Ordinal logistic regression under complex sampling: factors associated with the nutritional status of Chilean children, 2011

Angarita-Fonseca, A. and Roldan, C. and Navarro, C.
Master of Science (M.Sc.) in Statistics. Thesis. Universidad de Valparaiso, 2014

In this dissertation, the ordinal logistic regression under complex sampling was used as a method to model an ordinal outcome in a cross-sectional design. The database of the 2011 Chilean Socioeconomic Characterization Survey [CASEN] was analyzed in order to establish the factors associated with nutritional status in children under seven years. Firstly, standard errors for the percentage of each category of nutritional status were compared under simple random sampling and under complex random sampling. Univariate and multivariable ordinal logistic models were fitted to establish the association between the independent variables studied and the nutritional status of boys and girls. In the multivariable analysis, the final model included the following variables: total household income, education of the head of the household, global quality index of the home, child education, and social aid for food.

Leave a Reply

Your email address will not be published. Required fields are marked *