Local Polynomial Smoothing untuk Mengatasi Masalah Age heaping Data Jumlah Kematian Menurut Umur Hasil Sensus Penduduk 2010

Abstract

Age data collected through census or survey often experience errors, especially in reporting age data (age missreporting) related to the selection of certain numbers (digit preference) which are usually ages that end with the number 0 (zero) or 5 (five). The existence of this preference digit which results in the age distribution becoming enlarged or accumulated on the numbers ending with 0 and 5, which in demography is known as age-heaping. Evaluation of the quality of age data from the results of the 2010 Population Census in this study focused more on data on the number of deaths by age is very necessary. This data will be evaluated using the Whipple Index (IW) and Myers (IM) Index. The technique of smoothing local polynomials in non-parametric regression with IW and IM resulted in the conclusion that there was age heaping in the data that became the object of this study so that data needs to be smoothed with local polynomial techniques.