the technology acceptance model and the world wide web lederer.pdf

Ž .Decision Support Systems 29 2000 269–282www.elsevier.comrlocaterdsw

The technology acceptance model and the World Wide Web

Albert L. Lederer a,), Donna J. Maupin b,1, Mark P. Sena c,2, Youlong Zhuang d,3

a C.M. Gatton College of Business and Economics, Decision Science and Information Systems, UniÕersity of Kentucky,425C Business and Economics Building, Lexington, KY 40506-0034, USA

b Fiscal Affairs, 303 Administration Building, UniÕersity of Kentucky, Lexington, KY 40506-0032, USAc Accounting and Information Systems Department, XaÕier UniÕersity, 3800 Victory Parkway, Cincinnati, OH 45207, USA

d Department of Management, UniÕersity of Missouri, Middlebush Hall, Columbia, MO 65203, USA

Accepted 27 April 2000

Abstract

Ž .The technology acceptance model TAM proposes that ease of use and usefulness predict applications usage. The currentresearch investigated TAM for work-related tasks with the World Wide Web as the application. One hundred and sixty-threesubjects responded to an e-mail survey about a Web site they access often in their jobs. The results support TAM. They also

Ž . Ž .demonstrate that 1 ease of understanding and ease of finding predict ease of use, and that 2 information quality predictsusefulness for revisited sites. In effect, the investigation applies TAM to help Web researchers, developers, and managersunderstand antecedents to users’ decisions to revisit sites relevant to their jobs. q 2000 Elsevier Science B.V. All rightsreserved.

Keywords: World Wide Web; Technology acceptance model; Decision support systems utilization

1. Introduction

The World Wide Web has grown phenomenallysince its inception in 1990. The total value of goodsand services traded over it in the US alone will reachUS$327 billion in the year 2002, an average annual

w xgrowth rate of 110% 35 . Existing organizations,start-up firms, consultants, and end users are now

) Corresponding author. Tel.: q1-857-259-2536; fax: q1-857-259-8031.

Ž .E-mail addresses: [email protected] A.L. Lederer ,Ž [email protected] D.J. Maupin , [email protected]

Ž . Ž .M.P. Sena , [email protected] Y. Zhuang .1 Tel.: q1-857-259-2310; fax: q1-857-259-5555.2 Tel.: q1-513-745-3296; fax: q1-513-745-4383.3 Tel.: q1-573-882-7374; fax: q1-573-882-0365.

investing considerable resources in it. Corporationsare building Intranets and Extranets to help themaccomplish their objectives by assisting their em-ployees in doing their jobs better. Thus, an under-standing of the predictors of Web usage could servea multitude of stakeholders by helping them recog-nize how to promote that usage.

Researchers have conducted several studies toexamine the relationship between perceived ease ofuse, perceived usefulness, attitudes, and the usage ofother information technologies in recent yearsw x1,4,6,8–10,15–17,20,30–34 . Their research has

Ž .supported the technology acceptance model TAMw x8 . TAM posits that perceived ease of use andperceived usefulness can predict attitudes towardtechnology that then can predict the usage of that

0167-9236r00r$ - see front matter q 2000 Elsevier Science B.V. All rights reserved.Ž .PII: S0167-9236 00 00076-2

( )A.L. Lederer et al.rDecision Support Systems 29 2000 269–282270

technology. Several researchers have thus validatedTAM using several different applications includingprimarily e-mail, voice mail, word processing, andspreadsheets. Other researchers have recommended

w xthe investigation of Web user behavior 28 .The first purpose of the current research was to

validate TAM with the Web as the users’ applica-tion. The second purpose was to identify antecedentsto Web ease of use and usefulness. Doing so couldidentify features of the Web that might contribute toits ease of use and usefulness. It could thus provideimplications about ease of use and usefulness forWeb developers and managers.

2. TAM: the theoretical background

w xDavis 8 has shown that TAM can explain theusage of information technology. He applied the

w xtheory of Ajzen and Fishbein 2 about reasonedaction to show that beliefs influence attitudes whichlead to intentions, and therefore generate behaviors.Davis thus conceived that TAM’s belief–attitude–in-tention–behavior relationship predicts user accep-tance of IT.

Davis asserted that perceived usefulness and easeof use represent the beliefs that lead to such accep-tance. Perceived usefulness is the degree to which aperson believes that a particular information system

Žwould enhance his or her job performance i.e., byreducing the time to accomplish a task or providing

.timely information . Perceived ease of use is thedegree to which a person believes that using a

w xparticular system would be free of effort 8 .Two other constructs in TAM are attitude towards

use and behavioral intention to use. Attitude towardsuse is the user’s evaluation of the desirability of

Fig. 2. The TAM and Web usage.

employing a particular information systems applica-tion. Behavioral intention to use is a measure of the

w xlikelihood a person will employ the application 2 .TAM’s dependent variable is actual usage. It has

typically been a self-reported measure of time orfrequency of employing the application.

