Utilización del método de encuestas en la recopilación de estadísticas agrícolas

Abstract

Some of the problems encountered and the progress made in establishing an agri-cultural estimate service in Puerto Rico are reported During 1950 four interview surveys were conducted. A two-stage sample design was followed in which the first stage units were small political subdivisions (barrios or groups of barrios) and in the second stage units were farms These farms were selected from the first stage units previously drawn on a stratified geographical basis. Field work lasted around three weeks and from 25 to 45 interviewers were hired Special attention was given to the selection and training of interviewers and supervisors. Farmers were asked for informa tion as to size and location of farm, animal inventories, acreage in crops, land use and farm production by enterprises. A check was made with the 1951 agricultural census on a sample basis. Time it took to obtain lists of farmers in selected barrios, difficulty in making listing within the "barrios" in a systematic form, and discrepancies between names appearing in the listings and the questionnaires, are discussed.

From the census coverage check it was found that interviewers failed to list around 20 to 25% of all the farms, from 33 to 44% of the plots, and included erroneously in the lists about 12 to 16% of the names. It was also found that in about 4 to 5% of the cases, the names listed by interviewers were not the real farm operators. The percentage of distribution of farms by size was practically the same in the census and in the survey. Cattle inventories were checked with the records of a tick eradication program and jurther field checks were made to corroborate figures. While the number of farmers having cattle agreed in the two sources, it was found that the number of beads reported per farm was on the average 1.07 beads less that shown in the tick records. A check on tobacco production was made with the planting quota allotments and with the records of cooperatives and buyers. In velation to the Buyer's Records, farmers under reported on the average their tobacco crop by about 16% A comparison of the total crop handled by wholesalers and an expansion of the data gathered from farmers showed that the latter is 27% smaller than the real crop harvested. The methodological research is being continued.