Aplikasi Splus dalam Model Regresi dengan kesalahan Pengukuran Menggunakan Metode ICM

Abstract

Measurement data are often assumed to be without error is very rare indeed. Generally, any data obtained would contain an error due to measurement / factor measurement error (ME). For example in the regression modeling if X is a random  variable or with measurement error. Then the complex calculations will not be separated from the role of computers and technology. As well as to estimate the following model, given the data (Xi, Yi), the regression models are Yi = g(X i ) + ε i

where Xi is the i-th element of the predictor variable X and Yi is the i-th element of the response variable Y. Variable X which is a predictor of the outcome variable is usually
a certain constant observation, however, X is generally found that a random variable or variables where the value is not a fixed constant. For that reason in this case is called
the regression models with regression models with measurement error. There are two methods, namely parametric and Nonparametric Approach. Parametric approach used Ordinary Least Square (OLS), and for Nonparametric measurement errors are ignored when used B-spline method and the measurement error is not negligible when used methods Iterative Conditional Modes (ICM). In the previous study, examined the effect of measurement errors on the estimation of the model. Furthermore, in this study the process of calculating a regression model using the S-Plus2000 Software as a computational tool to estimate the regression model. Data will be generated using the program will be carried out in this study, which in turn is used to estimate program regression models with measurement error.