Causal Structure, Endogeneity, and the Missing Data Problem in Modeling the Impact of Information and Communication Technology (E-Government) Use:? Theoretical and Methodological Challenges
2014 47th Hawaii International Conference on System Sciences (2008)
Waikoloa, Big Island, Hawaii
Jan. 7, 2008 to Jan. 10, 2008
DOI Bookmark: http://doi.ieeecomputersociety.org/10.1109/HICSS.2008.88
Despite the abundance of empirical research on the impact of information and communication technology, their relationship still remains partially answered because of conflicting results. Empirical research reports positive, negative, and negligible effects depending on data and methods employed. This puzzling circumstance results largely from the lack of rich data and sophisticated knowledge and skills. This paper reviews data analysis methods frequently used in the literature and then discusses key modeling issues, such as causal structure, endogeneity, and the missing data problem, which traditional methods rarely address. In order to deal with those issues, the propensity score matching, treatment effect model, and recursive bivariate probit model are suggested as alternatives. These methods do not replace but supplement traditional approaches. This paper concludes with the emphasis on careful examination of the characteristics of dependent variables and prudent consideration of the key modeling issues.
Hun Myoung Park, "Causal Structure, Endogeneity, and the Missing Data Problem in Modeling the Impact of Information and Communication Technology (E-Government) Use:? Theoretical and Methodological Challenges", 2014 47th Hawaii International Conference on System Sciences, vol. 00, no. , pp. 196, 2008, doi:10.1109/HICSS.2008.88