ABSTRACT
This chapter explains how to create centile curves using gamlss. It examines the lambda, mu, sigma (LMS) method of centile estimation; the different functions for centile estimation within gamlss; and how to use the functions effectively. The chapter is important for practitioners involved in centile estimation since Generalized additive models for location scale and shape (GAMLSS) has become one of the standard tools for creating centile curves. Centile estimation includes methods for estimating the age-related distribution of human growth. The standard estimation of centile curves usually involves two continuous variables: includes the response variable, and the explanatory variable. Erratic centile curves may indicate the need to increase the values of the smoothing parameters. The chapter describes the use of quantile sheets regression for constructing centile curves. Quantile sheets were developed by S. K. Schnabel and P. H. C. Eilers in order to overcome some of the problems associated with quantile regression.
