nology,” MIS Quarterly, vol. 13, no. 3, pp. 319–340, 1989.

[11] F. D. Davis, R. P. Bagozzi, and P. R. Warshaw, “User acceptance of computer technology: A com-parison of two theoretical models,” Management Science, vol. 35, no. 8, pp. 982–1002, 1989.

[12] W. J. Doll, A. Hendrickson, and X. Deng, “Using Davis’s perceived usefulness and ease-of-use in-struments for decision making: A confirmatory and multigroup invariance analysis,” Decision Sci-ences, vol. 29, no. 4, pp. 839–869, 1998.

[13] A. L. Lederer, D. J. Maupin, M. P. Sena, and Y. Zhuang, “The technology acceptance model and theWorld Wide Web,” Decision Support Systems, vol. 29, no. 3, pp. 269–282, 2000.

[14] J. C. C. Lin and H. Lu, “Towards an understanding of the behavioral intention to use a web site,” In-formation Management, vol. 20, no. 3, pp. 197–208, 2000.

[15] V. Venkatesh, “Determinants of perceived ease of use: Integrating control, intrinsic motivation,and emotion into the technology acceptance model,” Information Systems Research, vol. 11, no. 4, pp.342–365, 2000.

[16] V. Venkatesh and F. D. Davis, “A theoretical extension of the technology acceptance model: Fourlongitudinal field studies,” Management Science, vol. 46, no. 2, pp. 186–204, 2000.

[17] A. Battacherjee, “Acceptance of e-commerce services: The case of electronic brokerages,” IEEETrans. Syst., Man, Cybern., vol. 30, pp. 411–420, 2000.

[18] K. Mathieson, E. Peacock, and W. W. Chin, “Extending the technology acceptance model: The in-fluence of perceived user resources,” Advances in Information Systems, vol. 32, no. 3, pp. 86–112,2001.

[19] M. Fishbein and I. Ajzen, Belief, Attitude, Intention and Behavior: An Introduction to Theory and Re-search. Reading, MA: Addison-Wesley, 1975.

[20] J. K. Liker and A. A. Sindi, “User acceptance of expert systems: A test of the theory of reasoned ac-tion,” Journal of Engineering and Technology Management, vol. 14, no. 2, pp. 147–173, 1997.

[21] I. Ajzen, “From intentions to actions: A theory of planned behavior,” in Action Control: From Cogni-tion to Behavior, J. Kuhl and J. Beckmann, Eds. New York: Springer, 1985, pp. 11–39.

[22] K. Mathieson, “Predicting user intentions: Comparing the technology acceptance model with thetheory of planned behavior,” Information Systems Research, vol. 2, no. 3, pp. 173–191, 1991.

[23] G. C. Moore and I. Benbazat, “An empirical examination of a model of the factors affecting utiliza-tion of information technology by end-users,” Working Paper, University of British Columbia,Nov. 1993.

[24] M. C. Chuang, C. C. Chang, and S. H. Hsu, “Perceptual factors underlying user preferences towardproduct form of mobile phones,” International Journal of Industrial Ergonomics, vol. 27, no. 4, pp.247–258, 2001.

[25] A. S. Taylor and R. H. R. Harper, “The gift of the gab?: A design oriented sociology of young peo-ple’s use of ’mobilZe!, Working Paper, Digital World Research Centre, University of Surrey, Eng-land, 2001.

[26] M. Eldridge and R. Ginter, “Studying text messaging in teenagers,” presented at the CHI 2001Workshop on mobile communications, Seattle, WA, 2001.

[27] N. Green, R. H. R. Harper, G. Murtagh, and G. Cooper, “Configuring the mobile user: Sociologicaland industry views,” Personal and Ubiquitous Computing, vol. 5, no. 2, pp. 146–156, 2001.

[28] L. Palen, M. Salzman, and E. Youngs, “Discovery and integration of mobile communications in ev-eryday life,” Personal and Ubiquitous Computing, vol. 5, no. 2, pp. 109–122, 2001.

220 PEDERSEN

Dow

nloa

ded

by [

New

Yor

k U

nive

rsity

] at

00:

56 0

6 O

ctob

er 2

014

Page 20: Adoption of Mobile Internet Services: An Exploratory Study of Mobile Commerce Early Adopters

[29] D. Gant and S. Kiesler, “Blurring and boundaries: Cell phones, mobility and the line between workand personal life,” in Wireless World: Social and Interactional Aspects of the Mobile Age, B. Brown, N.Green, and R. Harper, Eds. London: Springer, 2001, pp. 121–132.

[30] V. Oksman and P. Raitiainen, “Perhaps it is a body part. How the mobile phone became an organicpart of the everyday lifes of children and teenagers,” presented at the Nordiska konferensen förmedie- ock kommunikationsforskning, Reykavik, Iceland, 2001.

[31] R. Ling, “The diffusion of mobile telephony among Norwegian teens: A report from after the revo-lution,” presented at the ICUST 2001, Paris, France, 2001.

[32] B. Skog, “ Mobiles and the Norwegian teen: Identity, gender and class,” in Perpetual Contact, J. E.Katz and M. Aakhus, Eds. New York: Cambridge University Press, 2002, pp. 255–273.

[33] M. E. Karlsen, I. Helgemo, and M. Gripsrud, “Useful, cheap and fun: A survey of teenagers de-mands for mobile telephony,” Telenor R&D, Oslo, Norway, Research Rep., 2001.

[34] R. Ling, “It is ‘in.’ It doesn’t matter if you need it or not, just that you have it: Fashion and the do-mestication of the mobile telephone among teens in Norway,” Telenor Working Paper, TelenorR&D, Oslo, Norway, 2001.

