Author : Martinez, Beverly F.

Summary : The study investigated response errors in rice and corn surveys conducted by the Bureau of Agricultural Economics (BAEcon) through its Integrated Agricultural Surveys (IAS). Measurement of the reliability of the data was made possible by the design of thereliability of the data was made possible by the design of the questionnaire which repeated questions on specific items for a particular reference period in two or more IAS rounds. It was found out that for the subsample of matching households, summarized results were more stable over different rounds than individual responses. The variability in responses of the same household was studied by examining the number of unequal responses, the arithmetic and absolute differences between responses for different rounds, and the response deviations from the arithmetic mean. The extent to which the responses varied from one round to another was related to the item considered for investigation, to some characteristics of the farm household, and to the difference in time between the dates of interview. Measures of intraclass correlation and response variance were also estimated.