Fig. 1 shows the generic TAM model. Someauthors have considered additional relationships.Some have ignored intention to use or attitudew x1,10,17,30,33 , and instead studied the effect of easeof use or usefulness directly on usage. Findingsabout the effects of attitude and intention have notalways been significant. Hence, to maintain instru-ment brevity and permit the study of the antecedentsof ease of use and usefulness, the current researchsimilarly studied the direct effect of ease of use andusefulness on usage. Fig. 2 shows the model in thecurrent study.

Such theories and models as self-efficacy theory,cost–benefit research, expectancy theory, innovationresearch, and channel disposition have supportedTAM. Table 1 summarizes several TAM studies inIS research.

Two studies have investigated TAM using theWeb as the application. One found that usefulnessand ease of use predicted usage, but that usefulness

w xhad a stronger effect 33 . Another found that ease ofw xuse predicted usage 21 . By supporting TAM, both

Fig. 1. The TAM.

( )A.L. Lederer et al.rDecision Support Systems 29 2000 269–282 271

studies suggest the importance of the antecedents tousefulness and ease of use. What makes the Webuseful and easy to use? Therefore, in addition toemploying previous measures of ease of use andusefulness, antecedents specific to the Web weresought.

3. Ease of use and usefulness on the Web

Researchers have investigated features potentiallypredictive of the perceived ease of use of the Web.

Ž .The Graphic, Visualization, and Usability GVUCenter at the Georgia Institute of Technology hasconducted Web user surveys every 6 months since

w x1994 23 . The results from the most recent surveyidentified some key ease of use problems. Mostfrequently cited was the slow speed of downloadingor viewing Web pages. Other problems includedbeing unable to perform such tasks as finding a pagethat users knew existed, organizing the pages andinformation they gathered, finding a page once vis-ited, and visualizing where they had been and couldgo to find information.

w xA qualitative study raised similar problems 19 .In the qualitative approach, respondents cited slowdata access as the issue that they disliked most aboutthe Web. They also cited difficulty searching forspecific information, information clutter, time delaysdue to images, the unreliability of sites, and incom-plete category searches.

A third study identified eight usability principles:Žspeaking the users’ language use words, phrases,

. Žconcepts familiar to the user ; consistency similar.concepts, terminology, graphics, layout, etc. ; mini-

Žmization of the user’s memory load do not force.users to recall information across documents ; flexi-

Žbility and efficiency of use accommodate a range of.user sophistication and diverse goals ; aesthetic and

Žminimalist design visually pleasing displays with no. Žirrelevant or distracting information ; chunking short

.documents with one topic ideally on a single page ;Žprogressive levels of detail organize information

hierarchically with general information before spe-. Žcific detail ; and navigational feedback allow user to

. w xdetermine document position 18 . A fourth studyw xsuggested similar issues 5 .

Much less research has considered potential pre-dictors of perceived usefulness of the Web. TheGVU survey did, however, list the most common

Ž .uses of Web users as browsing 79% , followed byŽ . Ž .entertainment 64% , work 52% , and shopping

Ž . w x11% 23 . Another survey identified the amount ofinformation on the Web as the issue most liked by

w xrespondents 19 .Usefulness measures related to the work environ-

ment were also identified as possible features of aw xWeb site. According to Griffin 12 , general informa-

tion is more abstract than information related to thetask environment. Griffin further asserted that man-agers can identify environmental factors of specificinterest to organizations more easily than the abstractdimensions of general information. He identifiedseven task-related uses of information including in-formation about competitors, customers, suppliers,government regulators, labor, company owners, andcompany relationships.

Information related to functional support withinan organization might similarly provide usefulnessaspects to a Web site. Such functions typically in-clude marketing, finance, human resources, produc-

w xtion, and research and development 12 .Four factors that differentiate between good and

bad information might also provide a basis for use-w xfulness of the Web 12 . They are accuracy, timeli-

ness, completeness, and relevance.w xFinally, Anthony 3 identified three types of

managerial decision making. They were operational,managerial, and strategic decision making. Presum-ably, information to support those decision typescould make a Web site useful.

4. Methodology

4.1. Instrument deÕelopment

The authors developed an e-mail survey instru-ment that contained instructions asking the respon-dent to identify a Web site that hershe uses often forwork and then to answer questions pertaining to thatsite. Focusing a subject on a specific site followsChurchill’s recommendation to define a unit of anal-ysis for a more precise response and greater validityw x7 .


Table 1Previous TAM researcha

Authors Constructs Applications Methodology Findings

w xDavis 8 U, EOU, Usage PROFs, XEDIT, Survey, U™usage,Chart-Master, experiment EOU™usagePendraw

w xDavis et al. 9 U, EOU, A, BI, WriteOne Experiment EOU™U, U™A,Usage EOU™A, A™BI,

U™BI, BI™Usage

w x w xHaynes and Thies 15 U, EOU, Usage Automated teller, Survey Same as Davis 8self-service gas

w x w xMathieson 20 EV, U, EOU, Spreadsheet, Experiment Same as Davis 8A, BI, Usage calculator

w xAdams et al. 1 U, EOU, Usage E-mail, V-mail, Survey EOU™Usage,WordPerfect, U™Usage EOUlU123, HarvardGraphics

w x ŽBagozzi et al. 4 U, EOU, BI two WriteOne Experiment U™BI, EOU™BI,.time intervals , BI™Usage

