engleski

Decision tree analysis of the predictors of Internet affinity

A recently developed model of Internet affinity was used for survey design and data collection on variables that have potential influence on affinity for Internet use. A total of 600 Croatian students with access to the Internet at their colledge participated in this survey. The collected data were used for investigation of the relation between decision tree analisys and regression analisys of predictor variables of Internet affinity. Different predictors were found to influence two distinct criteria of Internet affinity (1) frequency of Internet use and (2) desire to use the Internet. Decision tree analisys and regression analisys manifested much similarity in this uncovered lists of variables that affected the criteria. However, the results of decision tree analisys were more informative since they provided greater insight into the structural relations of important predictor variables and also into specific subgroups of subjects in the survey.

Internet affinity; survey; decision tree analisys; regression analisys

nije evidentirano