Nonparametric Estimation for a Slope of a Replicated Linear Functional Relationship Model

Abstract

In this paper, we propose a nonparametric method to estimate the slope of a replicated linear functional relationship model. The nonparametric method is a robust method in nature and does not affect when the observations have outliers. Additionally, the nonparametric method does not require the normality assumption. Using simulation studies, we compared the performance of the proposed nonparametric method with the traditional method using maximum likelihood estimation. It is found that without any outlier, the maximum likelihood estimation works well but when outliers exist in the data, our proposed nonparametric method gives a small mean square error, thus suggesting a better estimate. Keywords: linear functional relationship model; nonparametric method; outliers