||Constructing samples for characterizing household food security and for monitoring and evaluating food security interventions: theoretical concerns and practical guidelines
||Reliable information on household food security is a prerequisite for the accurate and
effective design, monitoring, and evaluation of development projects. In part due to the
commitment, on the part of many development agencies, to work in marginalized areas, this
information is often either not available or grossly out-of-date. But collecting data is not a
costless exercise. This guide discusses how random sampling techniquesmethods that use
some mechanism involving chance to determine which farms, households, or individuals are to
be studiedcan economize on the costs of gathering information while increasing the likelihood
that it will be both accurate and available in a timely fashion.
The guide has been divided into two parts: an overview and a series of technical
appendices. The overview is written in a largely nontechnical fashion and is designed to be
accessible to a wide audience. It begins with a brief explanation of why random sampling
techniques are a powerful means of obtaining information on household characteristics such as
food security. It then takes the reader through a step-by-step process of constructing a random
sample. Having outlined these issues, a worked example is then presented. The second part
consists of a number of technical appendixes that extend the discussion found in the overview.
These are designed for individuals with some familiarity with statistics. The reader interested in
pursuing the issues raised in this guide is encouraged to consult Bernard (1988), Casley and Lury
(1987), Casley and Kumar (1988), Devereux and Hoddinott (1992), and Newbold (1988). More
technical discussions are found in Kish (1965) and Cochran (1977).
||Training/Learning Support Material;
||Improved food security and better nutrition