Methods of Associating Segments with Reporting Units Workshop on Area Sampling Frame Methods of Associating Segments with Reporting Units Hassan Serghini Budapest 15-18 October 2012 This presentation is preparation by using references mentioned at the end of this presentation NOM Prénom Fonction
Weighted segment method 17/01/2019 CONTENTS: Introduction Closed segment method Open segment method Weighted segment method
Introduction It is not possible to delineate segments so that no farms will overlap segment boundaries: practical problems of associating farms with segments. 3 methods are used to associate reporting units (holdings and tracts) to the sampling units (segments) and define the value of the survey variables for each segment; Closed segment, open segment, and weighted segment;
Introduction These three methods refer to three different ways of defining an area sampling unit; Closed segment: variables for which tract is the reporting unit; Open segment and weighted segments: variables for which the holding is the reporting unit;
Closed, open and weighted segments B G C D F Segment limits, Tract limits, Farmer residence
Closed segment method Closed segment: A, B, C, D (2), E, F, G tracts ; The value of a variable in a segment is the sum of its value of the tracts of the segment: the totality of a holding is not needed except for holding totally included in the segment; Data on specific items or activities, are collected within the boundaries of the segment either by personal interview with the holders or by checking on the ground and identification in the aerial photograph and their estimation;
Closed segment method The closed segment method is independent of the definition of the holding and it is very simple; For data related holdings closed segment method is not applicable. It is not efficient in estimating rare items or items concentrated in small geographic areas; Response errors can occur trough misidentification of segments and respondent unfamiliarity with the tract reporting concept;
Closed segment method Crop areas can be estimated very precisely: coverage errors are low as tract data can be observed accurately; Reporting errors are low because of the possibility of verifying respondent response on the aerial photos; But no variables related to the whole holding can be surveyed using the closed segment.
Open segment method Holdings: A, C, D, G; Open segment method associates a segment to all holdings with headquarters included in the segment; Need for a clear rules to define a unique reference point for each holding. The holder must be uniquely identified. The usual definition of headquarter is the residence of the holder;
Open segment method Difficulties: one-to-one correspondence between holders and holdings and locating the residence of the holders in urban areas. More than one person or household are involved in the operation of the holding. Urban areas have to be included in the area frame: Difficulty of screening for holders in densely populated areas (underestimation of the farm population);
Open segment method Probable high level of non respondent; Estimates all agricultural variables pertaining to the holding (livestock, grain stock, number of holdings and other economic characteristics;
Weighted segment method Holdings of A, B, C, D, E, F and G Reporting unit: holding associated with any land area in the sample segment; Variable in each tract is defined as the value of the variable in the holding multiplied by a factor equal to the ratio between the area of the tract divided by the area of the holding: E/(E+E’); Weighted segment can be used for all survey variable related to the holding; No need for the location of the holding headquarters;
Weighted segment method Necessity of defining the holding and finding an informant that can provide information for the totality of the holding; No need for the enumeration of urban segments with no agricultural land or activity; In the case of communal land, where livestock activities take place, it is necessary to record these activities by building another frame. Farmers may have tendency to underreport total holdings areas outside the segment, this biases the weighted estimates upward;
References used Earl E. Houseman, Area Frame Sampling in Agriculture, Statistical Reporting Service • U.S. Department of Agriculture, SRS No. 20; FAO, Multiple Frame Agricultural Surveys, Volume 1, Current Surveys Based on Area Sampling Methods, FAO Statistical Development Series, FAO LIBRARY AN: 375713, Rome 1996; FAO, Multiple Frame Agricultural Surveys, Volume 2, Agricultural Survey Programmes Based on Area Frame or Dual Frame (Area and List) Sample Designs, FAO Statistical Development Series, FAO LIBRARY AN: 401697, Rome 1998;
شكرا Thank you Merci