在实践中,往往发生测产与实产之间有较大的差误,即测产误差大于5%。我们发现误差主要来源于决定产量三因素中的每穗粒数。这是因为每穴稻株的内部、穴与穴之间,由于秧苗素质和环境条件的差异,每穗粒数变化较大,如取样较少(少于50穗),其代表性就不够。而按数理 In practice, there is often a big discrepancy between production and production, that is, the measurement error is greater than 5%. We found that the error mainly came from the grain yield per spike of three factors. This is because the number of grains per spike varies greatly between the inside and the hole of each rice plant due to differences in the qualities of the seedlings and the environmental conditions. If sampling is less (less than 50 spikes), the representativeness is not enough. And by mathematical