The limited sample in survey activities is an obstacle to providing data in smaller domains and areas. Small Area Estimation (SAE) can be solve this problem. However, this indirect estimation technique requires the assumption of a linear relationship between the mean of a small area and the accompanying variables. This problem can be solved by using a nonparametric approach, one of the nonparametric approaches that can be used is the Nadaraya-Watson Kernel base. To facilitate the implementation, the researcher built a Package R for Small Area Estimation of a nonparametric approach based on the Nadaraya-Watson Kernel with the package name "saekernel". The results show that the "saekernel" package that has been built is suitable and feasible to use. The package that has been built is also applied to the BPS survey, which is to estimate per capita expenditure at the sub-district level in the D.I Yogyakarta Province based on data National Socio-Economic Survey (Susenas) from the March 2019.
                        
                        
                        
                        
                            
                                Copyrights © 2022