Setting Up the Frame - Sampling Spatial Units for Agricultural Surveys - page 94

units of type 1 is a more difficult task. Delays in the production of estimates may
happen.
In the case of units of type 2, the spatial frame construction and management and
sample selection are similar but simpler than with units of type 1. Also in this case
delays in the production of estimates may happen.
Units of type 1 and type 2 have an economic advantage with respect to units of
type 3, since in one visit of the enumerator it is possible to collect information for a
wide area. Moreover, geometric errors resulting from incorrect positioning can be
minimized or corrected because the overlay of satellite data is easier.
An advantage of spatial units of type 3 is that the spatial frame (in this case
spatial point frame) construction and the sample selection is simpler than for a
spatial sample design with units of type 1. Estimation methods are simpler than with
units of type 1 and 2, but usually they are less efficient given the sample size n . Data
collection and processing are easier than the other two kinds of units. Another
advantage is that some survey variables are point-specific and cannot be measured
over extended polygons.
5.3 Main Advantages and Disadvantages of Different
Frames Typologies
On the basis of the broad classification of the statistical units outlined in the
previous paragraph, the frames can be divided in two basic types: list frames and
spatial frames.
A list frame is a list of all the statistical units in the survey population and the
frame components are simply lists of addresses for holders or farmers. The agri-
cultural survey design in this case is called list-sampling design. List frames are the
most common type of frame found in agricultural probability sample surveys
(Ferraz 2013 ). If their component units are clusters, multistage sampling design
must be adopted to select elements (holders). They can be set up with information
collected from the most recent agricultural or population census, administrative
data, previous surveys or a combination of several data sources (Wallgren and
Wallgren 2007 , 2010 ).
A spatial frame is a complete and exhaustive list of non-overlapping geographic
areas and the frame component units are segments or portions of land. These
polygons can be defined by using some geographic features such as rivers and
streets. The selection procedure applied to this list is generally called spatial
sampling design (see Chap. 7 ) .
In agricultural surveys, farms are traditionally sampled on a list (often a census)
of farms. When a satisfactory list frame does not exist, and it would be too
expensive or complex to create or when it is too expensive or complex to maintain
it, a spatial frame may be the best alternative (Kennel 2008 ). Generally, agricultural
censuses are made every 10 years so there may be a substantial difference between