ABSTRACT
This chapter summarizes the limitations on designs which have emerged from the discussion of the characteristics of the attributes which must be measured by farm survey. It provides some pointers to survey organization in the field, emphasizing the selection of and conditions of work for enumerators and on thorough preparations in the community to be investigated. The farm population under investigation should fully accept the implications of the survey before data collection begins. Various types of frames have been used for farm survey work, many out of expediency. While preenumeration is superior to any existing frame, the cost and time involved need careful consideration. For farming systems where conditions allow the use of either limited- or frequent-visit techniques, the limited techniques offer advantages of speed and cost. The problem of selection criteria is a repetition of that in selecting a representative farm: finding criteria to reflect the values of the population over a wide range of attributes.
