Sampling with Probability Proportional to Aggregate Size Using Nonparametric Bootstrap in Estimating Total Production Area of Top Cereals and Root Crops Across Philippine Regions

Year: 2012       Vol.: 61       No.: 1      

Authors: Maria Sofia A. Poblador; Iris Ivy M. Gauran

Abstract:

Cereal and root crop production are of primary interest to the country’s agricultural industry. The need to obtain reliable estimates of total area of production is therefore crucial. This paper examines the Sampling with Probability Proportional to Aggregate Size (PPAS) in terms of unbiasedness and precision of estimates as compared to two known sampling designs, Simple Random Sampling without Replacement (SRSWOR) and Sampling with Probability Proportional to Size Without Replacement (PPSWOR). Among several crops included in the 2002 Philippine Census of Agriculture, rice and corn are considered for cereals, while cassava and sweet potato for root crops. Crop area, which is believed to be highly correlated with total production area, is utilized as auxiliary information. Estimates of total production area are obtained under 1%, 5% and 10% sampling rates. To be able to evaluate precision of PPAS estimates, nonparametric bootstrap variance estimation is performed. It was found out that PPAS estimates are generally better than the two other sampling designs when it comes to precision but almost at par when it comes to unbiasedness.

Keywords: probability proportional to aggregate size sampling; probability proportional to size sampling; simple random sampling; nonparametric bootstrap estimation

Download this article: