Scribd
Upload
316 views17 pages

Two Phase Sampling

This document provides an overview of two-phase sampling, which is a sampling design where selection of samples occurs in two phases. In phase one, an auxiliary variable is measured for a large sample. In phase two, a subsample is selected from the phase one sample to measure the variable of interest. Two-phase sampling is useful when directly measuring the variable of interest is expensive but an auxiliary variable is cheap to measure. It allows incorporating auxiliary information to improve estimation. The document provides examples of applications in forestry and agriculture surveys and outlines the basic steps of two-phase sampling and how ratio estimation can be used to estimate population totals.

Uploaded by

Curla Jorge
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
316 views17 pages

Two Phase Sampling

This document provides an overview of two-phase sampling, which is a sampling design where selection of samples occurs in two phases. In phase one, an auxiliary variable is measured for a large sample. In phase two, a subsample is selected from the phase one sample to measure the variable of interest. Two-phase sampling is useful when directly measuring the variable of interest is expensive but an auxiliary variable is cheap to measure. It allows incorporating auxiliary information to improve estimation. The document provides examples of applications in forestry and agriculture surveys and outlines the basic steps of two-phase sampling and how ratio estimation can be used to estimate population totals.

Uploaded by

Curla Jorge
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd

BIOSTATISTICS 212

TWO-PHASE SAMPLING
C ARL A T. JORGE
OBJECTIVES
At the end of the lesson, the students should be able to:
! Define two-phase or double sampling
! Explain how to design a two-phase sample
! Explain how to use two-phase sampling to collect
information for ratio estimation
WHAT IS TWO-PHASE SAMPLING?

! Introduced by Neyman(1938)
! Also called double sampling
! A sampling design where the selection of samples is performed in
two phases:
! Phase I:  auxiliary variable x is observed
! Phase II: study of interest variable y is observed
WHEN IS IT USEFUL?

! When the variable of interest y is relatively expensive to measure,


but a correlated variable x can be measured fairly easily and used
to improve the precision of the estimator of t y .
! Useful in obtaining auxiliary variables for ratio and regression estimation.
! Useful for finding information for stratified sampling.
! To adjust for nonresponse
! To sample rare populations
! To improve sampling frame
DISADVANTAGES

! Time consuming
! More complex to execute and analyse
SOME APPLICATIONS OF TWO-PHASE SAMPLING

! Forest surveys - when travelling to remote areas to make ground


determination is difficult and expensive
! Agricultural surveys - when investigating crops acreage
! Military surveys - when sampling veterans
STEPS
! Phase I:

! a probability sample of n1 units, called phase I sample is taken


! Measure the auxiliary variable x for every unit in phase I sample

! Assuming that auxiliary information is inexpensive to obtain,


phase I sample is generally relatively large. This should provide
accurate information about the distribution of x’s.
STEPS
! Phase II:

! Treat the phase I sample as the population


! Select a probability sample of size n_2 units, called phase II sample
from phase I sample.
! Measure the variables of interest y for each unit in the subsample.

! Auxiliary information from phase I maybe used when designing the phase II
sample since phase I was treated as the population.
CAN’T IMAGINE HOW IT WORKS?
HERE’S AN EXAMPLE.
A forest resource manager is interested in estimating the total number of dead
trees in a 400 acre area of heavy infestation. She subdivides the area into 200
plots of equal sizes and uses photo counts to find the number of dead trees in
18 randomly sampled plots. She then randomly samples 8 plots out of these
18 plots and conducts a ground count on these 8 plots.

Let  x  denote the number of dead trees in the plot by photo count and  y  the
number of dead trees by ground count.
DATA ARE GIVEN AS
PHASE I: 1S T, 18 PLOTS ARE RANDOMLY SELECTED FROM 200 PLOTS OF EQUAL SIZES

x’ is the number of dead trees in the plot by photo count.

PHASE II: 2ND, THE 18 PLOTS FROM PHASE I ARE TREATED AS THE POPUL ATION .
THEN , 8 SAMPLES ARE RANDOMLY SELECTED 8 FROM THE 18 PLOTS.

x is the number of dead trees in the plot by photo count


y is the number of dead trees in the plot by ground count
HOW ARE WE GOING TO ESTIMATE THE NUMBER OF
DEAD TREES?

We can calculate the ratio estimate for the population total.

HOW?
PROCEED TO THE NEXT SLIDE
RATIO ESTIMATION WITH TWO-PHASE SAMPLING
DONEC QUIS NUNC
DONEC QUIS NUNC
WANT TO LEARN MORE ABOUT TWO-PHASE SAMPLING?
CHECK THE REFERENCES BELOW.

REFERENCES
Legg, J.C, & Fuller, W.A. (2009). Two-Phase Sampling.  In Chakraborty, R., Rao, C.R., &
Sen, P. , Handbook of Statistics 28 (pp. 55-70). Elsevier.
Lohr, S. L. (2010). Sampling: Design and Analysis, Second Edition. Boston, MA:
Brooks/Cole.
The Pennsylvania State University (n.d.). Double or Two-Phase Sampling. Retrieved
from [Link]
Thank you!

You might also like