Usage

w xTaylor and Todd 32 U, EOU, A, Computing Survey EOU™U, U™A,Subjective norm, resource center EOU™A, A™BI,Perceived behavioral SN™BI, PBC™BI,control, BI, BI™B, PBC™BBehavior

w xStraub et al. 30 U, EOU, Usage, V-Mail Survey U™Usage,Social presencer EOU™Usageinformation SPIR™U

Ž .richness SPIR

w xIgbaria et al. 17 EV, EOU, U, Micro-computer Survey EV™EOU, EV™U,Usage EOU™U, EOU™Usage,

U™Usage

w xSzajna 31 U, EOU, BI, E-mail Experiment EOU™U, U™BI,Usage BI™Usage

Hendrickson and U, EOU, Usage 1-2-3, Experiment EOU™U, EOU™Usage,w xCollins 16 WordPerfect U™Usage

w xChau 6 EOU, Near-term U, Word, Excel Survey EOU™Near-term U,Long-term U, BI EOU™BI,

Near-term U™Long-term UNear-term U™BI,Long-term U™BI

w xMorris and Dillon 21 EOU, U, A, BI, Netscape Survey EOU™U, U™A,Usage EOU™A, U™BI,

A™BI, BI™Usage


Ž .Table 1 continued

Authors Constructs Applications Methodology Findings

w xGefen and Straub 10 Gender, U, EOU, E-Mail Survey Gender™SPIR,Usage, SPIR Gender™U,

Gender™EOU,SPIR™U, U™Usage

w xThompson 34 U, EOU, A, BI Access, Survey EOU™U, EOU™A,Web page U™A, U™BI, A™BI,development Motivation™BI,software Social factors™A

w xTeo et al. 33 U, EOU, Usage, Internet Web-based EOU™U, EOU™Usage,Perceived survey EOU™PE, U™Usage,

Ž .enjoyment PE PE™Usage

a Legend: A, attitude; BI, behavioral intention; EOU, ease of use; U, usefulness.

The survey had the following major sections.Ø Nineteen items asking the extent to which the

Web site meets ease of use characteristics. Ratingson a 1–7 scale with end points of Astrongly agreeBand Astrongly disagreeB allowed the respondent toindicate the extent. Table 2A lists the three general

w xitems of Davis 8 which were used. Sixteen wereWeb-specific antecedents condensed from Refs.w x5,18,23 . Table 2B lists 18 Web-specific measuresfrom which these 16 were drawn.

Ø Twenty-two items asking the extent to whichthe Web site meets usefulness characteristics. Theseitems also used the same 1–7 scale. Three were

w x Ž .general measures from Davis 8 see Table 2C .w x ŽSixteen were Web-specific 3,12 antecedents see

.Table 2D .Ø Two items measuring Web site usage. One

asked the extent to which the respondent used theWeb site on 1–7 scale with AfrequentlyB and Ainfre-quentlyB as anchors. The second asked the respon-dent how many times hershe used the site in thepast 30 days.

Ø Demographic questions about the respondent’sage, work experience, functional area, organizationsize, Web experience, browser, speed of connection,and Web site location.

Six professionals who used the Web in their jobsparticipated in a pilot of the survey instrument. Atleast two of the authors observed the pilot subjects asthey completed the survey. Feedback from the sub-

jects and observations by the authors resulted inminor changes to the survey instructions, changes tothe order of selected items, and refinement to thewording of several items.

4.2. Subjects

The study focused on individuals who use theWeb for their jobs. Potential subjects were selectedfrom work-related Internet newsgroups. The news-groups featured discussions of various topics, includ-ing general business, consulting, finance, law, sci-ence, and biology. The authors accessed a Web sitethat archived the newsgroup submissions to identifye-mail addresses of the participants. The addresseswere then sorted and duplicates were removed. Acomputer program submitted an electronic copy ofthe survey to each e-mail address. Completed sur-veys came from 163 subjects for a 5% response rate.

This response rate may be low in comparison toconventional paper-based postal surveys. However,the novelty of unsolicited email surveys precludes asubstantial basis of comparison. In any case, the totalnumber of subjects suffices for the analysis de-scribed below.

w xThe single method test of Harman 13 was usedw xto test for common method variance 24 . The factor

analyses produced neither a single factor nor onegeneral factor that accounted for the majority of thevariance. Each factor accounted for more than the

w xviable cut-off of 5% 14 . This test thus failed to


Table 2

Ž . w xA TAM ease of use items 8

Ø Getting the information I want from the site is easyØ Learning to use the site was easyØ Becoming skillful at using the site was easy

Ž .B Antecedent ease of use items

w xEvaluation of Web prototypes 18Ø The site uses terms familiar to meØ The site makes it easy to recognize key informationØ The site displays visually pleasing designØ Each display page focuses on a single topicØ Display pages provide links to more detailed informationØ The site provides more than one method of navigationØ I can determine my position within the siteØ The site allows easy return to previous display pagesØ The site uses consistent termsØ The site uses consistent graphics

w xWeb user survey 23Ø The site loads quicklyØ The information I need is easy to find within the siteØ The site is easy to navigate

w xUsability testing criteria 5Ø The site uses understandable graphicsØ The display pages within the site are easy to readØ The site uses understandable termsØ The information I need is easy to find within the siteØ The site is easy to navigate

