<![CDATA[The science of statistics is required to continue to develop following the times, because the more the characteristics of the data available in the field are increasingly diverse. The more types of data, the more statistical analysis methods are developed, including hierarchical structured data. P3KE data is new data complementing DTKS which is the basis for the government in distributing social assistance. P3KE data becomes a reference in determining KPM BLT DD. Wanasari is a village that is accustomed to determining KPM BLT DD based on the results of deliberations at the kedusunan level (Musdus). In Wanasari Village, there is often a problem of inconsistency between the KPM candidates from the Musdus and the P3KE data from BKKBN provided through the Cianjur District government. Therefore, it is necessary to analyze the components that have a significant effect on the Welfare Decile of Wanasari Village P3KE data. The data is considered to be hierarchically structured with ordinal response variables. Therefore, multilevel ordinal logistic regression analysis with Bayesian parameter estimation will be used to obtain the best model. Normal (0.10) and Cauchi (0.2.5) priors were compared to find the best model. The results show that the P3KE data of Wanasari Village is hierarchical data because the results of two-level logistic regression analysis are better than one level. The study also concluded that Bayesian parameter estimation is better when using Cauchy prior (0.2.5) both for β coefficient estimation and inter-departmental diversity estimation. The best model obtained is able to explain the diversity between neighborhoods by 1.07 and has an accuracy of 63.23%. Predictor variables that have a significant effect include civil registration equivalents, having money/jewelry/livestock/etc. saved, wall type, cooking fuel, drinking water source, stunting risk, and number of households.]]>
Copyrights © 2025
