Extending the Technology Acceptance Model with Motivation and Social Factors

摘要: The Technology Acceptance Model (TAM) proposed by Davis (1986) has been tested in a number of studies, and received considerable support. model is parsimonious, easy to understand, provides reasonable explanatory value under variety conditions. As part larger study, two factors (Motivation complete the task Social Factors) were added TAM model, with data samples. In first test, Motivation Factors exerted much stronger direct influences on Intentions than constructs, amount variance explained increased significantly (26 % alone; 47% added). alone over 40% Intentions. second results less dramatic. Differences between samples are used interpret observed results. Specifically, context where respondents had no choice selection technology, played very strong influencing When tools was offered, Attitude toward using tool strongest influence Introduction numerous studies (e.g., et al., 1989; Mathieson, 1991; Taylor Todd, 1995), each case shown explain use technology and/or actual technology. From these it would seem that we could conclude primary components (beliefs about Ease Use Perceived Usefulness, Attitudes technology) do fact provide parsimonious way predicting Use. study reported here, test made this conclusion. sample 63 undergraduate students their intentions Access database management system (DBMS) course project. involved 82 students, HomeSite Web development software package Test Partial Least Squares (PLS) statistical analysis technique (Barclay 1995) context. Each construct measured minimum three questionnaire items, exception construct. one item (’my instructor/boss supports my computers’). maximum measures seven, for Expertise. A t-test differences conducted; Table 1 shows Factors, Two more positive (Access) second. factor loadings all items ranged from .54 .95. Fornell Larcker measure internal consistency (reliability) calculated construct, .74 .94 (this can be interpreted similarly Cronbach’s alpha). Discriminant validity also tested, found exhibit satisfactory reliability discriminant validity. Figure displays TAM. path coefficients R enclosed boxes those obtained (HomeSite) sample. seen 26% Access, 29% HomeSite. generally direction hypothesized. Usefulness sample, which but not substantive (.10). constructs as possible (Figure 2), (Access sample) 35% (HomeSite sample). Also, paths still positive, substantive. decreased .25 .18 Intentions, .34 .14 (in