作者: J�nos Podani
DOI: 10.1007/BF00031684
关键词: Mathematics 、 Statistics 、 Statistical hypothesis testing 、 Data type 、 Procrustes analysis 、 Data set 、 Multivariate statistics 、 Plot (graphics) 、 Multiple comparisons problem 、 Vegetation
摘要: The methodology of comparing the results multivariate community studies (resemblance matrices, ordinations, hierarchical and nonhierarchical classifications) is reviewed from two viewpoints: basic strategy measure employed. determined by 7 choices concerning type results, consensus methods or resemblance measures, hypothesis testing exploratory analysis, lack presence reference basis, data set congruence algorithmic effects, number factors responsible for differences among properties considered in comparison. Included a brief summary applicable to vegetation studies. Examples grassland survey demonstrate utility comparisons evaluating effects plot size, type, standardization, taxonomic level species on classifications ordinations.