Ž . w xC TAM usefulness items 8

Ø Using this site enhances my effectiveness at my jobØ Using this site in my job increases my productivityØ Using this site improves my job performance

Ž .D Antecedent usefulness items

w xCharacteristics of useful information 12Ø I use this site for accurate information for my jobØ I use this site for thorough information for my jobØ I use this site for timely information for my jobØ I use this site for relevant information for my job

w xTask environment information 12Ø I use this site for information about my company’s ownersØ I use this site for information about my company’s competitorsØ I use this site for information about my company’s suppliersØ I use this site for information about companies that

work with my companyØ I use this site for information about my company’s

customersØ I use this site for information about laborØ I use this site for information about government

regulators of my company

w xStrategic areas for corporate decisions 3Ø I use this site for strategic information for my job


Ž .D Antecedent usefulness items

w xStrategic areas for corporate decisions 3Ø I use this site for managerial information for my jobØ I use this site for operational information for my job

w xFunctional area information 12Ø I use this site for research and development informationØ I use this site for human resources informationØ I use this site for marketing informationØ I use this site for production informationØ I use this site for financial information

identify that common method variance was a prob-w xlem 11,25–27 .

5. Demographics and descriptive statistics

As Table 3A indicates, survey respondents weregenerally well educated with over 34% holding anadvanced degree and another 35% having a 4-yeardegree. Table 3A also identifies the respondents’functional work areas, browser used at work, andmethod of Internet connection.

Table 3B gives means and standard deviations forsubjects’ age, years of work in present position,years of work with present firm, years of Web usefor job, years of Web use, and number of employeesin organization. Respondents had an average age of37.4 and had used the Web for an average of over 3years. This indicates that the subjects were somewhatolder and more experienced than Internet users in the

w xgeneral population 23 .Table 4A, B, C, and D show the means and

standard deviations of the general ease of use andusefulness items and antecedents ordered by theirmean. Table 4E shows means and standard devia-tions of the usage items.

6. Data analysis

The sample of 163 subjects was first split ran-domly into two groups. Two-factor analyses wereperformed on 95 subjects. One examined the Web-specific ease of use antecedent items and the other


Table 3

Ž .A Demographic information

Percent of respondents

Education leÕelSome high school 1High school graduate 2Some college 142-year college degree 64-year college degree 35Masters 23Doctorate 13Other 7

Functional work areaAccountingrFinance 8Human resources 4Information systems 23Marketing 13Production 9Purchasing 5Sales 9Other 29

Browser used at workNetscape Navigator 71Microsoft Internet Explorer 23America Online browser 2Other 5

Internet connectionModem 61ISDN connection 7T1 connection 14Ethernet connection 12Other 6

Ž .B Demographic information

Item Mean S.D.

Age 37.4 10.3Years of work in present position 5.2 6.9Years of work with present firm 4.5 5.7Years of Web use for job 2.2 1.4Years of Web use 3.0 2.1Employees in organization 4500 24,800

analyzed the Web-specific usefulness antecedentitems. The purpose of these analyses was to reducethe number of those items, and identify the dimen-sions of the antecedents to ease of use and useful-ness. This group had 95 subjects to preserve a ratioof five subjects to each item for the usefulness items,

w xthe larger of the two sets of items 22 .

Each factor analysis used principle componentsextraction with Varimax rotation and required eigen-values of at least 1. Any item that failed to load on asingle factor at 0.5 or greater was dropped and thefactor analysis was redone. This process of droppingan item and rerunning continued until all items loadedat 0.5 or greater on one and only one factor.

Table 5A and B show the final factor structures.The authors named each factor based on their inter-pretation of its items.4

The testing of the relationships then used multipleregression on the 68 subjects in the holdback sample.

Ž .The models were see Fig. 2 :

UsagesUsefulnessqEase of use;Ease of usesEase of understandingqEase of

findingq Information focus;Usefulnesss Information for support activitiesq

Information qualityq Information for primaryactivitiesq Information for managementq

Information for research and development.

The results appear in Table 6A, B and C.Variance inflation factors did not exceed 10 for

w xany regression 29 . In fact, they were less than 2.Hence, multicollinearity was not extensive.

7. Summary of findings

ŽThis research provided support for TAM see.Table 6A . With usage measured by the 1–7 fre-

quency scale, the effect of usefulness and ease of useŽ 2 .was significant p-0.001 and R s0.15 . Useful-

Ž .ness p-0.01 had a stronger effect than ease ofŽ .use p-0.05 .

The research provided weak support for TAMwhere usage was measured by the number of times

Ž 2 .used in the past 30 days p-0.10 and R s0.04 .Ž .The effect of usefulness p-0.10 was weak. The

4 Among the Usefulness categories was Useful Information forResearch and Development. It had a single item. Although re-searchers sometimes drop such single-item factors, in this case,the authors chose to keep it to maintain the richness of thecategories. They also conducted the statistical tests described laterin this paper without this category and found very similar results.


Table 4

Ž .A Descriptive statistics for TAM ease of use items

TAM ease of use items Mean S.D.

