Wednesday 19 July 2 p.m. - 3:20 p.m. (Canada/Eastern)
Abstract
The Producer Price Survey (PPS) for Manufacturing is a nationwide survey conducted by the Philippine Statistics Authority (PSA) to gather monthly data on producer prices of industry products produced by sample manufacturing establishments. The collected data are used in the generation of the monthly Producer Price Index (PPI) for the manufacturing sector. Sample establishments with no collected producer prices are inevitable due to various reasons and are therefore critical in the estimation of PPI. Imputation techniques such as the Last Observation Carried Forward (LOCF) and geometric mean are applied to reduce the bias in the estimation of producer price of non-responding sample establishments. However, it was observed that the preliminary and revised estimates of PPI had large discrepancies when using varying imputation technique for non-responding establishments during the preliminary estimation. The research study compared LOCF with geometric mean in terms of accuracy through simulation. Results show that in general, LOCF yield higher accuracy than geometric mean in imputing producer price of non-responding establishments.
