The problem of plot size in large-scale yield surveys

Abstract

This paper gives the results of the three investigations carried out on a district-wide scale for comparing the efficiency of small size plots with those of the order of 1/80th of an acre used under the existing official procedure in India. The first investigation was on wheat in the United Provinces, and the other two on paddy (rice) in the provinces of Bihar and Madras. The experiments were conducted by the staff of the departments of Revenue and Agriculture posted in the districts, who ordinarily carry out crop-cutting experiments under the official orders. The results of all the three investigations show that there is a definite risk of obtaining over-estimates of the average yield with small size plots. In contrast, plots of the order of 1/80th of an acre appear to be free from bias. The results also show that small plots fail to furnish unbiased estimates of the different components of the true variance.