Learning to use the site was easy 5.61 1.38Becoming skillful at using the site was easy 5.45 1.45Getting the information I want from the site is easy 5.40 1.39

Ž .B Descriptive statistics for antecedent ease of use items

Antecedent ease of use items Mean S.D.

The site uses terms familiar to me 5.77 1.54The site uses consistent terms 5.72 1.28The display pages within the site are easy to read 5.66 1.24The site uses understandable terms 5.64 1.41The site uses consistent graphics 5.60 1.45Display pages provide links to more detailed information 5.54 1.59The site uses understandable graphics 5.44 1.48The site is easy to navigate 5.33 1.44The site allows easy return to previous display pages 5.28 1.67The site makes it easy to recognize key information 5.19 1.52The information I need is easy to find within the site 5.01 1.56I can determine my position within the site 4.71 1.84The site loads quickly 4.71 1.73The site displays visually pleasing design 4.70 1.48Each display page focuses on a single topic 4.57 1.91The site provides more than one method of navigation 4.40 1.83

Ž .C Descriptive statistics for usefulness items

TAM usefulness items Mean S. D.

Using this site improves my job performance 5.80 1.37Using this site enhances my effectiveness at my job 5.75 1.36Using this site in my job increases my productivity 5.71 1.32

Ž .D Descriptive statistics for antecedent usefulness items

Antecedent usefulness items Mean S.D.

I use this site for relevant information for my job 5.99 1.23I use this site for accurate information for my job 5.89 1.39I use this site for timely information for my job 5.81 1.35I use this site for thorough information for my job 5.41 1.59I use this site for strategic information for my job 5.16 1.81I use this site for research and development information 4.90 2.12I use this site for operational information for my job 4.44 1.99I use this site for managerial information for my job 3.59 2.05I use this site for marketing information 3.58 2.41I use this site for production information 3.42 2.40I use this site for information about my company’s customers 2.95 2.22I use this site for information about companies that work with my company 2.90 2.16I use this site for financial information 2.78 2.17I use this site for information about my company’s competitors 2.72 2.13I use this site for information about my company’s suppliers 2.72 2.07I use this site for information about government regulators of my company 2.49 2.06I use this site for human resources information 2.36 2.02I use this site for information about labor 2.13 1.71I use this site for information about my company’s owners 1.75 1.56



Ž .E Descriptive statistics for usage items

Usage items Mean S.D.

How frequently did you use this site in the past 30 days? 5.58 1.50How many times did you use the site in the past 30 days? 20.80 20.75

stronger effect of usefulness than ease of use isw xconsistent with previous Web research 33 .

The research also provided some understanding ofŽ .ease of use see Table 6B . The antecedents pre-

Ž 2 .dicted ease of use p-0.01 and R s0.50 withŽ .ease of understanding p-0.01 having a stronger

Ž .effect than ease of finding p-0.05 .The research also provided some understanding of

Ž .usefulness see Table 6C . The antecedents predictedŽ 2 .usefulness p-0.01 and R s0.58 , but only infor-

Ž .mation quality had a significant effect p-0.01 .

8. Implications for researchers

This study supports TAM. It thus helps re-searchers understand the relationships between easeof use and usefulness, and the acceptance of Webtechnology by users. It confirms that use of Websites depends on the usefulness and ease of use ofthe site. It also helps us understand the predictors ofusefulness and ease of use for the Web.

The study provides two new instruments tailoredto the Web. On one hand, future researchers coulduse these instruments for assessing the ease of useand usefulness of Web sites. On the other hand,these two instruments could stimulate future re-searchers to develop better instruments for assessingthose characteristics of Web sites. Alternative word-ing of the items might be tried. With further refine-ment of the Web-specific items, greater varianceexplained might be achieved.

In this research, the highest predictive power be-Ž .longed to information quality for usefulness and

Ž .ease of understanding for ease of use . Perhaps theformer occurred because its individual items weremore general to all users, whereas the others haditems more specific to individual’s jobs. Perhapsease of understanding had higher predictive powerthan ease of finding because users more easily adjust

to difficulties navigating through frequently usedWeb sites. Nevertheless, future researchers mightempirically investigate why these factors had thehighest predictive power for their respective con-structs. Future researchers might also investigate howto improve these apparently important factors inWeb site design.

This research examined frequently visited sites. Itthus facilitates deduction about the specific ease ofuse and usefulness characteristics of sites that moti-vate revisiting. However, future researchers mightconsider sites that users do not revisit. Data contrast-ing more often and less often visited sites mightfurther help explain why some sites are used morefrequently. Future research could also ask subjects torespond in general about their impressions of theease of use, the usefulness, and their own usage ofthe Web.

One limitation of the current research is the as-sumption that work usage is approved and construc-tive rather than games or chatting. None of thesubjects in the current research responded about agame or chat site. Nevertheless, future research couldconsider predictors of games, chatting and otherpotentially detrimental activities.

Although Harman’s single method test did notidentify common method variance as a problem, itstill might have been. To ensure that it is not aproblem and to prevent the consistency effect result-ing from the same subject reporting both indepen-dent and dependent variables, future research mightuse more objective measures of the dependent vari-able. Software for monitoring precise usage wouldprovide such an objective measure.

