ABSTRACT

CHAIRUNNISA EKA MAULIDA, “Estimate The Probability of Obesity to children elementary school in Samarinda ulu Based on Risk Factor” . Guided by Mr. Drs. Suyitno, M.Sc as first advisor and Mr. Syaripuddin, S.Si, M.Si as second advisor.
In statistic science, known all of regression method. Regression method the most popularity is linier regression where is this method using if data analysis have interval scale, ordinal and nominal. If the measure scale is nominal, especially on dependent variable, therefore exactly using regression logistic method.
The purpose of this research is looking for probability obesity a children probability based on risk factor age (X 1), sex-matched (X), mother educated level(X3), Father educated level (X4), Pocket money (X52), Genetic factor (X), Activity(X7) and Eat quantity (X) among Samarinda ulu elementary school children using logistic regression analysis. From the eight variables, the variable which influence the obesity is pocket money, Activity and eat quantity. 8
Based on the regression model chosen with the backward method, the logistic regression model concurred is :
Y =1  e e  XXX 759.0946.0314.0466.2 875  XXX 759.0946.0314.0466.2 875
From the regression equation obtained obesity a children probability based on risk factor pocket money, activity and eat quantity.
Based on odds ratio therefore a quotation children probability with much pocket money is 1.369 time if to compare a children probability with some money to obesity, a children probability too little activity is 0.388 time if to compare a children probability too much activity to obesity, a children probability with eat quantity to much is 2.136 time if to compare a children probability with a little eat quantity to obesity.
Key words : logistic regression, obesity, factor 6