Xu, Jing and Wang, J. (2008) Two-stage estimation of inequality-constrained marginal linear models with longitudinal data. Journal of Statistical Planning and Inference 138 (6), pp. 1905-1918. ISSN 03783758.
Text (Pre-print (unrefereed))
Restricted to Repository staff only
Inequality-constrained regression models have received increased attention in longitudinal analysis during recent years. Regression parameters are usually obtained from iteration algorithms. An analytical formulae of the estimators cannot be provided. Therefore, the asymptotic behavior of estimators has not been fully clarified yet. This paper presents a TS estimation (TS for short) and the asymptotic distribution of the estimators. Simulations are conducted to compare constrained TS estimation, constrained ordinary least squares (OLS) estimation and TS estimation in terms of sample bias, sample mean-square error (MSE) and sample variance of the estimators.
|Keyword(s) / Subject(s):||Longitudinal data, inequality constraints, correlation, two-stage estimate, asymptotic distribution|
|School or Research Centre:||Birkbeck Schools and Research Centres > School of Business, Economics & Informatics > Economics, Mathematics and Statistics|
|Date Deposited:||08 Dec 2010 12:15|
|Last Modified:||17 Apr 2013 12:33|
Archive Staff Only (login required)