Please use this identifier to cite or link to this item: http://hdl.handle.net/10419/180320
Authors: 
Carson, Scott A.
McDonald, James B.
Year of Publication: 
2018
Series/Report no.: 
CESifo Working Paper No. 7058
Abstract: 
Assumptions about explanatory variables and errors are central in regression analysis. For example, the well-known method of ordinary least squares yields consistent and efficient estimators if the underlying error terms are independently, identically, and normally distributed. Additionally, the conditional distribution of the dependent variable is symmetric. The modern obesity epidemic is a well-known health dilemma where the BMI distribution was initially positively skewed but has become more symmetric, which may affect inferences about health and public resource allocation. This study applies partially adaptive estimation methods with flexible error distributions to account for possible skewness and leptokurtosis in the distribution of BMI.
Subjects: 
obesity epidemic
partially adaptive estimation
skewed generalized T distribution
JEL: 
C10
D13
J13
Document Type: 
Working Paper

Files in This Item:
File
Size





Items in EconStor are protected by copyright, with all rights reserved, unless otherwise indicated.