我邦に於ける水稻品種の生産力に關する研究 (続) : 再び基準偏差法に就て

抄録

1. This report is the methodical development of my long investigation on the yielding efficiency of Japanese rice variety. I use this method of calculation in some past of my investigation. 2. I use the "Standard deviation method" as the relative expression of heading period, stem length, numbers of stems and yield per tan of rice variety and I think it is a convenient method and it is a method or reducing the eror to use the standard deviation method. And I proved this by the following example. 3. In a group of variants I call the deviation of each variant from the representative value of variants (I call this Standard value, and it is convenient to use the mean value, ) standard deviation. When I need the relative value only, and when I can regard the representative value of variants as the same, by useing the "Standard deviation method" I can eliminate the effect of the deviation of mean, such as year factor effect and soil heterogenity effect and can express the relative factors with less errors and by more exactness. 4. In calculating the error of "Standard deviation method" I take the standard deviation value as a variant, calculate the standard deviation, proobable error, and probable error of mean by the normal metbod. 5. The reliability of the estimate, the error, calculated by "Standard deviation method" is normaly increased by a great scale. In my example it increased twice. 6. I think, the "Standard deviation method" may be inferior by the accuracy to the "Student method", but I thiv it is more conenient and more distinct. 7. I think, the "Standard deviation method" does not abandon other method of calculation. The "Standard deviation method" can express the factors accurately by the co-operation of other method, such as the ordinaly method useing absolute number.