Development of an Integrated Agricultural Survey (IAS): The Case of the Backyard Livestock and Poultry Survey (BLPS)
Relative to the approaching availability of the 2002 Census of Agriculture (CA) results, the Bureau of Agricultural Statistics (BAS) is in the midst of planning and preparation of a major redesign in its agricultural survey. At present, BAS employs a master sample designed using the results of the 1991 Census of Agriculture and Fisheries (CAF). Said master sample employs a stratified multi-stage design and its made up of four independent replicates. This master sample is used by regular BAS surveys, such as the Rice and Corn Production Survey (RCPS) and the Backyard Livestock and Poultry Survey (BLPS). While the RCPS uses the full sample, BLPS only utilizes a quarter sample or simply a single replicate. Using the results of the January 2001 round of the BLPS, estimates of sampling errors were computed based on selected characteristics, such as the total number of animal/poultry raised and total number of animals/poultry disposed. The results show that for the BLPS, a single replicate of the current master sample allows the generation of national level estimates of adequate precision. In some cases, the design permits the generation of precise estimates at the regional level and in few cases even at the provincial levels. Further, this study will look into the perceived weaknesses of the current BLPS. Recommendations were also made to improve the current master sample design.