Factor analysis is a popular and useful tool for thereduction of data and the identification of key themesin the data. However, because items do not load at agiven arbitrary level, they might still be relevant.Hence, future research should replicate this study.Perhaps other constructs would emerge.


Table 5

Ž .A Factor analysis of antecedent ease of use items

Factors and items Factor loadings

F1 F2 F3

Factor 1: Ease of understandingThe site uses understandable graphics. 0.89The site uses consistent graphics. 0.86The site uses consistent terms. 0.81The site uses understandable terms. 0.76Display pages provide links to more detailed information. 0.67The site displays visually pleasing design. 0.60The display pages within the site are easy to read. 0.59

Factor 2: Ease of findingThe site allows easy return to previous display pages. 0.80I can determine my position within the site. 0.71The site is easy to navigate. 0.71

Factor 3: Information focusEach display page focuses on a single topic. 0.90The site makes it easy to recognize key information. 0.62

Eigenvalues 5.17 1.43 1.07Percent of variance explained 43.2 11.9 8.9

aCronbach’s alpha 0.88 0.70 0.46

Ž .B Factor analysis of antecedent usefulness items

Factors and items Factor loadings

F1 F2 F3 F4 F5

Factor 1: Information for support actiÕitiesI use this site for information about my company’s competitors. 0.77I use this site for information about labor. 0.71I use this site for information about my company’s suppliers. 0.70I use this site for information about my company’s customers. 0.70I use this site for information about companies that work with my company. 0.66I use this site for human resources information. 0.65I use this site for information about government regulators of my company. 0.58

Factor 2: Information qualityI use this site for relevant information for my job. 0.83I use this site for accurate information for my job. 0.82I use this site for timely information for my job. 0.73I use this site for thorough information for my job. 0.65

Factor 3: Information for primary actiÕitiesI use this site for marketing information. 0.81I use this site for production information. 0.77I use this site for financial information. 0.77

Factor 4: Information for managementI use this site for operational information for my job. 0.84I use this site for managerial information for my job. 0.75I use this site for strategic information for my job. 0.61



Ž .B Factor analysis of antecedent usefulness items

Factor 5: Information for research and deÕelopment

I use this site for research and development information. 0.85Eigenvalues 5.18 3.52 1.43 1.13 1.05Percent of variance explained 28.8 19.6 7.9 6.3 5.8Cronbach’s alpha 0.83 0.81 0.79 0.77 N.A.

a w xCronbach’s alpha in this research was 0.70 or greater for every factor except Information focus 22 . However, Information focus hadŽ .only two items and hence, alpha is not so meaningful. Also, the two items correlated significantly p-0.001 . Hence the factor remained in

the analysis.

Table 6

Ž .A Two multiple regressions: UsagesEase of useqUsefulnessaUsage

Scale of 1–7 Number of times

Coefficients p- values Coefficients p- values) )Ease of use 0.25 0.02 1.19) ) ) )Usefulness 0.30 0.01 2.46 0.08

2R 0.15 0.04a )F 13.08 0.001 2.85 0.06

Ž .B Multiple regression: Ease of usesEase of understandingqEase of findingq Information focus

Coefficients‡Factor 1: Ease of understanding 0.46†Factor 2: Ease of finding 0.20

Factor 3: Information focus 0.152R 0.50

‡F 21.87

Ž .C Multiple regression: Usefulnesss Information for support activitiesq Information qualityq Information for primary activitiesqInformation for managementq Information for research and development

Coefficients

Factor 1: Information for support activities y0.01£Factor 2: Information quality 0.83

Factor 3: Information for primary activities y0.04Factor 4: Information for management 0.06Factor 5: Information for research and development 0.08

2R 0.58£F 16.14

a The p- values appear in this table because two of them are AcloseB to more commonly accepted cut-off values in social sciences, i.e.,Ž . Ž .Ease of use for Scale of 1–7 0.02 is AcloseB to 0.01 and F for Number of times 0.06 is AcloseB to 0.05.

) 0.10 significance level.)) 0.05 significance level.))) 0.01 significance level.a0.001 significance level.‡ 0.01 significance level.†0.05 significance level.£ 0.01 significance level.


The response rate of 5% may be a limitation inthis study. Little is known about e-mail surveys andthe Internet may motivate more of them due to theirlow cost. In fact, the future may bring a growingpopularity of even the simple posting of a Websurvey where no response rate can be calculated.One recent TAM study advertised such a site and

w xthus could not calculate a rate 33 . More needs to beunderstood about e-mail and Web surveys.

While the current research examined why someusers access Web sites more than others do, addi-tional research could consider why some people stilldo not use it at all in their jobs. TAM could not bethe theoretical basis for such research because itassumes subjects can assess ease of use and useful-ness. Nevertheless, such research could be useful.

Finally, most of the respondents in this studywere highly educated and experienced at using theInternet. Investigating ease of use and usefulnessmeasures with less educated and more inexperiencedWeb users may provide additional validation of TAMand interesting insights about ease of use and useful-ness. As increasing numbers of workers use the Webin their jobs, findings about such users could proveuseful to both employers and Web site developers.

