Graphical Chain Models and their Application
Graphical models are a powerful tool to analyze multivariate data sets that allow to reveal direct and indirect relationships and to visualize the association structure in a graph. As with any statistical analysis, however, the obtained results partly reflect the uncertainty being inherent in any type of data and depend on the selected variables to be included in the analysis, the coding of these variables and the selection strategy used to fit the graphical models to the data. This paper suggests that these issues may be even more crucial for graphical models than for simple regression analyses due to the large number of variables considered which means that a fitted graphical model has to be interpreted with caution. Sensitivity analyses might be recommended to assess the stability of the obtained results. This will be illustrated using a data set on undernutrition in Benin.
KeywordsGraphical Model Markov Property House Quality Association Structure Graphical Chain
Unable to display preview. Download preview PDF.
- Foraita, R. & Sobotka, F. (2008). gmvalid: Validation of graphical models. R package version 1.2.Google Scholar
- Friedman, N., Goldszmidt, M. & Wyner, A. (1999a). Data analysis with bayesian networks: a bootstrap approach.Google Scholar
- Friedman, N., Goldszmidt, M. & Wyner, A. (1999b). On the application of the bootstrap for computing confidence measures on features of induced bayesian networks.Google Scholar
- Gorstein, J., Sullivan, K., Yip, R., de Onis, M., Trowbridge, F., Fajans, P. & Clugston, G. (1994). Issues in the assessment of nutritional status using anthropometry., Bull World Health Organ 72: 273–283.Google Scholar
- Klasen, S. (2003). Malnourished and surviving in South Asia, better nourished and dying young in Africa: what can explain this puzzle?, Measurement and Assessment of Food Deprivation and Undernutrition, FAO, Rome, pp. 283–287.Google Scholar
- Lauritzen, S. L. (1996). Graphical Models., Clarendon Press, Oxford.Google Scholar
- Macro (1996). MEASURE DHS datasets Bangladesh, Benin. www.measuredhs.com, last accessed: 10. Aug 2009.
- Steck, H. & Jaakkola, T. S. (2004). Bias-corrected bootstrap and model uncertainty, in S. Thrun, L. Saul & B. Schölkopf (eds), Advances in Neural Information Processing Systems 16, MIT Press, Cambridge, MA.Google Scholar
- UNICEF (1998). The State of the World’s Children: Focus on Nutrition., UNICEF, New York.Google Scholar
- WHO (1995). Physical status: The use and interpretation of anthropometry, WHO Technical Report Series 854, WHO, Geneva.Google Scholar