APLIKASI BOOTSTRAP PADA ANALISIS REGRESI UNTUK DATA KECELAKAAN KERJA

  • Toto Hermawan

Abstract

To find out the relationship between two or more variables, regression analysis can be used. The definition of regression analysis itself is a data analysis method that utilizes the relationship between two or more variables. One concern in regression analysis is one of them is the standard error of estimation of the regression coefficient. In a regression there is already a formula for estimating standard errors. In addition, the standard error can also be estimated by the resampling method, which is bootstrap. Bootstrapping is very useful as an alternative to estimating parameters or standard errors when researchers feel hesitant to meet the assumptions in their data, for example the data are not normally distributed. In addition, bootstrapping is also useful when parametric inference requires a very complicated formula for calculating standard errors (Widhiarso, 2012). In this paper we will compare the standard error estimates obtained through existing formulas with the standard error estimates obtained through bootstrap resampling.

References

Efron, B. & R. J. Tibshirani. 1993. An Introduction to the Bootstrap. New York: Chapman and Hall.

Krisnawardhani, Tanti. Salam, Nur. Anggraini, Dewi. 2010. Analisis Regresi Linear Berganda dengan Satu Variabel Boneka (Dummy Variable). Program Studi Matematika Universitas Lambung Mangkurat. Banjarbaru.

Seber. George A.F & Alanj.LEE. 2003. Linear Regression Analysis. Canada.

Widhiarso, Wahyu. 2012. Berkenalan dengan Bootstrap. Fakultas Psikologi UGM. Yogyakarta.

Published
2019-01-07
How to Cite
Hermawan, T. (2019). APLIKASI BOOTSTRAP PADA ANALISIS REGRESI UNTUK DATA KECELAKAAN KERJA. Academy of Education Journal, 10(01), 55-62. Retrieved from https://jurnal.ucy.ac.id/index.php/fkip/article/view/271