A multi-group analysis of structural invariance: an illustration using the technology acceptance model

  • Authors:

  • Xiaodong Deng;William J. Doll;Anthony R. Hendrickson;Joseph A. Scazzero

  • Affiliations:

  • Management Information Systems, Department of Decision and Information Sciences, School of Business Administration, Oakland University, Rochester, MI 48309, USA;MIS and Strategic Management, The University of Toledo, 2801 West Bancroft St., Toledo, OH 43606, USA;Management Information Systems, Department of Logistics, Operations, and Management Information Systems, College of Business, Iowa State University, 310 Carver Hall, Ames, IA 50011, USA;Department of Accounting and Finance, College of Business, Eastern Michigan University, 406 Gary M. Owen Building, Ypsilanti, MI 48197, USA

  • Venue:
  • Information and Management
  • Year:
  • 2005

Quantified Score

Hi-index 0.00

Visualization

Abstract

Selecting the appropriate mix of functional and/or interface characteristics to achieve user acceptance has proven to be a more challenging and difficult decision than expected. While numerous studies have shown that the technology acceptance model (TAM) is useful for predicting acceptance, estimates of its structural weights are not consistent across studies. Using initial exposure data from 742 users of office suite applications (i.e., spreadsheet, database, word processing, and graphics), our research illustrated the use of multi-group analysis of structural invariance (MASI) to test differences in structural weights across population subgroups for latent variables in TAM. We argue that, for large sample studies containing latent variables, MASI may be a more appropriate test of differences for structural weights/regression coefficients than analysis of covariance. The managerial implications of the results in setting functionality and interface goals and allocating resources to continued development efforts are discussed.