著者

沼田 眞

出版者

公益社団法人 日本植物学会

雑誌

植物学雑誌 (ISSN:0006808X)

巻号頁・発行日

vol.63, no.745, pp.149-154, 1950

被引用文献数

2

---

1) I investigated the quantitative method analysing vegetation according to the theory of small samples.<BR>2) It is recognized by most investigators that we ought to use the random sampling in our field survey, but we often based its method on subjective randomizaton.<BR>3) Sampling error is controlled by vegetational homogeneity and sampling ratio. And I could not compare several sampling methods by means of sampling error.<BR>4) The best stratification is the worst clusterization. Those order is expressed by coefficient of interclass correlation (&rho;') and also the quality of sampling method is decided by the coefficient.<BR>5) The stratification by means of a dominant does not coincide with the one by means of the whole physiognomy.The latter is better.<BR>6) The significance of difference between several sampling methods is indicated by analysis of variance.Relative number of sampling units required for a given accuracy according to samling methods is calculated from average comparable variance.