水稲圃場試験の生育調査に関する考察
スポンサーリンク
概要
- 論文の詳細を見る
In this paper the authors investigated the variance of height and number of stem of rice plant, and also considered the relation of sample size and reliability, comparing some sampling methods. (1) The variance of height and number of stem are almost constant in various species and in various stage of growth expecting every early stage of growth. The standard deviation of rice plant is 3〜5 (cm) in height, 2〜3 in number of stem, so the deviation is 5〜7% of its height and 20〜30% of its number of stem. (2) The relation of sample size and confidence interval of height and number of stem are considered in one sample field in which the deviation and mean are already known. Results are shown in Table 3. From this results the confidence interval at ten sample size is X^^- ± 2.2(α=10%) in height, X^^- ± 1.7 in number of stem, so that the comparision of growth from this sample size has considerable risk. Between plots to detect 1 cm difference of height 120 samples must be investigated. But it seems that the investigation of 15〜30 samples is sufficient for ordinary field experiment of rice plant. (3) Three sampling method-simple random sampling, stratified random sampling and systematic sampling-are compared. 10 samples are sampled by each method from the field, and calculated the confidence interval. It is interesting problem that the systematic sampling method, which is most easy for resarch workers, has not necessarily low reliability.
- 社団法人日本土壌肥料学会の論文
- 1955-02-15