9. Implications for practitioners

This research has potential for practical applica-tion in the development and use of Web sites. Byconfirming TAM, it suggests that Web site develop-ers should provide ease of use and usefulness fortheir Web sites to encourage people to revisit theirsites. It also suggests both specific factors and itemsthat those developers might emphasize when theycreate new Web sites. For example, it suggests thatinformation quality — i.e., relevance, accuracy,timeliness and thoroughness of information — maybe more important than the various other more spe-cific information uses in this study. Also, it suggeststhat ease of understanding may be more importantthan ease of finding it in the decision to revisit.

The research has also provided two instrumentsthat could be useful to both Web site developers andWeb site managers in organizations that encourage

Ž .employees to use specific especially Intranet Websites. Those developers and managers could haveusers complete the instruments about specific sites.

The responses could be used to identify strengthsand weaknesses in existing sites. Developers andmanagers could investigate the factors and itemswith lower scores. The responses might thus beuseful in improving those sites.

In fact, normative data about many could beŽaccumulated using these or future, improved ver-

.sions of these instruments. Comparisons of scoresfor individual sites to such data could help develop-ers and managers assess their sites. Comparisonscould also stimulate competition among Web sitedevelopers and thus foster the improvement of suchsites.

10. Conclusion

This research has validated TAM in the context ofthe World Wide Web. It has also contributed byapplying TAM to lay the groundwork for under-standing antecedents to ease of use and usefulness.Such antecedents might effect Web usage. An under-standing of them could guide both Web site researchand development.

References

w x1 D.A. Adams, R.R. Nelson, P.A. Todd, Perceived usefulness,ease of use, and usage of information technology: a replica-

Ž . Ž .tion, MIS Quarterly 16 2 1992 227–247, June.w x2 I. Ajzen, M. Fishbein, Understanding Attitudes and Predict-

ing Social Behavior, Prentice-Hall, Englewood Cliffs, NJ,1980.

w x3 R.N. Anthony, Planning and Control Systems: A Frameworkfor Analysis, Harvard University Press, Cambridge, MA,1965.

w x4 R.P. Bagozzi, F.D. Davis, P.R. Warshaw, Development andtest of a theory of technological learning and usage, Human

Ž . Ž .Relations 45 7 1992 659–686, July.w x5 E. Boling, Usability testing for Web sites, Learning for the

Global Community: Seventh Annual Hypermedia ’95Conference, URL s http:rrwww.indiana.edur ; iirgrARTICLESrusabilityrusability.main.html, February 22,1997.

w x6 P.Y.K. Chau, An empirical assessment of a modified tech-nology acceptance model, Journal of Management Informa-

Ž . Ž .tion Systems 13 2 1996 185–204, Fall.w x7 G.A. Churchill, A paradigm for developing better measures

of marketing constructs, Journal of Marketing Research 16Ž .1979 64–73, February.

w x8 F.D. Davis, Perceived usefulness, perceived ease of use, anduser acceptance of information technology, MIS Quarterly 13Ž . Ž .3 1989 319–339.


w x9 F.D. Davis, R.P. Bagozzi, P.R. Warshaw, User acceptance ofcomputer technology: a comparison of two theoretical mod-

Ž . Ž .els, Management Science 35 8 1989 982–1003, August.w x10 D. Gefen, D.W. Straub, Gender differences in the perception

and use of e-mail: an extension to the technology acceptanceŽ . Ž .model, MIS Quarterly 21 4 1997 389–400, December.

w x11 C.N. Greene, D.W. Organ, An evaluation of causal modelslinking the perceived role with job satisfaction, Administra-

Ž .tive Science Quarterly 18 1973 95–103.w x12 R.W. Griffin, Management, 3rd edn., Houghton Mifflin,

Boston, MA, 1990.w x13 H.H. Harman, Modern Factor Analysis, University of Chicago

Press, Chicago, IL, 1976.w x14 L. Hatcher, A Step-by-Step Approach to Using the SAS

System for Factor Analysis and Structural Equation Model-ing, SAS Institute, Cary, NC, 1994.

w x15 R.M. Haynes, E.A. Thies, Management of technology inŽ .service firms, Journal of Operations Management 10 3

Ž .1991 388–397, August.w x16 A.R. Hendrickson, M.R. Collins, An assessment of structure

and causation of IS usage, The DATA BASE for AdvancesŽ . Ž .in Information Systems 27 2 1996 61–67, Spring.

w x17 M. Igbaria, T. Guimaraes, G.B. Davis, Testing the determi-nants of microcomputer usage via a structural equation model,

Ž . Ž .Journal of Management Information Systems 11 4 199587–114, Spring.

w x18 M.D. Levi, F.G. Conrad, A heuristic evaluation of a WorldŽ . Ž .Wide Web Prototype, Interactions 3 4 1996 50–61,

July–August.w x19 N. Lightner, I. Bose, G. Salvendy, What is wrong with the

World Wide Web? A diagnosis of some problems and pre-Ž . Ž .scription of some remedies, Ergonomics 39 8 1996 995–

1004.w x20 K. Mathieson, Predicting user intentions: comparing the tech-

nology acceptance model with the theory of planned behav-Ž . Ž .ior, Information Systems Research 2 3 1991 173–191,

