Page 49 - Contributed Paper Session (CPS) - Volume 1

CPS863 Abubakar S. A. et al.

Redesigning of the corn production survey in the
Philippines
1,2
1
Abubakar S. Asaad , Sarah B. Balagbis , Jay M. Manlapaz 1
1
3
*Melanie C. Estrada , Divina Gracia L. del Prado , Erniel B. Barrios 4
1 Statistical Methodology Unit, Philippine Statistics Authority
2 College of Public Health, Department of Epidemiology and Biostatistics, University of the
Philippines Manila
3 Economic Sectoral Statistics Section, Philippine Statistics Authority
4 School of Statistics, University of the Philippines Diliman

Abstract
Redesigning of the Philippine Statistics Authority’s (PSA’s) nationwide
quarterly Corn Production Survey (CPS) was done to reflect the current
behavior of corn production in each province. Simulation results show that
stratified Lavalle-Hidiroglou (LH) algorithm is the best stratification with
probability proportional to size systematic (PPS-SYS) and simple random
sampling without replacement (SRSWOR), as the best selection methods for
sample barangays and sample households, respectively. Further, design-based
estimation is recommended over bootstrap method, in terms of precision and
accuracy.

Keywords
Sampling frame; sampling design; design-based; bootstrap method

1. Introduction
The CPS is a nationwide quarterly agricultural survey conducted by the PSA
during the months of April, July, October and December. The CPS aims to
generate provincial estimates and forecasts on corn production, area and yield
as inputs for government policies and programs on corn1. The CPS uses the
1991 Census of Agriculture and Fisheries (CAF) as a sampling frame, with the
total farm area based on the 1980 Census of Agriculture (CA). This served as
the stratification variable and the number of strata is either 10 or 5 depending
on domain size. The coverage of the survey include Isabela, Laguna and
Bukidnon and National Capital Region (NCR)2. The histogram of corn
production and corn farm area (1st Semester) in the Philippines using the 2013
CPS3 data and 2012 CAF4 respectively, were positively skewed for each
province. The correlation coefficient of corn production and corn farm area by
quarter is from moderate to high.
The redesigning of the survey was done to reflect the current behavior of
corn production in each province in the country. The redesigning of CPS is
aimed to come up with a better sampling designs that will generate reliable

38 | I S I W S C 2 0 1 9