Quasi‐least squares fitting. (8th April 2015)

• Record Type: Journal Article
• Authors: Shults, Justine
• Is Part Of: Wiley interdisciplinary reviews. Volume 7:Number 3(2015)
• Page Start: 194
• Page End: 204
• Publication Date: 2015-04-08
• Subjects: binary -- clustered -- correlated -- GEE -- longitudinal -- QLS -- semi‐parametric; Mathematical statistics -- Data processing -- Periodicals; Science -- Data processing -- Periodicals; Social sciences -- Data processing -- Periodicals; Mathematical statistics -- Periodicals; 519.50285
• DOI: 10.1002/wics.1349
• Languages: English
• ISSNs: 1939-5108

Abstract

Abstract : Quasi‐least squares (QLS) is a computational method for estimation of the correlation parameters in the framework of generalized estimating equations (GEE) [Liang and Zeger. Biometrika 1986, 73:13–22]. QLS is a complementary approach to GEE, that might be considered if convergence is an issue, or if the implementation of a plausible model for the correlations is not straightforward for GEE. Software for QLS is available in SAS, Stata, MATLAB, and R. In addition, a comprehensive review of the approach [Shults and Hilbe. Quasi‐Least Squares Regression . Boca Raton, FL: Chapman & Hall/CRC; 2014] was recently published in the highly regarded Chapman & Hall/CRC Statistical Monograph & Applied Probability Monograph Series. WIREs Comput Stat 2015, 7:194–204. doi: 10.1002/wics.1349 This article is categorized under: Statistical Models > Fitting Models Algorithms and Computational Methods > Least Squares