PERBANDINGAN HASIL REGRESSION IMPUTATION DAN MULTIVARIATE IMPUTATION BY CHAINED EQUATION PADA PEMBENTUKAN MODEL STRUKTURAL (Analisis Faktor yang Mempengaruhi Umur Pertama Kali Hubungan Seksual Pada Remaja Menurut Survei RPJMN 2015)

BERLIANA DEVIANTI PUTRI, 101514153016 (2018) PERBANDINGAN HASIL REGRESSION IMPUTATION DAN MULTIVARIATE IMPUTATION BY CHAINED EQUATION PADA PEMBENTUKAN MODEL STRUKTURAL (Analisis Faktor yang Mempengaruhi Umur Pertama Kali Hubungan Seksual Pada Remaja Menurut Survei RPJMN 2015). Thesis thesis, Universitas Airlangga.

[img]
Preview
Text (abstrak)
abstrak.pdf

Download (26kB) | Preview
[img] Text (fulltext)
full text.pdf
Restricted to Registered users only until 17 January 2021.

Download (1MB)

Abstract

Data collection activites have a higher risk of missing data. Missing data may produce biased estimates and standard errors were greater, so the imputation technique is needed. The aim of this study was analyze and compare between regression imputation and multivariate imputation by chained equation (MICE) method then establish a model of factors influencing the timing of first sexual intercourse among adolescent based on Survey of RPJMN 2015.. This study was non-reactive study and used raw data taken from BKKBN East Java Province. The population was all respondents of Survei RPJMN 2015 which had dating and pre-marital sexual intercourse. Simulation data of this study was from complete data which eliminated as much as 5%, 10%, and 15%. The comparative parameters was MSE, GFI, and RMSEA. Friedman Test analysis showed that there was no different between the imputed data and the original data. Based on MSE analysis, MICE better than Regression Imputation. Based on GFI and RMSEA values, Regression Imputation as good as MICE in missing data 5% and 10%, but in missing data 10% MICE better than Regression Imputation. The conclusion was MICE better than Regression Imputation. Hopefully, this result could help other researchers, especially for BKKBN, not to deleting unit which has missing data. The advantages of the MICE method are applicable to not only for normality data, and can be applied to vacancies up to 15%.

Item Type: Thesis (Thesis)
Additional Information: KKC KK TKM 02/18 Put p
Uncontrolled Keywords: Missing Data, Regression Imputation, MICE, Path Analysis
Subjects: K Law > K Law (General) > K1-7720 Law in general. Comparative and uniform law. Jurisprudence > K(520)-5582 Comparative law. International uniform law > K3566-3578 Public health
Divisions: 10. Fakultas Kesehatan Masyarakat > Magister Ilmu Kesehatan Masyarakat
Creators:
CreatorsEmail
BERLIANA DEVIANTI PUTRI, 101514153016UNSPECIFIED
Contributors:
ContributionNameEmail
ContributorHari Basuki N, Dr.,dr.,M.KesUNSPECIFIED
Depositing User: Unnamed user with email indah.fatma@staf.unair.ac.id
Date Deposited: 17 Jan 2018 00:24
Last Modified: 17 Jan 2018 00:24
URI: http://repository.unair.ac.id/id/eprint/69054
Sosial Share:

Actions (login required)

View Item View Item