September.w x21 M.G. Morris, A. Dillon, How user perceptions influence

Ž .software use, decision support systems, IEEE Software 199758–65, July–August.

w x22 J.C. Nunnally, I.H. Bernstein, Psychometric Theory, Mc-Graw-Hill, New York, 1994.

w x23 J.E. Pitkow, C.M. Kehoe, Emerging trends in the WWW userŽ . Ž .population, Communications of the ACM 39 6 1996

106–108, June.w x24 P.M. Podsakoff, D.W. Organ, Self-reports in organizational

research: problems and prospects, Journal of Management 12Ž . Ž .4 1986 531–544.

w x25 P.M. Podsakoff, W.D. Todor, R.A. Grover, V.L. Huber,Situational moderators of leader reward and punishment be-haviors: fact or fiction? Organizational Behavior and Human

Ž .Performance 34 1984 21–63.w x26 C.A. Schreisheim, The similarity of individual-directed and

group-directed leader behavior description, Academy ofŽ .Management Journal 22 1979 345–355.

w x27 J.F. Schreisheim, The social context of leader–subordinaterelations: an investigation of the effects of group cohesive-

Ž .ness, Journal of Applied Psychology 65 1980 183–194.

w x28 M.J. Shaw, D.M. Gardner, H. Thomas, Research opportuni-ties in electronic commerce, Decision Support Systems 21Ž .1997 149–156.

w x29 J. Stevens, Applied Multivariate Statistics for the SocialSciences, Lawrence Erlbaum Associates, Mahwah, NJ, 1996.

w x30 D. Straub, M. Limayem, E. Karahanna-Evaristo, Measuringsystem usage: implications for IS theory testing, Manage-

Ž . Ž .ment Science 41 8 1995 1328–1342, August.w x31 B. Szajna, Empirical evaluation of the revised technology

Ž . Ž .acceptance model, Management Science 42 1 1996 85–92,January.

w x32 S. Taylor, P. Todd, Assessing IT usage: the role of priorŽ . Ž .experience, MIS Quarterly 19 4 1995 561–570, Decem-

ber.w x33 T.S.H. Teo, V.K.G. Lim, R.Y.C. Lai, Intrinsic and extrinsic

Ž .motivation in internet usage, Omega 27 1999 25–37.w x34 R. Thompson, Extending the technology acceptance model

with motivation and social factors, Proceedings of Associa-Ž .tion for Information Systems Annual Conference 1998

757–759, August.w x35 H. Yang, R.M. Mason, The internet, value chain visibility,

and learning, Proceedings of the 31st Annual Hawaii Interna-Ž .tional Conference on System Sciences 1998 23–32.

Albert L. Lederer is Professor of man-agement information systems in theCarol M. Gatton College of Businessand Economics at the University ofKentucky. He holds a BA in psychologyfrom the University of Cincinnati, anMS in computer and information sci-ences from the Ohio State University,and a PhD in industrial and systemsengineering from Ohio State. His re-search has appeared in the Journal ofOrganizational Computing and Elec-

tronic Commerce, Communications of the ACM, Journal of Man-agement Information Systems, MIS Quarterly, and elsewhere. Hismajor research area is information systems planning.

Donna J. Maupin is the Web Managerfor the University of Kentucky FiscalAffairs Division. She holds a BS inretail marketing from the University ofKentucky and an MBA from the Univer-sity of Kentucky, and is completing herPhD in decision sciences and informa-tion systems in the Gatton College ofBusiness and Economics at the Univer-sity of Kentucky. Her research has ap-peared in the Journal of Computer In-formation Systems and the Journal of

Small Business Strategy. She has presented her work at theAmerica’s Conference on Information Systems, Association forComputing Machinery Special Interest Group on Computer Per-sonnel Research Conference, and Informs National Meeting. Herresearch interests include electronic commerce, strategic planning,and information management.


Mark Sena is an Assistant Professor ininformation systems at Xavier Univer-

Ž .sity OH . He holds a BBA in businessanalysis from Texas A&M University,and an MBA from Miami UniversityŽ .OH , and is completing his PhD indecision sciences and information sys-tems in the Gatton College of Businessand Economics at the University ofKentucky. His research credits includean article in the Journal of InformationTechnology and Management. He has

presented his work at the America’s Conference on InformationSystems, Association for Computing Machinery Special InterestGroup on Computer Personnel Research Conference, InformsNational Meeting, and Summer Computer Simulation Conference.His research interests focus on electronic commerce, decisionsupport systems, and enterprise systems.

Youlong Zhuang is an Assistant Profes-sor in management information systemsin the College of Business at the Univer-sity of Missouri, Columbia. He holds aBS in systems engineering from Shang-hai University of Science and Technol-

Ž .ogy China an MBA from Indiana StateUniversity, and a PhD in decision sci-ences and information systems from theGatton College of Business and Eco-nomics at the University of Kentucky.He has presented his work at the Annual

Meeting of the Decision Sciences Institute, the America’s Confer-ence on Information Systems, Association for Computing Machin-ery Special Interest Group on Computer Personnel Research Con-ference, and Informs National Meeting. His major research area iselectronic commerce.

the technology acceptance model and the world wide web lederer.pdf

Documents