[35] R. Ling, “We release them little by little: Maturation and gender identity as seen in the use of mo-bile telephony,” Personal and Ubiquitous Computing, vol. 5, no. 2, pp. 123–136, 2001.

[36] A. S. Taylor and R. H. R. Harper, “Talking activity: Young people and mobile phones,” presentedat the CHI 2001 Workshop on Mobile Communications, Seattle, WA, 2001.

[37] A. Weilenmann and C. Larsson, “On doing ‘being teenager’. Applying ethnomethodology to theanalysis of young people’s use of mobile phones,” presented at the IRIS 23, University ofTrollhättan, Uddevalla, Sweden, 2000.

[38] P. S. Alexander, “Teens and mobile phones growing-up together: Understanding the reciprocal in-fluence on the development of identity,” presented at the Wireless World Workshop, University ofSurrey, England, 2000.

[39] T. Dörsch and U. Fiebig, “Scenarios and their consequences, Part 2,” presented at the WirelessWorld Research Forum III, Stockholm, Sweden, 2001.

[40] C. H. Marcussen, “WAP for travel and tourism services,” presented at the IFITT Workshop on Mo-bile Applications, Innsbruck, Austria, 2001.

[41] J. Cattell, “The mobile internet and its implications for research,” presented at the NetEffects4, Bar-celona, Spain, 2001.

[42] L. Fortunati, “The ambiguous image of the mobile phone,” Farsta, Sweden: Telia, Cost 248 Report,1998.

[43] A. Bandura, “Self-efficacy mechanism in human agency,” American Psychologist, vol. 37, no. 2, pp.122–147, 1982.

[44] M. Rosenbaum, “A schedule for assessing self-control behaviours: Preliminary findings,” Behav-iour Therapy, vol. 11, no. 1, pp. 109–121, 1980.

[45] M. D. Dickerson and J. W. Gentry, “Characteristics of adopters and non-adopters of home comput-ers,” Journal of Consumer Research, vol. 10, pp. 225–235, 1983.

[46] R. E. Goldsmith and C. F. Hofacker, “Measuring consumer innovativeness,” Journal of the AcademyMarketing Science, vol. 19, no. 3, pp. 209–221, 1991.

[47] R.E. Goldsmith, “Using the domain specific innovativeness scale to identify innovative Internetconsumers,” Internet Research: Electronic Networking Applications and Policy, vol. 11, no. 2, pp.149–158, 2001.

[48] J. Carroll, S. Howard, F. Vetere, J. Peck, and J. Murphy, “Just what do the youth of today want?Technology appropriation by young people,” presented at the HICSS 35, Big Island, Hawaii,2002.

[49] J. L. Arbuckle and W. Wohtke, Amos 4.0 User’s Guide. Chicago: SmallWaters, 1999.

AN EXPLORATORY STUDY OF MOBILE COMMERCE SERVICE ADOPTION 221

Dow

nloa

ded

by [

New

Yor

k U

nive

rsity

] at

00:

56 0

6 O

ctob

er 2

014

Page 21: Adoption of Mobile Internet Services: An Exploratory Study of Mobile Commerce Early Adopters

222 PEDERSEN

APPENDIXItems Used in Questionnaire

Concept Item

User friendliness Learning to use mobile commerce services is easy to meIt is easy to make the mobile commerce services do what I want them toMy interaction with mobile commerce services is clear and understandableI find it easy to interact with mobile commerce servicesI find it easy to use mobile commerce services

Usefulness Using mobile commerce services makes me save timeMobile commerce services make me a better consumerUsing mobile commerce services improves my efficiency as a consumerMobile commerce services are useful to me as a consumerMobile commerce services increases my effectiveness as a consumer

Externalinfluence

Media is full of reports, articles and news suggesting using mobile commerce services is agood idea

Media and advertising consistently recommend using mobile commerce servicesIn my profession it is advisable to use mobile commerce services

Interpersonalinfluence

Almost all of my friends use mobile commerce servicesAlmost all my colleagues thinks using mobile commerce services is a good ideaMy friends/colleagues think that we should all use mobile commerce servicesSome of my friends/colleagues recommended I should try out mobile commerce services

Self-control Generally speaking I want to do what my friends think I should doGenerally speaking I want to do what my superiors think I should doMy friends/colleagues and I use the same kinds of mobile services

Self efficacy I am able to use mobile commerce services without the help of othersI have the necessary time to make mobile commerce services useful to meI have the knowledge and skills required to use mobile commerce servicesI am able to use mobile commerce services reasonably well on my own

Facilitatingconditions

I am given the necessary support and assistance to use mobile commerce servicesI have the financial and technological resources required to use mobile commerce servicesI have access to the software, hardware and network services required to use mobile

commerce servicesThe mobile commerce services I use are well integrated and provided in a stable service

infrastructureMy service provider/operator facilitates the use of mobile commerce servicesThere are no compatibility problems related to the mobile commerce services I use

Attitude towarduse

Good/badWise/foolishFavorable/unfavorableBeneficial/harmfulPositive/negative

Subjective norm People important to me think I should use mobile commerce servicesPeople who influence my behavior think I should use mobile commerce servicesPeople whose opinion I value prefer me to use mobile commerce services

Behavioralcontrol

I feel free to use the kind of mobile services I like toUsing mobile commerce services is entirely within my controlI have the necessary means and resources to use mobile commerce services

Dow

nloa

ded

by [

New

Yor

k U

nive

rsity

] at

00:

56 0

6 O

ctob

er 2

014