Robust Regression Estimation Mm Bisquare On The Factor of DHF In East Java

Abstract

Background: MM Estimation is an alternative of regression on the data outlier. East Java is one of area that has CFR > 1%. There are some factors that can affect IR DHF such as climate change, community behavior, and environmental. Aims: to know the regression factors that affects the IR DHF in East Java in 2017. Settings and Design: This research has non-reactive (unobtrusive) design by using secondary data. Methods and Material: The independent variables are population density, percentage of PHBS, percentage of healthy home, and precipitation of East Java in 2017. The dependent variable is IR DHF in 2017.The population is 38 regencies and cities in East Java with the sample are 35 regencies and cities by simple random. Statistical analysis used: Regression on MM estimation with Tukey’s Bisquare. Results: Regression MM-Bisquare estimation is the effective on the data outlier. The regression model is �� ̂= 22.325 + 0.010 (population density) + 0.207 (% of PHBS) – 0.527 (% of healthy home) + 0.006 (precipitation). The density of population and percentage of healthy home (p value < 0,05) affect the IR DHF of East Java in 2017 on the significant standard 0.05, while the percentage of PHBS and precipitation (p value > 0.05) have no effect on the IR DHF in East Java in 2017.Conclusions: Regression MM-Bisquare estimation is used as the alternative regression on the data outlier. The factors that affect IR DHF can be the main focus of DHF prevention program.