PENERAPAN MODEL REGRESI ORDINAL MAJEMUK DALAM MENAKSIR PARAMETER REGRESI

HERLINA JUSUF, 090013907 M (2002) PENERAPAN MODEL REGRESI ORDINAL MAJEMUK DALAM MENAKSIR PARAMETER REGRESI. Thesis thesis, UNIVERSITAS AIRLANGGA.

[img]
Preview
Text (ABSTRAK)
gdlhub-gdl-s2-2006-jusufherli-731-tkm_18-03.pdf

Download (788kB) | Preview
[img]
Preview
Text (FULL TEXT)
gdlhub-gdl-s2-2006-jusufherli-731-tkm_18-03.pdf

Download (2MB) | Preview
Official URL: http://lib.unair.ac.id

Abstract

Relation between one or several explanatory variables with ordinal scale and response variable, also with ordinal scale, in ordinal multiple regression (OMR) may be made by mean of threshold approach or proportional odds. The objective of this study was to apply ordinal multiple regression model in estimating regression parameter in predicting ordinal response, to learn testing technique, either partial or simultaneous, and to identify variables that have influence on the response. Data used in this study were those on immunization practice, in which immunization completeness served as the dependent variable (1), while the independent variables, which were expected to influence the immunization completeness, were respondent's age (X,), education (X2), number of children (X3), level of knowledge (X4), attitude (X5), and information obtained from promotion attended (X6). Partial analysis results showed there were four significant variables, while two variables, i.e., age and attitude, were not significant. Analysis results of simultaneous test revealed an estimation value that the six independent variables simultanoeusly influenced the dependent variables, i.e., children immunization completeness. At the end of parameter test to identify variables that mostly influenced response variable, i.e., immunization completeness, there were only two variables that influenced most, i.e., respondent's knowledge on immunization and information obtained from promotion attended, while the other four variables were not significant. The final models of ordinal multiple regression were as follows 1. Logit 1 model, less complete immunization status : g, =2.067 - 0.9588 Xa(2)– 2.5825 X4(3) -1.2160 X6(3) – 2.1160 X6(4) – 3.5527 X6(5) 2. Logit 2 model, most complete immunization status : 9, =4.773 - 0.9588 X4(2)– 2.5825 Xa(3) – 1.2160 X6(3) – 2.1160 X6(4) – 3.5527 X6(5) Interpretation with odds ratio revealed that respondents with average knowledge on immunization might have complete immunization status with an increase of 0.9588 (-In 0.38) times, and respondents with adequate knowledge on immunization might have complete immunization status with an increase of 2.5825 (-In 0.08) times than those who had less knowledge on immunization. As to information obtained from promotion attended, those who attended three times had an increase of complete immunization status of 1.2160 (-In 0.30) times, four times 2.1160 (-In 0.12) times, and five times 3.5527 (-In 0.03) times than those who attended promotion only twice.

Item Type: Thesis (Thesis)
Additional Information: KKC KK TKM 18/03 Yus p
Uncontrolled Keywords: Threshold, proportional odds, ordinal multiple regression
Subjects: Q Science > QA Mathematics > QA319-329 Functional Analysis
R Medicine > RA Public aspects of medicine > RA1-1270 Public aspects of medicine > RA421-790.95 Public health. Hygiene. Preventive medicine
Divisions: 10. Fakultas Kesehatan Masyarakat > Magister Ilmu Kesehatan Masyarakat
Creators:
CreatorsEmail
HERLINA JUSUF, 090013907 MUNSPECIFIED
Contributors:
ContributionNameEmail
ContributorKuntoro, H.,dr.,MPH.,Dr.PHUNSPECIFIED
ContributorSunarjo, Dr.,dr.,MS.,M.ScUNSPECIFIED
Depositing User: Nn Anisa Septiyo Ningtias
Date Deposited: 2016
Last Modified: 11 Aug 2016 01:28
URI: http://repository.unair.ac.id/id/eprint/35808
Sosial Share:

Actions (login required)

View Item View Item