【摘要】 本研究结合近20亩玉米空白试验,通过不同抽样方法,进行了玉米大面积田间测产的研究与探讨。本文引述了已故谭启栋教授提出的典型众数抽样的理论依据和方法步骤。这一方法不仅在林业上适用,而且在农业方面起到同样作用。试验结果表明,典型众数抽样较同时同地进行的随机抽样平均亩产量更接近空白试验所测得的每亩平均产量。从亩产总体平均值95%的置信范围看,典型众数抽样亦较之随机抽样为精确。即使两种抽样方法结果相近似,但典型众数抽样在布点、经费、人力及时间等方面比随机抽样为少,仅及随机抽样的3/5与3/20。本文通过统计估计与分析阐述了典型抽样的误差小于同大小的随机抽样,且是最能代表总体的抽样。最后论述了典型抽样的必要性,并且结合农业生产的特点,提出了改进这一方法的设想。更多还原

【Abstract】 In the experiment,based on blank test with corn on 20 mu of land and test with different methods of sampling,the problem of field estima- tion of corn yield for large area was studied and probed.The paper cites the theoretical basis and procedures for typical mode sampling provided by late Professor Tan Qidong.Tests showed that the average yield per mu obtained by typical mode sampling was more close to the average yield per mu actually measured in blank test than that obtained by random sampling. Judging by the 95% confidence range of the population average per mu yield,typical sampling is more accurate than random sampling.Even when the results under the two sampling systems were similar,typical mode sam- pling saved substantially cost and labor(by 2/5)and time(by 17/20)over random sampling.The paper notes that statistical estimation and analysis showed the error involved in typical sampling was smaller than that in ran- dom sampling of the same volume,and that typical sampling was the one that could best represent the population.Finally,the paper emphasizes the necessity for typical sampling and envisages in the light of the character- istic features of farm production measures for further improvement of the method.更多还原