AN ANALYSIS OF METHODS USED TO EXAMINE GENDER DIFFERENCES IN COMPUTER-RELATED BEHAVIOR

A comprehensive review of the literature of gender differences in computer-related behavior reveals a myriad of conflicting results. A critical analysis of the empirical methods used to collect data is offered as one means of sorting out the numerous inconsistencies found. Nine areas are discussed where common mistakes are made, including 1) sample selection, 2) sample size, 3) scale development, 4) scale quality, 5) the use of univariate and multivariate analyses, 6) regression analysis, 7) construct definition, 8) construct testing, and 9) the presentation of results. It is concluded that although a number of researchers have fallen prey to making these mistakes, in most cases, easy remedies are available. Above all, researchers are encouraged to present more detailed and clear information regarding methods and results.