LSMS Sample Template

Session IV: The practice of
sampling for household surveys
Kristen Himelein and Juan Muñoz
Module 2: Sampling for Surveys
March 12, 2012
Exercise 1: Taking a sample of PSUs with
Probability Proportional to Size (PPS)
• The Ministry of Social Protection is conducting a survey to evaluate
the impact of an additional education program for school-age
children in the Forest region
• A sample of 1,056 households will be selected in two stages:
– First, 88 Enumeration Areas (EAs) will be selected with PPS
– Second, 12 households with children will be selected in each EA
• Open the Excel book “Exercises.xls” and activate spreadsheet
“Forest”
• A written explanation of this exercise can be found in Chapter XVG of “Household Sample Surveys in Development and Transition
Countries” (pps 319-332). Downloadable at
http://unstats.un.org/unsd/publication/seriesf/seriesf_96s.pdf
Exercise 2: Household
selection
• You are the leader of one of field teams
• Your team has completed the household listing in one of
the 88 PSUs selected in the Forest region. The full listing is
four pages long.Your interviewers flagged the eligible
households with a tick  on column “School-age children?”
• You now need to complete the cover page and randomly
choose the 12 households to be interviewed, following the
instructions
• Pretend that one of your classmates is the local facilitator
who helped your team identify the area and its boundaries
• This random selection method is called “Circular
systematic sampling”
Exercise 3: Choosing staff members
• You are the leader of the teams responsible for
interviewing three staff members in each health
facility
• You have asked one of your interviewers to fill a
form with the names of all eligible staff members,
sorted alphabetically
• Choose three of them, following the instructions
in the form
• Compare your choice with those of your
classmates